Attack and Defence

How Creative Thought Develops in a Chess Player

Mark Dvoretsky and Artur Yusupov

With contributions from:

Mikhail Krasenkov
Beniamin Blumenfeld
Sergei Dolmatov
Vladimir Vulfson
Igor Belov

Translated by John Sugden

B. T. Batsford Ltd, London

f;.:s, pw,ht-hed t ~~
C M.-nt ~tet.$.k.y aAdArlti.t Yu ..~'>• l ~◊$
Enghsh:f.!l,.'l!J~ljf' tnm;l,,!)(',i, e Jclm SuiJ~n l '¥i$
R~ed t'9!>

ISBN 0 7134 8214 1

British Library Cataloguing-in-Publication Data.

A catalogue record for this book is
available from the British Library.

All rights reserved. No part of this book may be reproduced, by any

means, without prior pennission of the publisher.

Edited by Graham Burgess and typeset by John and Petra Nunn for
Gambit Publications Ltd, London.

Printed in Great Britain by

Redwood Books, Trowbridge, Wilts for the publishers,
B. T. Batsford Ltd,
9 Blenheim Court,
Brewery Road,
London N7 9NT

A member of the Chrysalis Group plc

A BATSFORD CHESS BOOK

Contents

Symbols 5
Preface (Mark Dvoretsky) 7

Part 1: Calculating Variations 11

1 The Technique of Analysis and Decision-Taking
(Mark Dvoretsky) 11
2 Wandering in the Jungle (Mikhail Krasenkov) 38
3 Visual Imagination and Chess Analysis
(Beniamin Blumenfeld) 46

Part 2: Intuitive Decisions 53

4 How Chess Intuition Develops (Mark Dvoretsky) 53
5 Chess in the Style of Jazz (Sergei Dolmatov) 76

Part 3: The Practical Elements in Decisions 100

6 Practical Chances in Chess (Beniamin Blumenfeld) 100
7 Is it Worth Complicating? (Vladimir Vulf son) 110
8 Thoughts about a Book (Mark Dvoretsky) 127

Part 4: Attack 151

9 Missed Brilliancy Prizes (Artur Yusupov) 151
10 A Controversial Position (Mark Dvoretsky) 176
11 Attacking after Castling on Opposite Wings
(Mark Dvoretsky) 182

Part 5: Defence 199

12 Training Session: Making Difficult Choices (Igor Belov) 199
13 Virtuoso Defence (Mark Dvoretsky) 216
14 Errors and What Lies Behind Them (Mark Dvoretsky) 231
4 Contents

Part 6: Further Examples 238

15 Analysis of a Game (Mark Dvoretsky) 238
16 Some Achievements of Our Pupils (Artur-Yusupov) 251

Solutions to Chapter 1 Exercises 269

Solutions to Chapter 4 Exercises 273
Solutions to Chapter 11 Exercises 276

Index of Players 282

Index of Composers 284
Index of Games 285
Index of Openings 288
Symbols
+ check
++ double check
X captures
0-0 castles kingside
0-0-0 castles queenside

!! brilliant move
! good move
!? interesting move
?! dubious move
? bad move
?? blunder

1-0 the game ends in a win for White

½-½ the game ends in a draw
0-1 the game ends in a win for Black

(n) nth match game

(D) diagram follows
Preface
Mark Dvoretsky

The book you have before you is the fifth and final volume in the series
'School for Budding Chess Champions', based on material from the
Dvoretsky/Yusupov school for gifted young chess players.
Our little school existed for only three years (1990-2). Its sessions were
attended by 10-15 young people. Nearly all of them first joined us at the
age of 12-15 years. Five years on, I am proud to be able to say that eight of
our students have become grandmasters, some of them very strong ones
enjoying world fame. Here are their names: Alexei Alexandrov, Inoa Ga-
ponenko, Vasily Emelin, Vadim Zviagintsev, llakha Kadymova, Sergei
Movsesian.,.Ella Pitem and Peter Svidler. In the very near future I am sure
that Vladimir Baklan and Peter Kiriakov will attain the grandmaster title
too. Hardly any other children's chess school can boast of such a high
success rate.
In listing the school's achievements I am nonetheless perfectly aware
that the successes were gained first and foremost by the students them-
selves and their regular coaches. Was there much that we could teach (for
instance) Peter Svidler - three-times Russian Champion - at the three
ten-day sessions of the school which he attended (the second, fourth and
sixth)? The chief ingredients of Svidler's successes were of course his
own immense talent and the help of his outstanding coach Andrei Lukin.
Yusupov and I saw our role as that of stimulating the young players'
further development; helping them to come to terms with themselves,
with their own strengths and weaknesses; formulating plans for the fu-
ture; discussing the problems they faced at the chessboard, and the means
of overcoming them; pointing out the basic policies and methods for per-
fecting their chess. That was all. It was not an immense amount, but it was
not negligible either. The students' results confirm that this was the right
approach, and that as a whole we performed our task successfully.
Among other things, it was at a session of our school that Peter Svidler
was advised by me to tum to Lukin for assistance.
 8 Preface

This same approach is reflected in our books. We have not attempted to

write textbooks to clarify this or that topic fully and precisely. Our aim
has been to equip our readers with high-quality material and a variety of
ideas conducive to independent thinking and work along the lines we
have mapped out. The ideas are not all our own; they include those_ of
other specialists (particularly trainers working together with us at the
school). Naturally this method of exposition requires readers to adopt an
inventive (and sometimes critical) attitude to the text they are studying; it
is unsuited to lovers of ready-made precepts. To judge from the popular-
ity of our books, a great many chess players like this approach.

The issues discussed in this book are not confined narrowly to chess;
they are all situated on the border between chess and psychology. The
chess player's thoughts at the board, the characteristic processes of
decision-taking in a variety of situations - this, in a nutshell, is the basic
content of the book. Many of the examples that are offered for your atten-
tion are highly complex and ambiguous; they require deep investigation
. of a position, ingenuity and bold, risk-taking actions. In comparison with
its predecessors, then, this volume has less to do with instruction and
more to do with creative problem-solving.
The division of the book into six parts is to some extent arbitrary, since
the themes of the chapters are closely interrelated. It is obvious, for exam-
ple, that the discussion about the accurate and deep analysis of variations
is merely begun in Part 1 and continues right to the end of the book.
The analysis of variations is by no means an omnipotent force; in the
course of a game a player needs to guess as well as calculate. The problem
of developing your intuition has hardly attracted any serious discussion in
chess literature. I am not a professional psychologist and make no claim
to treating the subject scientifically, but I hope my practical ideas and rec-
ommendations will prove useful to the reader.
Many chess players make the serious mistake of devoting ~eir free
time solely to the study of opening theory. The fact is that errors commit-
ted at later stages have just as much bearing on the result as a poor start to
the game. Some weaknesses which can and must be overcome with inten-
sive will-power and persistence are shared by chess players of all levels.
To emphasize this point, the book includes critical analyses of games not
only by young·masters and candidate masters but also by high-ranking
 Preface 9

grandmasters - such as Artur Yusupov (who performs the task himself, in

Chapter 9, Missed Brilliancy Prizes) and Garry Kasparov.
Opening theory is subject to speedy revision, so that opening manuals
are sometimes out of date even before they are published. By contrast, as-
tute observations and conclusions about the game of chess as a whole re-
tain their value over a period of many years. You may confirm this by
acquainting yourself with the two articles written some decades ago by
Beniamin Blumenfeld, a master who was also a penetrating researcher
into practical chess psychology. Don't be put off by the mode of presenta-
tion, which seems slightly outdated by present-day standards. The writ-
er's thoughts are what matters, and they remain as relevant as ever.
I have already devoted one book - Secrets of Chess Tactics - to the
problems of attack and defence. However, these themes are inexhausti-
ble, and I hope that the fresh material discussed in the relevant chapters
will be useful to you.
In what has become in this series the traditional concluding chapter,
Yusupov an._alysessome games by our students. Whereas in our previous
books he placed the emphasis on instructive errors, this time the grand-
master concentrates on displaying the young players' creative achieve-
ments. The book ends with a brilliant game by Vadim Zviagintsev, which
the experts rated as the best in lnformator 62. It is extremely rare for
young players to receive this honour, given the jury's susceptibility to ti-
tles and big names. I wish our readers the same competitive and creative
success as our best students. I hope you will be aided in this by ideas that
you derive from the books in our series.
 1 The Technique of Analysis
and Decision-Taking

Mark Dvoretsky

What do we think about during a who perceptively investigated the

game of chess? We look for prom- psychology of chess thought.
ising possibilities, compare them The problem of thinking about
with each other, work out varia- moves has always interested me. I
tions, try to neutralize our oppo- have devoted a number of articles
nent' s counterplay, and so forth. to it, as well as several chapters in
All this,is a creative process my previous books. I have studied
which admits of no recipes suited some methods and devices (for ex-
to all cases. Yet there definitely do ample 'prophylactic thinking' -
exist rules, precepts of thought, see the book Positional Play) in
which in some way help us to or- considerable detail, others only in
ganize and regulate this process, to broad outline. I cannot formulate a
make it more reliable, to avoid the precise scheme for optimal think-
simplest mistakes and economize ing at the chessboard (I am con-
our thinking time - in short. to raise vinced that in principle no such
the level of our decision-taking. scheme exists), but I will give you
A fair amount has been written some advice which I hope will be
on this subject. Grandmaster Alex- of use to you in future contests.
ander Kotov. for example. has set
out his ideas on the technique of The thought-processes which
calculation in Think like a Grand- we are going to examine can be
master. I would also advise you to conveniently divided into two cate-
consult the interesting article by gories:
Mikhail Krasenkov included in the 1) Methods of searching for a
next chapter; and also the old but move and calculating variations;
by no means antiquated articles by 2) Means of economizing time
\
Beniamin Blumenfeld, a master and effort; rational thinking.
12 The Technique of Analysis and Decision-Taking

Technique of searching
for moves and calculating
variations

1. Candidate moves

Kotov was perhaps the first to high-

light this device of calculation. He
recommended that you should im-
mediately identify all the possi-
ble candidate moves, not just for B
your first move but for subse-
quent ones - and not just for your- Alexander - Euwe
self but for your opponent. If you Nottingham 1936
read the article by Krasenkov that I
mentioned, you will see that this side. Also 33 ...'ikd5 34 'iVxd5 %hd5
principle (like others that we shall 35 lted7 leads to roughly the same
discuss) by no means always thing. The attacking try 33 ... ltd5
'works'. Nonetheless it is very (reckoning on 34 llxg7 J:t.xg7 35
good advice for many situations. ltxg7 'ifbl + 36 q.>g2 lth5) is re-
Why is it so important to apply futed by 34 l:tc8!. Finally, if Black
the principle of 'candidate moves'? plays 33 ... l:tb8, White should not
In the first place, it helps us to sur- reply 34 1i'e4? l:txb4 (threatening
vey the variations rationally, to 35 ... 1!i'bl+) or 34 'ila7? J:r.xb4 35
single out all the continuations .:Xg7 llbl+ 36 ~g2 'itd5+, but 34
that have to be calculated. 1!i'c6! is a perfectly playable move
(34 ...l:r.xb4 35 .:.Xg7!; 34 ... 'ifbl+
Black is two pawns up, but his 35 ~g2 'iixb4 36 'ifxf6!; 34 ... d3
opponent's pieces are very active. 35 'ati>g2 !? d2 36 l:tcd7).
The g7-pawn is under attack. If I wouldn't say it was e~sential to
Black defends it with 33 ... 'iVbl+ work out all these variations accu-
34 'it>g2 'ikg6, then after 35 :Cd7 rately. You only need to satisfy
d3 36 'Wd5 White wins back the d- yourself that the opponent retains
pawn and will probably soon liqui- possibilities of defence. The point
date to a drawn ending with three is that Black has just one more re-
pawns against two on the same source available: he can simply
 The Technique of Analysis and Decision-Taking 13

push his passed d-pawn, allowing Let us begin with the queen's
l:txg7 and defending h7 from bl move tof7:
with his queen. This is the most a) 36 ~h2 d2 37 'fkf7 °ii'f5!.
forcing and hence the most tempt- Now 38 nxh7+ is impossible as
ing line, and naturally the one to the rook is taken with check, while
examine first. It is important to de- on 38 l:tg4 Black has the decisive
cide whether the opponent has per- 38 ... °ifxf2+ 39 'ith3 "iffl + 40 @h2
petual check. If not, Black should 'ii'hl+! (or 40 ... 'ife2+ 41 @h3
definitely play this line as he will 'ifxg4+!).
acquire an overwhelming material b) 36 'itg2 d2 37 'fkf7. This
plus. If there is no escaping the time 37 ... 'iff5? is bad due to 38
perpetual, he can go back to the .:txh7+! 'it'xh7 39 1Wxf6+ °ikg7 40
other continuations and study them 'fkxd8+ and 41 'Wxd2, when White
more thoroughly. comes out a pawn up, while the
33 ... d3! line 37 ... dl'fi'? 38 ltg8+! ltxg8 39
The precise order of moves is 'Wxf6+ leads to perpetual check.
important. On 33 ...l!i'bl+? 34 <iti>g2 However, Black can win by insert-
d3, White has the additional possi- ing 37 .. Ji'e4+!. Then 38 'ith2 'ii°f5
bility of 35 l:tcd7 ! d2 36 l:txd8 transposes to variation 'a•. If 3 8 f3,
.:txd8 37 'ifd7 !, leading to a draw. the simplest answer is 38 ...'ii°xf3+ !
After the text-move, however, 34 39 ..t.ixf3 dl'fi+ leading to a quick
ltcd7 is useless, as after 34 ...d2 the mate, though another possibility is
pawn will queen with check. 38 .. .'~i'e2+ 39 <if->h3'iffl + 40 'ifiih2
34 lhg7 lhg7 °ii'h 1+ ! 41 ~xh 1 d 1'fi+ 42 'iPh2
35 lhg7 Whl+ l:t.d2+ 43 <iti>h3 'iih 1+ 44 @g4 h5+ !
Black's next move will be 36 ...d2. 45 'iti>f4ltd4+ 46 'ifiie3'iigl+.
This is where we have to concen- We will now look at the rook's
trate on picking out the candidate move to g4.
moves. White has two ways of con- c) 36 'itg2 d2 37 l:tg4 'ii°hl+!
tinuing the attack: 37 ltg4 (threat- 38 'iti>xhldl'ii'+ and 39 ... 'i!Vxg4.
ening mate on g7) or 37 °iif7 (with d) 36 <if.>h2 d2 37 l:tg4 (D).
the idea of 38 l:t.g8+or 38 :.Xh7+ ). The hl-square is controlled by
In each case the white king may be the white queen. Black gets no-
on either g2 or h2. So there are four where with 37 ... 'iigl+? 38 'it>h3!
possibilities, and it was essential to 'iffl + 39 <iti>h4. The only possibil-
calculate all of them before play- ity is 37 ...'iig6! 38 .:lxg6 hxg6. Let
ing j3 ... d3. us see if White can give perpetual
14 The Technique of Analysis and Decision-Taking

even more important. It enables us

to avoid a typical mistake that nearly
all chess players make time and
again - plunging straight into ana-
lysing the first continuations that
come into their head. That way
some powerful possibilities may
be neglected, resulting in a great
waste of tim~ and energy. If we
have concentrated on seeking out
all the sensible candidate moves,
we sometimes discover resources
the existence of which we never
check with his lone queen: 39 'ilf7 suspected at the outset.
dl 'ii 40 'ilxf6+ ~h7 41 'ii'e7+ (af-
ter 41 'iff7+ ~h6 421i'f4+ ~g7 43
'ife5+, Black has either 43 ... ~h7
44 'ife7+ ~h6- see the main line-
or 43 .. !~f7 44 'iff4+ ~g8 45 'ii'c4+
'Wd5) 41...~h6 42 'ii'h4+ (after42
'ife3+ ~h5 ! 43 'ile5+ g5 the checks
run out) 42 ... 'ilhS, and the rook
cannot be taken because the queen
is pinned. We can now see why, af-
ter 36 ~g2 d2 37 l:tg4, the reply
37 ...'if g6? would be insufficient to
win; the sole correct move is w
37 ...1i'hl+!.
It remains to add that in the Najdorf - Kotov
game, after 36 ~h2 d2, White re- Mar del Plata 1957
signed (0-1).
The first thing that strikes you is
Thus, a preliminary review of the that White can win the h-pawn with
candidate moves ensures precision check: 21 .i.xf6 .i.xf6 22 1Wxh7+
and reliability in our calculation of <oli>f8.
There is nothing to calculate
variations. However, the 'explor- here - the resulting position just
atory function' of this method is needs to be assessed. White has no
 The Technique of Analysis and Decision-Taking 15

chance of giving mate (the king- .i.xf6 and especially 21 ~g4 were
side is solidly defended by the f6- too tempting) let us recall the prin-
bishop ). Black keeps some posi- ciple of 'candidate moves', and
tional compensation for his pawn, look for some other possibilities
in the shape of control of the c-file for White.
and the weakness of the white It turns out that there are two
pawn on d4. other methods of conducting the
Another, more tempting, con- attack:
tinuation is 21 ~g4. Obviously a) 21 ~dl (with the idea of 22
21 ... h6? 22 ~xh6+ leads to mate. ~hS).
It is also easy to see the variation b) 21 ~c2 (with the threat 22
21. .. ~xb3? 22 ~xf6+ ~xf6 23 ~xh7+ and justified by the varia-
'ii'xh7+ ~f8 24 'ifh8+! ~e7 25 tion 21...l:.xc2 22 .i.xf6 ~xf6 23
'ifxg7 and wins. However, we must 'iixh7+ and 24 'fi'xc2).
not jump to conclusions. First we Method "b' is more forcing, so
need to check that we have taken we should consider this one first.
all the def~nsive resources into ac- 21 itc2!? lhc2
count (we have to look for the op- Not 2 l...g6 22 .i.xf6, or 2 l...h6
ponent's candidate moves as well 22 ~xh6. On 21...ffl, White wins
as our own). We find the sole de- with 22 ~xh7 fuh7 23 'ifh5 !.
fence in 21...~fS!. There are vari- 22 ~xf6 h6
ous ways to win the h-pawn, but 23 'ifbS! ~xf6 (D)
none of them are entirely clear. For Or 23 ...l:.fS 24 ~xg7.
instance, after 22 ~xf6 ~xf6 23
~xf6 'ifxf6 24 'ii'xf6 gxf6 25
~xdS exdS 26 l:.xh7 ~g8, the out-
come is obscured by the activity of
Black's rooks on the open c- and
e-files.
Of course, no one guaranteed
that we could do more than win a
pawn here. The knight's move to g4
looks very strong, especially if we
notice that after 21...<~fS White can
continue the attack with 22 .i.h6!?.
Wait, though. A little belatedly w
(which is quite excusable - 21
16 The Technique of Analysis and Decision-Taking

24 'iixf7 + 'itb7 White's threat by giving advance

Or 24 ... 'ith8 25 l:xh6+ ! gxh6 26 protection to the f7-point: 21...:C7 !
tl)g6#. (22 ~h5 tl)xh5 23 'ii'xh5? ~xg5).
25 l:.xb6+! '.txb6 Instead Kotov carelessly played
26 'iig6# 21 .. .'~aS?, and after 22 ~hS! the
This winning combination (as white attack was irresistible. The
pointed out by Igor Zaitsev) imme- concluding moves were: 22. ..l:Ced8
diately settles the question of the (22 ... ltJxhS 23 'ii'xh5; 22 ... llf8 23
strongest way to continue the at- ~xf6 ~xf6 24 ~xf7+ l:txf7 25
tack. There is no need to analyse 'iixh7+) 23 ~xf7+ 'itf8 24 ~h6!
any of the other lines. As you see, tl)e8 25 'iff4 ~f6 26 ~xg7+ @e7
it is not only important to draw 27 ~xe8 il..xg7 28 ltxb7 l-O.
up a complete list of candidate
moves; you must also decide on
the best order in which to exam-
ine them.
The combination would be much
easier to find if White clearly had
no other promising possibilities.
However, in this case he did have
some, and they began by distract-
ing our attention. In such circum-
stances you could easily miss 21
il..c2 ! ! even if you were an excel-
lent tactician. The search tech- B
nique that we have been practising
('candidate moves') significantly Dvoretsky - Butnorius
increases our chances of success. Dubna 1970
However, no technique is any
help if a player lacks keen com- I had mishandled the opening,
binative vision. This quality must and at this point Black c9uld have
be developed and trained by regu- obtained an excellent position with
larly solving appropriate exer- the simple continuation 16... dxc4!
cises. 17 bxc4 tl)c6, forcing 18 tl)4b3 or
In the actual game, Miguel Naj- 18 lt:)4f3 (not 18 tl)xc6? bxc6 19
dorf played the weaker 21 il..dl ?!. l:te2 W'd3 threatening 20 ... W'xh3
His opponent could have parried and 20 ...ltd8). Without a doubt,
 The Technique of Analysis and Decision-Taking 17

the game would not yet be won for the normal development of Black' s
Black -it would 'merely' be excel- queen's knight.
lent. Alternatively he can try for 18 ... ii..b4
more by playing his knight to e4, I 8 ...ii..a5 !? .
but in that case he is falling be- 19 4Jf3! J:le8
hind in development, so the conse- 20 'ir'd4 ii..f8
quences need to be calculated in What does Black want to do
detail. Luckily for me, my oppo- now? To bring his knight out to d7,
nent made his move without prop- of course. Can this be stopped?
erly examining the variations. 21 4Je5! 'ir'b6?!
16 ... tt'ie4? Black would lose at once with
Clearly White must sacrifice the 21...tt'id7? 22 tt'ixd7 ii'xd7 23
exchange (17 tt'i4f3? tt'ixd2 18 4Jf6+ !. It was worth considering
tt'ixd2 dxc4 is thoroughly bad for 21 ... f5, but White would then have
him), but which way should he do maintained the advantage with 22
it? Once we locate all the candi- lbg3 ~f6 23 f4.
date possil:,tlities, it is fairly easy to 22 'ii'd3
decide which is strongest: Black has no moves. If 22 ... tt'ia6,
a) 17 :.Xe4 dxe4 18 tt'ixe4; then 23 tt'id7 followed by 24 tt'if6+
b) 17 tt'ixe4 .i.xe 1 18 'iixe 1 is decisive.
dxe4 19 'it'xe4; I must explain that despite the
c) I 7 tt'ixe4 Jlxe 1 18 cxd5. obvious excellence of my position,
In the first two cases, all we can I wasn't at all in an optimistic mood.
speak of is some compensation for In the Dubna Young Masters'
the lost exchange; whereas in the Tournament I had been playing
third case a double-edged position atrociously, constantly committing
arises, and it is not clear whose oversights which were chiefly due
game is preferable. to a lack of patience, a wholly un-
So you see that sometimes we justified haste in the taking of deci-
are not looking for candidate sions. In the previous game I had
moves but for 'candidate possi- thrown away what was virtually a
bilities' - short variations which won position with a single hurried
may begin identically. move. In the present game, my
17 tt'ixe4! hel opening play had been abysmal. It
18 cxd5! was sheer luck that my opponent
White's pieces dominate the cen- had handed me the initiative for the
tre, 'and his d5-pawn is preventing proverbial 'mess of pottage'.
18 The Technique of Analysis and Decision-Taking

"Sure", I thought, "I've got a de- for not just five minutes but twenty,
cent position, but I bet I'll lose my I worked out a winning combina-
nerve again and make some blun- tion.
der. Well, whatever he plays, I'm 23 d6!! l:txeS
going to take a full five minutes over On 23 ... fxe5 I intended 24
my next move! I've got loads of lbf6+ ! (24 lbgS ! is equally strong)
time in hand - I must use it." 24 ...gxf6 (24 ...~f7 25 ltixe8 ~xe8
After a long think, Butnorius 26 'ibg6+ ~d8 27 'iVg5+!) 25
played: 'ifxg6+ ~h8 26 'iVxe8 'ii'xd6 27
22 ... f6 (D) :Cl 'fJ/e728 'fJ/c8,and Black is de-
fenceless.
24 .i.xe5 fxeS
25 'iic4+ ~h7
26 'ti'c8 'fJ/d4(D)
26 ....i.xd6 27 lbg5+ and mates.

w
This actually disconcerted me.
"It's quite clear- I take on g6, he
develops his knight on d7, and then
there'll be something to think about, w
but right now, what am I supposed
to do with my five minutes? Well 27 ltigS+
never mind, I'm sticking ,to my White talces it into his head to
word all the same." give up another rook for luck. The
Just for something to do, I sacrifice leads to a forced mate. In
started going over some other pos- principle, however, this kind of
sibilities apart from 23 ltlxg6. "combination for combination's
(There we are - candidate moves!) sake" (the phrase is Grandmaster
The result was that after thinking Vladimir Simagin's) should on no
 The Technique of Analysis and Decision-Taking 19

account be commended. After all, Here is another aspect of the

on the elementary 27 l:tel (with same rule. Don't be in a hurry to
threats of 'ii'xf8 and 'Wxb7) Black calculate too far ahead. H prob-
could only resign. The best way is lems arise that demand analysis
always the simplest! in depth, don't rush to do it. First
27 Wh6 ask yourself how essential it is.
28 tf:jf7+ q.>h7 Are there any improvements for
29 'ii'xf8 'ii'xal+ yourself or your opponent at an
30 q.,g2 lDc6 earlier stage? New ideas at the
31 'ii'xa8 gS start of a variation are a good deal
32 lhxgS+ ~g6 more important than refinements
33 'it'e8+! 1-0 at the end of it; they are far more
relevant to the process of fighting.
2. What could I have missed? I once gave Artur Yusupov this
study to solve:
Sometimes during our analysis we
need to come back to the task of
looking for candidate moves. (We
may simply have forgotten to per-
form that task at the right time; but
then again, we can't always be sure
that our list of J>C?SSibilitieswas
compiete.) If the variations aren't
turning out in our favour, it makes
sense to go back to the beginning
and ask, What else could "there be
0

in the position? What idea could I

have failed to notice?" Conversely,
if the prospects look particularly
bright, we should turn the question A. Wotawa
round - what other resources might Deutsche Schachzeitung 1938
the opponent have? From time to Draw
time, try throwing off the burden
of the lines you have analysed, Artur thought for a long time,
and take a look at the position vainly trying to find salvation
with fresh eyes. This practice of- among the intricacies of the un-
ten gives excellent results. pleasant rook ending.
20 The Technique of Analysis and Decision-Taking

..Stop analysing!" I said to him

at last ... Just look at the position and
think what you might have missed."
All at once Artur hit on the solu-
tion.
1 e4!! fxe4
2 q;g7 mis
3 q;g6 l:teS
4 ..W6 %te8
S <M7
The king endlessly pursues the
rook.

3. Should we re-check our F. Bondarenko and

calculations? A. Kuznetsov
Commendation,
Another of Kotov's principles - to Gorgiev Mem 1977
go down each branch of the 'tree
of variations' once and only once - 2 a5, Black manages to bring his
seems to me dubious. After all, we bishop to f3: 2 ...it.b5 3 <iPb7it.e2 4
are not disinterestedly exploring a6 it.f3 5 it.xf3 gxf3 6 a7 h I 'if, and
possibilities - we are looking for Black has the advantage. The
the strongest moves. They don't al- zwischenzug 1 l:E.d5+(with the idea
ways come into our head immedi- of 1...q;g6? 2 l:td6+ and 3 .i.d5) is
ately, and sometillles there is no met by I.. .f5 ! 2 l:txf5+ q;g6, for
reason why they should do so be- example: 3 l:th5 'it>xh54 Jl..d5it.xa4
fore our analysis of the position has followed by bringing the bishop- to
reached a certain point. Suppose the f3, or 3 it.d5 @xf5 4 a5 <it?e55 .i.b7
variations are not turning out right .i.f7 6 a6 .i.d5, and Black wins.
for us because some detail is lack- There remains 1 l:dl, but then
irtg. It then dawns on us that this I ...Jlxa4 ! 2 J:kI Ji.c6+ 3 ~b8 h I 'if
detail can be supplied by inserting leads to a drawn bishop endgame.
some intermediate move which This appears to be the best
looked pointless at first sight. White can do - but the appearance
is deceptive. Let's not be in such a
How is White to stop the enemy hurry to agree a draw. Let us try to
pawn? In the event of l it.d5 .i.xd7 invent something. This of course is
 The Technique of Analysis and Decision-Taking 21

where some imagination is called 7 l:tg6+ '.ii>h5

for, but the accurate analysis we 8 l:tg8+ Wh6
have already performed will also 9 l:th8+ and wins
be of use. Calculating variations right t.o
We can find the solution if we the end, one after the other, is,
think of the move 3 .:Xc6 (instead then, something we rarely have to
of 3 '.t>b8) in the last-mentioned do. After putting together the list
variation. Unfortunately it doesn't of candidate moves, perform a
work, but the idea can be im- quick assessment, a preliminary
proved. examination of them. Your pro-
1 l:tdS+!! rs visional conclusions are sure to
We already know that this is the come in useful as you pursue your
only move. calculations further. You may be
2 l:tdl! ha4 able to gauge how promising some
Here again, our foregoing analy- particular variation is, and es-
sis established that Black has no tablish a rational order for your
choice. If tbe a-pawn stays on the further analyses. Perhaps (as in
board, White easily wins the bish- Najdorf-Kotov, for instance) one
op ending. move will prove so strong that it
3 ltcl Jlc6+ is simply not worth analysing the
4 l1xc6! ht'if others.
5 ~f7+ 'iti;>gS--(D)
4. Keep a mental note of the
results of your calculations;
terminate each variation
with a definite conclusion

Sometimes a completely precise

verdict is required, as when ana-
lysing 33 ... d3! in the Alex:ander-
Euwe game. In that case an exact
result - win or draw - had to be
reached. If we had broken off the
calculations only half-way through,
with the verdict 'a bit unclear'. we
could not have taken the right deci-
6 f4+! gxf3 sion.
22 The Technique of Analysis and Decision-Taking

However, a precise evaluation is

by no means always essential. For
example, you may come to the
conclusion that a certain position
arises by force but is difficult to as-
sess and would require additional
ca!culations. If necessary you will
carry them out later, starting from
the position in question and not re-
peating the analysis that led up to
it. That is the point of retaining
your conclusions from the varia- w
tions already studied.
Lerner - Lukin
5. Prophylactic thinking USSR 1977

It often helps to begin thinking 21 a4!

about the position by asking, One threat is 22 .i.fl. If 21...bS,
''What does my opponent want? then 22 .i.n ! .i.c8 (22 ...:cs 23
What would he play if it were his ~d4 tbxe5 24 ti'f5 f6 25 axb5
move?" ~b7 26 lha7) 23 'iff4 a6 24 bc4
bxc4 25 'ifxc4 leaves White with
Readers familiar with my earlier an extra pawn.
books will surely need no convinc- The question arises whether
ing that a capacity for prophylactic White can reach the same position
thought is of immense value. All the with 21 .i.fl (and if 21...b5, then
same, let me give one more exam- 22 a4). Which move-order is more
ple. accurate? Here we have to concen-
trate on looking for resources for
What does Black want? Advanc- the opponent. We are bound to give
ing the f-pawn would only weaken preference to the pawn moye once
his position. A much better idea is we notice that 21 .i.fl can be met
to bring his bishop into play via c8 by the unexpected 21...1i'a3!-even
or b7. To do this, he first has to pro- though, as Utut Adianto pointed
tect his knight with ...b5. out, White still retains the better
This tells us how White should chances with 22 .i.xc4 .i.xc4 23
handle the position. lbg5 ! b5 24 ~-
 The Technique of Analysis and Decision-Taking 23

In the game, Konstantin Lerner 16 ... l:tab8

didn't think about prophylaxis. Let us apply some 'prophylactic
He simply played 21 l:adl ?!. His thinking' and ask ourselves what
opponent replied 21 ...b5 !, not wor- Black wants. Obviously, to play
rying about 22 l:td7 .i.c8 ! 23 l:lx.e7 ... b5, open the b-file and pressurize
(or 23 e6 .i.xd7 24 exd7 'i&'xel+ the b2-pawn. How can this be op-
25 ~xel l:xel + 26 ii.fl l:d8) posed?
23 ...ii.xg4 24 :.Xa7 .ilxf3 25 ii.xf3 17 :Cl!
~e5 with approximate equality. Now l 7 ... b5 will be answered
by 18 axb5 lhb5 19 l:tc2. Then the
6. What is the drawback to bishop will go to cl, solidly de-
my opponent's move? fending the pawn, while the rook is
transferred to e2 on the open file.
If he makes an unexpected and A fine concept!
disconcerting move, ask yourself Let us now put ourselves in
this question. Logical considera- Black's place and try to detect the
tions sometipies help you to find minus side of White's plan. With
the weak spot in your opponent's the white rook on c2, Black has
idea and the best way to counter ... d4 with the threat of ... .i.b3.
it. However, White replies c4, when
the blocking of the queenside
should be to his liking.
Another point is that after l:c2
the a4-pawn is vulnerable. How
can Black exploit this? If he is giv-
ing up the idea of ...b5, there is
nothing for the rook to do on b8.
So 17 ...l:be8 seems indicated, and
if 18 :C2 then 18... ~d8!. On the
other hand if White tries 18 .i.e3
b6 19 l:c2?, our previous idea will
work: 19 ... d4! (with tempo) and
B then ....i.b3.
17 ... ~c8?!
Ciocaltea - Liberzon This is another way to attack a4
Netanya 1983 (18 l:c2 ~b6), but a less effective
one. What is its disadvantage? The
24 The Technique of Analysis and Decision-Taking

knight has left the kingside, where

it was needed for defence.
18 lbgS!
Threatening 19 'iih5 and 20
lbg6#.
18 ... g6?!
18 ... .l:.f6 is better. Now the al-
h8 diagonal is weakened. How can
White utilize this weakness?
19 c4! d4
19 ... dxc4? is unplayable due to
20 .i.xc6. By forcing his opponent
to close the queenside, White has
freed his hands for active opera- Kotkov - Dvoretsky
tions on the other wing. His advan- Moscow Championship 1972
tage is now beyond doubt.
20 'iVe2 lbb6 21 b3 :C.be8 22 once with 54 ... .Ua3+. That move
'iVf.2lbc8 23 Af3 lbe1+ 24 :.Xel must be adequate to win, and yet to
:C.e825 lb:e8 'ir'xe826 g4! lbd6 27 me it seemed technically impre-
gxfS lbxrS 28 ~e4 tbe3 (28 ...lbce7 cise. Black's king is stuck on the
is more tenacious) 29 Ji.xc6! bxc6 edge of the board and at the mo-
30 lbe4 lbg4 31 'iVg2lbh6 32 jfgS ment is taking no part in the game.
'iVf8 33 'ifxcS 'ifxcS 34 lbxcS 1-0 Here is a sample variation (though
of course it is not forced), in which
7. What am I trying to achieve? this defect makes itself felt: 55 @d2
tbxd4? 56 4,)xd4 ltd3+ 57 <tic2
This too is a useful question. Clar- .:Xd4 58 :tel !, and White actually
ify your aims. Do you want to ex- wins.
change a pair of pieces, seize an Black would first like to bring
important square, prevent some his king a bit nearer, so as to sup-
active undertaking by your oppo- port the c-pawn or stop tl).e oppo-
nent, or what? A logical examina- nent's passed e-pawn if the need
tion of the position may suggest a arises. I therefore started consider-
direction for further analysis. ing a rook exchange.
54 ... llbS!?
Black has a healthy extra pawn. Now if 55 l:.al+, then after
It is tempting to go into action at 55 ...c;t>b?the king has moved closer
 The Technique of Analysis and Decision-Taking 25

to the centre, aid Black will still Quickly reaching this position
get his check on the third rank. The m my calculations, I had a momen-
only question i: what happens if rary fright - I didn't see how to
White exchanges on b5 and picks stop the pawns. However, I then
up the d-pawn w.th his knight. Let asked myself, "Where should my
us try to find theanswer, but it will knight be going?" To d6 of course,
have to be conrincing and rela- and if possible with tempo. So its
tively uncompltated. The initial route became clear.
position is too good to warrant the 60
slightest risk. 61 e7
55 lhb5 <it'xb5 62 'at>e2
56 lcic3+ <it'b4 This whole episode is instruc-
57 lcixdSt- @b3 tive from the viewpoint of the tech-
58 e6 (D, nical exploitation of an advantage.
If 58 ~d2 lcixl4, the outcome is Black is clearly more comfortable
obvious. playing with his king on b7. Once
it turns out that White cannot af-
ford to exchange rooks, Black can
make life easier for himself by
playing 54 ...l1b5. For that reason it
is worth making the effort and cal-
culating the forced variation, espe-
cially if you are able to do it quickly
and accurately (in my case it took
about three minutes). I think we are
now better placed to understand that
good technique is in many re-
spects founded on short, precise
tactical calculations.
It remains for me to show how
After the text-nove, Black may the game concluded: 54 ... l:tb5! 55
play 58 .. .-!be7!?,but first it is bet- l:.al+ 'at>b756 'at>f:2l:.b2 57 'it>e3
ter to look at so~thing more forc- :.b2 59 'at>e3.:t.b3+
~b3+ 58 ..ti>f:2
ing - winning the white knight. (sealed). Another point about tech-
58 c3 nical exploitation: in won posi-
59 lcixc3 <it'xc3 tions it doesn't pay to force events
60 d5 before the adjournment - though
26 The Technique of Analysis and Decision-Taking

of course new regulations to elimi- 1. What should you think

nate adjournments make this ad- about first7
vice redundant.
60 ~f2 h4 61 e6 ~c7 (this is In Krasenkov's article you will find
where Black's 54th move brings some interesting ideas on this topic,
results!) 62 l:.a8 h3 63 ltg8 t'i:Je7 but they are partly at variance with
64 lth8 ~d6 65 li:Jgl c3 66 <it>e3 my own recommendations. It is up
c2+ (another way is 66 ...l:.bl 67 to you to choose which ones you
too2 h2 68 lhh2 c2 69 <it>d2 c I 'if+ think are nearer the truth.
70 loxcl l:tb2+) 67 ~dl l:tg3! (but If there is an attractive continua-
not 67 ...Abl? 68 t'i:Je2h2 69 locl !) tion which forces the play, then of
0-1 course it makes sense to begin your
calculations with this line (as we
Let us now proceed to examine saw, for example, when analysing
the other aspect of decision-taking Alexander-Euwe). It Is usually sen-
technique. This aspect is even less sible to start by examining the
well studied. forcing moves such as exchanges,
moves that win material, or the
Principles of rational, opposite - sacrifices. Calculating
economical thinking a concrete line is often simpler
than appraising the consequences
First, the most general statement of of a quiet continuation.
aims. When thinking about your Once I was talking to Mikhail
move, your task Is not at all to cal- Katz, the famous draughts trainer
culate every variation to the end who has coached an entire galaxy
and attain an exhaustive under• of Women's World Champions
standing of the position. You have (Elena Altshul, Zoya Zadovskaya
one task, and one only: to take and others). He told me he taught
the correct decision, to play the his pupils to start by considering
best move.. As far as possible, try all possible moves that put men en
to minimize your expenditure of prise. As a result it was extremely
time and effort. You should only rare for them to miss unexpected
calculate the minimum number combinations.
of variations necessary for mak- If you discover a combinative
ing the right decision. idea and feel it is probably sound,
How is this to be achieved? it makes sense to examine the op-
Here are some suggestions. ponent's weakest-looking amwers
 The Technique of Analysis and Decision-Taldng 27

first. (Again, let us recall Alexan- force perpetual check or (for in-
der-Euwe. We did things in that or- stance) a level endgame. You are
der when we studied the position now justified in breaking off and
after Black's 35th move.) Quickly saying "Right, then - I'll play the
looking over the easiest variations, combination! I've got perpetual
you narrow down the field of in- check in hand. There should be
vestigation, possibly restricting it some more attacking chances, but I
to one or two continuations. After won't finish calculating them - I
that it is easier, from the psycho- can do it later, or force the draw if
logical viewpoint, to concentrate there's nothing else for it."
all your efforts on the crucial lines. The 'emergency exit', which al-
Conversely, if you suspect the lows you to do without calculating
combination will not work, be- the critical variation in advance,
gin by concentrating on the de- doesn't have to be a forced draw. It
fence that looks best. If it refutes is enough to note that somewhere
the combination, that is enough; along the line you have a move
there is no point in examining any which may actually be second-rate
other resources the opponent may but which you nonetheless judge
have. to be perfectly acceptable. A few
Obviously these recommenda- moves later you will find it much
tions are far from having absolute easier to decide whether to go into
validity. In chess the most varied the main variation or be content
situations arise, and you may have with a reasonable deal.
to cope with them in completely
different ways, but the guidelines I 3. Process of elimination
have given will apply to the major-
ity of cases. Sometimes there is no point in
precisely calculating the line you
2. The •emergency exit' intend to play. You just need to
conclude that the line makes
I shall not give any examples of some sense, that there is no im-
this, but simply explain the idea. mediate refutation, and that you
Suppose an obscure position have nothing better anyway - all
has arisen,and it is not clear who is other moves are bad. By such
better. You start to work out a com- means you may be able to save a
plex combination and notice that at great deal of time and energy.
some- point, if you want, you can Some chess players are prone to
28 The Technique of Analysis and Decision-Taking

forget this method of selecting a That is exactly what happened,

move, or apply it ineptly. for example, in the very important
In the years 1973-5 I worked game Yusupov-Zapata played in
with Valery Chekhov, preparing the 1977 World Junior Champion-
him for the World Junior Champi- ship at Innsbruck. The Colombian
onship. We gave a lot of attention Alonso Zapata played magnifi-
to perfecting his technique for cently in the second half of the tour-
decision-taking and especially for nament (scoring 6 out of 7) and
making economical use of his took second prize. His only loss to-
thinking time (Chekhov was then a wards the finish was against the
time-trouble addict). An instruc- player who went on to win the
tive episode occurred in one of the Championship - but see how easy
training games we played. this win was for Yusupov: 4 ...d6 5
d4 0-0 6 it.g2 lbbd7 (if Black is in-
Dvoretsky - Chekhov tending ...e5, it is better to play it at
Moscow 1974 once - 6 ... e5 - and meet 7 dxe5 by
Reti Opening 7 ... lbfd7) 7 0-0 eS?! (7 ...l:r.e8!?;
7 ...c6 !?) 8 dxeS ltlg4 9 ltlc3 dxe5
1 lbf3 lbf6 10 lbd2 .:te8 (10 ... fS would be an-
2 g3 g6 swered by 11 e4, but perhaps that
3 b3 it.g7 was the lesser evil) 11 l2Jc4 lbb612
4 it.b2 lbxb6! cxb6 (12 ... axb6 13 'iVxd8
Valery had a reason for asking .:.Xd8 14 .:tfd 1, and 14....:tesfails to
me to play this particular system 15 lbd5) 131i'xd8 :Xd814 lbdl
(of which I didn't know the theory JJ.f 5 (good ideas for Black are al-
at that time). This was how Alex- ready hard to come by) 15 it.xb7
ander Kochiev, his chief rival in l:.ab8 16 e4 Ji.d7 17 .i.dS, and
the coming qualifying tournament, White went on to exploit his extra
liked to play with White. It must pawn in comfort.
be said that this kind of opening 4 ... 0-0
scheme, for all its unassuming ap- Occasionally even such a natu-
pearance, contains a fair amount of ral move as castling can prove a
poison. If Black has no precise loss of time. The most precise
plan, he can soon land in a strategi- move-order is 4 ... d6! (intending
cally difficult situation - he only 5 ... e5) 5 d4 c5 !. The point is that
needs to commit two or three White has to do something about
barely perceptible inaccuracies. 6 ... cxd4 7 lbxd4 d5, but 6 c4 can
 The Technique of Analysis and Decision-Taking 29

be answered either by 6 ... cxd4 7 and White will aim to gain a space
tl)xd4 d5 8 .i.g2 dxc4, or by advantage with d4-d5.
6 ... tl)e4 with the unpleasant threat 10 tl)b4!? (D)
of 7 ...~a5+ (if the moves 4 ...0-0 5 The e4-knight is attacked. How
.i.g2, had been inserted, White would you continue for Black?
could simply castle here). Nor does What position would you aim for?
White achieve anything with 6
dxc5 'ii'a5+ 7 lbbd2 'ii'xc5 (threat-
ening 8 ...tl)g4; 8 a3 would now be
a mistake due to 8 ...0-0 9 Ji.g2
tl)g4 ! 10 tl)e4 tl)xf2 ! !) 8 Ji.d4 ~h5
(8 ...'ikc7 is also good) 9 il.g2 tl)c6
10 .i.b2 Ji.h3.
Another way of carrying out the
same idea is 4 ... c5 ! (threatening
5 ... d5) 5 c4 d6! (with a view to
6 ... e5) 6 d4~e4!.
5 .i.g2 cS
6 c4 tl)c6
7 0-0 d6
In such positions ... e5 is an un- Sharp attempts like 10 ... tl)gS?
pleasant positional threat, since 11 f4 tl)h3+ 12 <itihl and 10 ... d5?
the b2-bishop would be shut out of 11 cxd5 'i!fxd5 12 g4 clearly fail.
the game. White could only acti- 10 ... tl)f6? is also bad; the space-
vate it with e2-e3 and d2-d4, which gaining d4-d5 (either at once or af-
would be difficult to carry out. ter exchanging on f5) guarantees
8 d4 tl)e4 White a positional plus. That leaves
9 tl)bd2 only 10...tl)xd2. \
The game Korchnoi-Gligoric, When playing 10~h4, I was in-
USSR-Yugoslavia match 1956 con- tending to investigate whether I
tinued 9 e3 .i.g4 10 ~cl .i.xf3 11 could spoil Black's pawn forma-
.i.xf3 tl)g5 12 .i.xc6 bxc6 13 dxc5 tion by inserting an exchange on
dxc5 14 .i.xg7 ~xg7 15 f4 tl)e4 f5. The complications following
with equality. l 1 lbxf5 tl)xf112 lbxg7 might tum
9 ... ii.CS out in my favour, due to the power-
After 9 ... lbxd2 10 'i!fxd2 the pin ful b2-bishop. But of course I didn't
on the al-h8 diagonal disappears start working out the variations
30 The Technique of Analysis and Decision-Taking

beforehand; I merely satisfied my- Objectively the chances are about

self that the simple 11 1i'xd2 (the level here.
'emergency exit'!) was perfectly Taking on d4 with the pawn is
playable. weaker, since pawn weaknesses
In general terms 11 ltlxf5 looks are formed in the black camp. The
a suspect idea; therefore, as I have better pawn structure guarantees
said before, we should begin by White a small but lasting advan-
looking for the simplest refutation. tage.
There is no hurry to study the posi- What happened in the game?
tion where Black is the exchange After I played 10 ioh4, Chekhov
up (even if that position is in his fa- sank into thought. I spent a couple
vour). First ask yourself what the of minutes looking at the conse-
candidate moves are. I saw the re- quences of 10...t°oxd2 11 t°oxf5.
ply l 1...t°oxc4!, keeping a sound Then, having convinced myself (in
extra pawn for Black, and broke his thinking time!) that that line
off my calculations at that point. was unplayable, I got up and walked
So 11 1rxd2 is forced. Most about. There was nothing more to
likely Black should not allow the think about - it was up to my oppo-
advance d4-d5 (although 11....i.d7 nent to choose his move.
is possible) so he should choose Ten minutes passed, then an-
between l l...cxd4 12 ltlxfS gxf5 13 other ten. This rather began to irri-
hc6 bxc6 14 .i.xd4 and l 1...ltlxd4 tate me. What was he thinking of?
12 ltlxf5 ltlxi5 13 .i.xb7. Which of By process of elimination it is
these positions should Black pre- quite easy to see that Black must
fer? What is required now is not exchange knights. Why waste any
calculation but sure positional as- time on it, then?
sessment. 10 ... loxd2
I think the second possibility is 11 1rxd2
the right one. White has to conduct As you would expect, I made my
an attack on the kingside with h4- move instantaneously, whereupon
h5 or f4-f5 as appropriate, but the Chekhov once again immersed him-
success of such an attack is very self in thought for twenty-five min-
doubtful, because there are not all utes or so. Evidently he had still
that many pieces left on the board not come to a definite conclusion,
and meanwhile Black is ready to or else some new doubts had arisen,
counter-attack on the queenside at bringing him back to the problem
once by advancing his a-pawn. he had been struggling with.
 The Technique of Analysis and Decision-Taking 31

Of course, it was sheer waste to 17 J:lad1

spend nearly an hour on such a This threatens 18 l:.d5, winning
relatively uncomplicated problem. a pawn.
What's more, Chekhov's solution 17 ... 'ii'e6
was by no means the best. 18 .:re1
11 cxd4?! Black has to contend not only
12 ~S gxfS with l:.d5, but also with the line-
13 hc6 bxc6 opening move e2-e4.
14 hd4 .hd4?! 18 ... f6
14...e5 15 ~b2 'ile7 was prefer- If 18...'ilf6, White plays 19 ltd5
able. e5 20 'ir'd2 (from this variation we
1S 11i'xd4 cS? (D) can see why 18 e3? ! would have
After this Black has a difficult been less exact).
position. He had to choose be- 19 l:.dS aS
tween 15... 'ild7, intending ...e5, Black could put up stiffer resis-
and 15...e6, preparing ...d5. But tance with 19...'ile4 20 :Xf5 'ir'xf4
then, there is nothing surprising in 21 l:.xf4 a5. The advance of the a-
my opponent's weak play. If you pawn would promise him some
use up too much time and energy at counterplay in the endgame, but is
one particular stage, you won't not much use in the middlegame.
have enough left to solve your prob- White replies by organizing an at-
lems afterwards. tack on the king.
20 lhfS a4
21 e4 axb3
22 axb3 l:.fb8
23 l:.e3
White has a strategically won
position. He aims to open lines on
the kingside with g4-g5.

After the game I naturally asked

Chekhov what he had been think-
ing about when he took such a long
time over move 10.
w "Well, you know," he replied, "I
was deciding what to do after-
wards."
32 The Technique of Analysis and Decision-Taking

"But 10... ttlxd2 was forced, In Part 3 of Secrets of Chess

wasn't it?" Training I described an interesting
"Yes, it was." type of exercise - playing out end-
"Were you risking anything game studies over-the-board. In
when you played it? Could it have this way you can develop your
lost outright?" imagination and your analytical
"No." technique; in particular, you can
"Then why didn't you play it learn to make sensible use of the
and think afterwards, in the posi- process of elimination.
tion where you actually had to
make a choice? With that position
in front of you on the board, it
would be easier - there would be
less chance you'd miss something.
In any case, you might have been
able to do some of your thinking in
my time. H you'd taken on d2 at
once, I'd still have been looking at
11 fuf5."
Oddly enough, these arguments
were new to Chekhov at the time.
He was wholly unfamiliar with
such time-saving devices as the
process of elimination. But then, G. Kasparian
large numbers of players are prone 1st Prize, '30 Years of Soviet
to this same kind of inefficient Armenia'Toumey 1950
thinking.
As a result of analysing his typi- White's knight is trapped and
cal mistakes in the taking of deci- will hardl•i escape alive. The at-
sions, Chekhov almost eliminated tempt to extricate it with the aid of
the desperate time-scrambles that the bishop is easily refuted: 1 ~d5?
had characterized his earlier games. ttlf4 (not the immediate 1... e6? 2
In the following year (1975) he Jt.xe6 'iti>xe63 ttlg6) 2 Jt.c4 e6 3
successfully came through the ~g3 Jt.d6 and 4 ...~g7. 1 @xh3
elimination process and went on to looks tempting in view of the lines
gain the distinguished title of 1...Jt.g7 2 e5+ and l...cii;,g72 ~g4
World Junior Champion. itJf6+ 3 ~f5. However, Black has
 The Technique of Analysis and Decision-Taking 33

the much stronger 1...~f4+ 2 Wg4 While he is at it, we shall work out
e5 ! followed by ...Ji..g7. Playing it some variations too. After 4 ...@g8
this way, White has no chance. 5 c;t;g6, he can only strengthen his
What else can he do? The only position by 5 ...e6 6 c;t;f6Ji..b4. Now
continuation to offer any hope is 1 7 ii..c8? would be a mistake, as af-
e5+ <J;;g7,and now either 2 Ji..f3 ter 7 ... @f8 the e6-pawn cannot be
lbf 4 3 ~g3 or 2 <iitxh3<iti1xh8 3 'itg4 taken because of 8 ...ii..e7+; while on
!i)g7 4 cJi;g5, trying to exploit the 8 Ji..d7 ii..e7+ 9 <iitg6Ji.dB 10 ii..c8
temporary lack of co-ordination of Ji..h4 (zugzwang) 11 Ji..d7 Black
the black pieces. Of course, we places his king on e7 and consoli-
have to be quick to unearth the lit- dates his forces. We must attack
tle traps that make life easier for us the pawn from the other diagonal:
in some variations, for instance 2 7 ii..a6! <;Pf88 ii..c4 ii..e7+ 9 <;Pg6
<;Pxh3 ~f4+ 3 <;Pg4 and then Ji..h4 10 ii..a2, and Black is unable
3 ...~d3 4 lt)g6! or 3 ...lt)e6 4 lt)f7! to improve his position any further.
<J.xn5 Ji..d5 with a draw. 4 ...<:J.h7does not alter matters: 5
Thus our"first move is clear. ii..e4+ <la'g86 <;Pg6.
1 eS+! ~g7 Black's only remaining try is to
Now which way do we go? If 2 remove his knight from g7:
Ji..f3? !i)f4 3 'it.>g3,the black knight 4 ... tbe6+
finds an extra square, g2. After 5 <:J.g6 Ji..g7!(D)
3 ... lt)g2! 4 <;Pxh3 lt)e3 our defen- 5 ... tbd8 is useless: 6 ~d5 (in-
sive resources are exhausted. There- tending 7 e6) 6 ...e6 7 ~xe6! !i)xe6
fore we must capture the pawn at 8<;Pt7.
once.
2 @xh3!
An immediate draw would re-
sult from 2 ...tbf4+ 3 <;Pg4!i)e2 (we
already know how to meet 3 ...lt)d3
or 3 ... !i)e6) 4 <;Pg5<iitxh8 5 ~g6,
and Black cannot stop the threat-
ened 6 q;t7 (e.g. 5 ...<J;;gs6 Ji..d5+).
3 <J;;g4 ~g7
4 <iitgS
Now Black must have a long
think, as it is not at all simple to
convert his extra knight into a win.
34 The Technique of Analysis and Decision-Taking

Seeing that 6 ffi? ltld8+ is un-

playable, White must move his bish-
op, renewing the threat of @n.
Black will reply 6 ...t;)d8, and then
we play 7 e6 !. However, where
should the bishop go?
If 6 .i.c6?, Black replies 6, ..ll)d8
with tempo. Other tries that fail are
6 .i.a8? t;)c7, 6 .i.e4? t;)c5, 6 .i.f3?
t;)d4 and 6 .i.g2? t;)f4+. That
leaves only 6 .i.c8 or 6 .i.hl, but
after 6 .i.c8? li)d8 ! 7 e6 <itg8 the w
black knight comes into play via
c6 or b7. The bishop is compelled can see, there is no need to ana•
to remain on the long diagonal. lyse any long and complicated
6 .i.h 1! ! lZ)d.8 variations 'to the end'; it is much
7 e6! .i.b2 more important to examine the
8 .i.dS! essential short variations pre•
Not 8 .i.a8? t;)xe6 9 c;t;n li)c7. cisely, while trying to register all
8 •.• .i.a3 the vital resources both for our-
9 ~h6 .i.cl+ selves and for the opponent.
9 ...~g8 doesn't help: 10 <itg6
~8 11 ~h7. An amusing sight: 4. Comparison
the black knight and king are pris-
oners in their own camp! This is quite a sophisticated device.
10 ~g6 .i.d2 Sometimes a quick choice of move
11 .i.hl!! (D) can be made purely on the fol-
Again the only safe square for lowing lines: you realize that this
the bishop. move brings about a situation
11 ... .i.e3 which is in no respect worse, and
12 .i.dS! in some respect better, than the
The drawn result is now obvi- one you would obtain from a dif-
ous. ferent continuation. For example,
We have thus traced roughly it was just in this way that we ar-
what a chess player would be think- rived at the correct order of moves
ing if he had to play the position in (33 ... d3!, not 33 ...'ii'bl+?) in the
Kasparian"s endgame study. As you game Alexander-Euwe.
 The Technique of Analysis and Decision-Taking 35

Let us consider the following upon to elucidate how the game is

study. meant to end. Starting by compar-
ing the two possibilities, you
quickly place the knight on h4 and
leave it to your opponent to look
for a way to save himself. In his
thinking time you can form a better
picture of the ensuing variations.
1 ~gl
2 ll)f3+ r,t.,g2
3 ll)xb2 'itixh2
Now if-White goes after the a-
pawn, Black will lock him in by
running across to c7 with his king.
How can this be prevented?
4 eS! .i.xeS (D)
F..Bondarenko and M. Liburkin
2nd Prize, All-Union Physical
Culture and Sport Committee
1950

White has two moves, 1 lbh4

and 1 lbe 1. They come to the same
thing in the case of 1...~gl 2
lbf3+ ~g2 3 lcixh2. However, af-
ter 1 lbel .i.c3! White again has to
play 2 l£if3, whereupon 2 ... ~g2
brings about the same position
only with the black bishop on c3.
We don't even need to figure out
whether this has any significance. 5 ~e6!! ~g3
Why should we give the opponent 6 r.t.>d7 ~4
an additional defensive resource? 7 ~c8
What we should play is clear. Black's bishop gets in the way
1 loh4! of his own king.
When thinking about a move in Observe what happens in the case
practical play, you aren't called of 4 .. ..i.c3 (instead of 4 ....i.xe5) 5
36 The Technique of Analysis and Decision-Taking

e6 Jl.b4 6 q.,es q.igJ 7 '.PdS Wf4 8 you may find a precise solution
cj;;c6<it>eS9 <itib7q.,d6, and this to the problem that faces you,
time the king hampers the bishop: and that this solution will deci-
10 e7! and wins. sively affect the further course of
From this last variation it the game - in other words, at the
emerges that if White mistakenly key moments of the struggle (the
starts with 1 ll'lel ?, the tempo ability to identify them is very
Black gains by 1...1'.c3 ! is of vital important); alternatively if you
significance: 2 ll::if3 cj;;g2 3 ll'lxh2 can't see a continuation that is at
'it'xh2 4 c5 'it'g3 5 e6 .i.b4 6 'it'e5 all acceptable, and need to dis-
.i.f8 (or 6 ...cj;;g4) 7 <it>d5Wf4 8 cover one.
'it>c6 <itie5 9 <if.lb?Wd6 10 'iii>xa7
cJ;;c7and draws. With that, this chapter closes. I
don't recommend that you restrict
5. Don't spend too long yourself to the above advice; delve
analysing extremely complex into the matter more deeply,
variations - in such cases, evolve new principles of your own.
rely on your 'feel' Study examples which illustrate
the rational technique of searching
Quite often we encounter obscure for moves and taking decisions.
situations where it is practically Solve special training exercises;
impossible to arrive at the truth in analyse your own performance in
conditions of limited thinking time. tournament games. I hope that this
Even if you find the right move. manner of working at your chess
the price for using up an immense will gain your interest. But are we
amount of energy may prove too talking about chess alone? After
high. Most likely you will have too all, rational, precisely organized
little time and strength left for later thought is of value in any sphere of
decisions. life.
In what cases does it make
sense to spend a lot of time think- In conclusion, here are some
ing about a move? The answer quite difficult exercises for you to
is, when you understand that solve independently.
 The Technique of Analysis and Decision-Taking 37

Exercises

1. White to move 3. White to move

4. Black to move

Solutions are given on pages 269-272.

2 Wandering in the Jungle
Mikhail Krasenkov

The calculation of variations is one undergrowth' (masses of long, in-

of the fundamental elements deter- volved variations). In his opinion,
mining a chess player's choice of the chief rule for calculation is "to
move. The ability to calculate ac- go down each branch only once.
curately a long way ahead, in con- No going over things twice, no go-
ditions of limited time on the clock, ing back! Only in a few individual
is the paramount factor in your stan- cases, in especially complicated
dard of play, and depends to a large positions, will a grandmaster check
extent on organizing and disciplin- through his chosen variation a sec-
ing your thoughts. Perhaps the ond time. As a rule, he will not
foremost specialist in this field was keep climbing all over the tree."
Grandmaster Alexander Kotov, the I would observe that, essentially,
originator of a theory of chess the tree of variations is formed dur-
analysis which ought to be familiar ing the process of calculation itself,
to every accomplished player. (All so that we can speak of a particular
my quotations are from his book order in which it takes shape.
Think like a Grandmaster.) On the very important question
The fundamental concept in Ko- of assembling the branches of the
tov 's theory is the •tree of varia- tree - that is, deciding on the
tions•. "All possible continuations moves which call for examination.
in a given position can be visual- in any particular position - Kotov
ized as a 'tree' in which variations formulates the following rule:
and sub-variations are represented "When beginning our analysis we
as branches and twigs." Kotov dis- must first of all mentally enumer-
tinguishes between different types ate, and make an exact note of, all
of 'tree' - between a 'bare trunk' the possible candid.ate moves in
(a single variation or branch), a the given situation... Having de-
'coppice' (lots of short lines) and fined and collected them, we start
'impenetrable thickets' or 'jungle working out one variation after
 Wandering in the Jungle 39

another in turn." This must, of Here is a very simple example to

course, apply not only to the initial illustrate Zlotnik's first point.
position but to any position reached
in the analysis, where one side has
a choice of continuations. Thus, all
possible candidate moves are de-
termined once and for all, in ad-
vance of your analysis of the
position. They are selected on the
basis of general considerations, in-
tuition, and so forth.
Basically, these two rules of Ko-
tov 's are applicable to the majority
of positions in practical play, and if
a chess player is able to obey them w
as a matter of second nature, this
constitutes 'an immense step for- Alexander - Marshall
ward in his mental development. Cambridge 1928
Yet these rules also have signifi-
cant shortcomings which emerge From examining the line 1 l:tf4
particularly in complex, obscure po- exf 4 2 gxf 4, to which Black replies
sitions. International Master Boris 2 ...dxc3 - controlling the g I-square
Zlotnik, who has investigated this - you could hit on the idea of side-
question, writes as follows: stepping with the knight first, by 1
"1. In complex positions it is ex- ~d5 or 1 ~a4! (and then 2 l:tf4).
tremely difficult to compile the list The move 1 ~a4 is indeed the
of candidate moves at the very put- quickest way to win. But without
set of your calculations. In prac- any analysis (however rudimen-
tice, the candidates come to light tary) of the 1 l:tf4 line, it is hard to
during the process of fathoming see what reason there would be to
the position. include 1 ~a4 among your candi-
"2. A fine point in the analysis date moves.
of one variation often dawns on Zlotnik's second remark points
you in the course of calculating a to such characteristics of human
different line. Therefore in diffi- thought as the workings of the sub-
cult situations, repeating your cal- conscious and the faculty of asso-
culation is essential." ciation. Another major defect of
40 Wandering in the Jungle

Kotov's theory is also readily ap- position; and so on. The aim must
parent: he ignores the problem of be realistic, that is, it must be
the order in which to examine the based on your assessment of the
candidate moves, asserting that position and on intuitive consid-
this ..depends on the character and erations. Your sights may be raised
habits of the player, and on the pe- somewhat higher if enough think-
culiarities of the position". If, as ing time is available, or lowered if
Kotov assumes, it is obligatory to time is short.
examine all the candidate moves, 2. Lo<>k for ideas to achieve
then of course the order hardly the aim; select appropriate can-
matters. In fact, though, there are didate moves, and ( this is very
many cases where, in the interests important) decide on their order
of economy, the analysis of some of priority; that is, decide which
candidate moves can be dispensed ones are most or least likely to suc-
with; it has no bearing on your de- ceed.
cision, which can be reached sim- 3. Analyse the variations (as
ply by analysing the other moves. deeply as possible) in order of
In such cases the order of analysis priority (starting with those which
is of supreme importance. seem most likely to achieve the
In this chapter I shall try to pres- aim). Similarly, at each point in the
ent a more complex (though admit- analysis where a choice arises, the
tedly none too schematic) algo- order of calculation should depend
rithm for calculation. This algo- on the priority of the possible
rithm, which I shall illustrate by an moves (with your specific aim in
example, is in fact used (uncon- view).
sciously) by many chess players in 4. If you find a continuation
obscure positions (of the 'jungle' which achieves the aim, what fol-
type). lows depends on your available
1. Define the aim of your analy- thinking time. If time is short, the
sis -in other words the criterion by main part of your analysis should
which you will judge the varia- stop at this point (though there still
tions and decide whether they sat- remains the essential 'Blumenfeld
isfy you or not. The aim might be, re-check' - see point 8 below). If
for example, to attain a decisive there is plenty of time, the aim may
material plus; to increase your po:. be adjusted (that is, raised); the list
sitional advantage; to equalize the of candidate moves that remain to
game; to put up resistance in a bad be considered may be revised, and
 Wandering in the Jungle 41

the analysis may continue. :Should some forced lines of play will have
the new aim not be attained, you emerged. Often the new idea - the
will fall back on the line already new candidate - will not be found
found. in the initial position but at the end
5. If as a result of your analysis of a series of moves; together with
no way is found to achfove the them, it will constitute a 'candidate
aim, your further action depends variation'. You will now start cal-
once again on the clock sitlllation. culating the new possibilities (com-
With a time shortage you must pare point 3); this is stage two of
lower your sights, correct tthe list the analysis. (Sometimes, though
of candidate moves, and resume not often, you will go through this
the analysis. It often happens that cycle a third time.)
your new aim is already satisfied Generally speaking, a repeat
by a line you have examined - or a analysis is an admission of partial
suitable continuation is relatively failure. Ideally, all ideas for pursu-
simple to find. The essential thing ing your aim should be incorpo-
is not to lnake your move 'on rated in stage one of the analysis.
spec', without any calculation. However, as we have seen, this is
6. If on the other hand there is not always possible.
plenty of time left, and your intui- 7. It may be that while analysing
tion suggests that the aim ought to one variation you hit on a new idea,
be attainable (an accomp lished
1 a new candidate move, which does
player should put more faith in his not apply to this particular varia-
intuition, which after all embodies tion. In that case, decide where the
his accumulated understanding of new move comes in order of prior-
chess) - then you may (and must) ity, but don't start to examine it be-
deliberately perform a 'repeat fore finishing with the line you are
analysis' of certain lines. In so do- currently calculating. An excep-
ing you will be looking for new tion may be made when it is obvi-
ideas to achieve the aim. Accord- ous at first glance that the new idea
ingly, you will find new candidate is better (not just worthy of higher
moves and 'candidate variations'. priority) than the line you are look-
Let me explain what this means. In ing at.
many variations, when ana]ysing 8. One of the major defects in the
the first time round, you will al- mental equipment of many players
ready have discovered your oppo- is 'chess blindness', a proneness to
nent' s strongest, or only, replies; overlook elementary replies for the
 42 Wandering in the Jungle

opponent at a distance of one or

two moves. An antidote to this fail-
- ing is 'Blumenfeld 's rule' (of
which Kotov also speaks): on com-
pleting your calculations and male-
. ing your decision, pause for a
moment (write the move on your
scoresheet) and look at the posi-
tion with the fresh eyes of an out-
sider. Could the move you intend
be a blunder leading to immediate
disaster? Check that it is not, and B
only then make your move. If you
see that the move is an error, you Lagunov - Krasenkov
will have to renew your analysis. Dnepropetrovsk 1985
In this case, as a rule, you will need
to lower your 'aim' and seek sim- to combine attacks on the bishops
plifications, since a crude oversight with threats to penetrate on the back
signifies your unreadiness for a rank. I listed the following 'cand-
complex struggle. idate moves• in order of priority:
Blumenfeld's rule is well known
to the majority of accomplished a) 23 .. .'t!fdl.
players, but ... in the heat of battle b) 23 ...'ii'b4.
they often forget it. c) 23 ...'ii'a4.
I .would like to illustrate the d) 23 ...1!fg4.
foregoing by showing the conclu- e) 23 ...1!fe5.
sion of one of my own games. I
consider the winning manoeuvre I Stage one of the analysis then
found in this game to be one of my began.
best creative achievements. a) 23...'ii'dl 24 •~fl. and now:
al) 24 ...'.g4 25 .i.d-2 'ifc4+ 26
The awkward placing of White's 1!fe2,and Black has nothing.
pieces and the weakness of his first a2) 24 ...'IVhS 25 .i.d2, and
rank prompted me to look for a Black has two choices:
forced win, in other words a large a21) 25 .. .'ii'xh2 26 f3! 'ii'hl+
material gain (this constituted the 27 ~f2 't!t'h4+ 28 ~fl 11Vc4+29
'aim' ofmy analysis). The idea was 'ii'e2.
 Wandering in the Jungle 43

a22) 25 ... '6'b5+ 26 c;t_,gl'6'xb2 queen behind the rook; I would

27 .i.e3 and then 27 ...'i!ic3281Wfl, have to do this without allowing
or 27 ...'iib4 28 'ircl l:.c7 29 '6'dl. White to defend d 1 by .i.c2. The
a3) 24 ...Wa4 25 .i.d2 'ifb5+ 26 following 'candidate variations'
c;t_,g1,transposing to 'a22'. suggested themselves:
In the above variations, Black
merely gains a second pawn - the A) 23 .. .'ildl 24 c;t_,n 'ii'a4 25
aim is not achieved. .i.d2 l:.d6;
b) 23 ...'ii'b4 24 'ii'cl 'irc5 (or B) the same, but with 25 ...l:.d5;
24 ...l'k7 25 'ii'dl) 25 .i.c2 l:.c7 26 C) the same with 25 ...l:.d4.
'ii'd2 ! f6 and after 27 .i.b3 ! or 27
.i.f4!, Black gains nothing. D) 23 ...• a4 24 ~d2 J:d6;
c) 23 .. .'ii'a4 24 ~d2 (but not 24 E) the same, but with 24 ...l:.d5;
b3? 1Wa5),and there is nothing to F) the same with 24 ...l:td4.
be found.
d) 23 ...11i'g424 ~d2. G) 23.-..'ii'dl 24 ~fl 'ii'h5 25
e) 23 ...1fe5 24 .i.e3. ~d2 'ii'b5+ 26 'iii>gllid6;
In these last two lines, Black's H) the same, but with 26 ...J:r.d5;
possibilities are clearly exhausted. I) the same with 26 ...l:td4.
So stage one of the analysis has
failed to give the desired result. If In all cases Black threatens
Black were now short of time, he ... 'ii'd7. I quickly rejected the first
would have to revise his aim- let trio of variations in view of 26
us say, by seeking the best way to 'ii'e2. The third trio was discarded
win a second pawn. In that case, if only because of 27 .i.c3. In the
incidentally, he would have an ex- remaining cases, what I discovered
tra 'candidate move' to consider: was as follows:
t) 23 ...'ii'xb2 (which clearly falls D) 23 ...• a4 24 .i.d2 l:.d6 25
short of the original maximum aim). ii'e2 ! (25 ~l? 'ii'b5+ 26 'ife2
His choice would be between 'ilxb2 27 .i.d3 'i!id4) and then
'a21 ', 'a22' and 'f'. Fortunately, 25 ...'6'd7 26 .i.el or 25 ...... a6 26
however, I had enough time in hand 'ifel (better than 26 'ii'e3 'ifb6!),
(the to~mament was played at the intending to defend everything
'good old' rate of 40 moves in two with .i.c3 and .i.c2.
and a half hours), and I resolved to E) 23 ... 'ila4 24 ~d2 l:.d5 25
loqk for new ideas. The thought ~e4! and then 26 ~. as 25 ...l:.e5?
occurred to me to re-position my fails to 26 ~xh7+.
44 Wandering in the Jungle

F) 23 ... 'ii'a4 24 .i.d2 25 'ii'e2 and defences. On the other hand,

(better than 25 c;!;>fl'iib5+ 26-'ife2 the remaining calculation (stage
'ii'xb2 27 .i.d3 'ii'xa2) and then three already!) is not complicated.
25 ... 'ii'd7 26 .i.el or 25 ... 'ii'a6 26 With White's queen on e2, the black
'ii'e3!. queen's penetration to dl is deci-
In all variations Black is at a loss sive: 26Jlel 'ifdl 27 'ife4 f5, or26
how to wfo. So stage two of the 'i!ib5 l1d8 27 Jlel 'ii'dl 28 'iia5 b6.
analysis has also ended without The entire three-stage analysis
producing the results. (including re-checking) took ex-
Should Black after all renounce actly an hour. The game continued
the maximum aim and revise his 23 ...'ii'a4! 24 .i.d2 l1d6! (DJ.
analysis accordingly? I decided to
keep on searching, and I was not
disappointed. An idea came into
my head like a flash of lightning.
23 ..... a4 24 .id2 .:Ld6 25 'ii'e2
'i!fd4!! (DJ.

Now my opponent from Novo-

sibirsk unfortunately played 25
ffl(?), and after 25 ... ifbS+ he re-
signed (0-1). Black's main idea,
alas, remained off-stage ...
I may say with some pride that
Yes, the idea of going back with out of the many strong players
the queen is far from obvious. I be- (masters and grandmasters) to
lieve it would have been impossible whom I have shown this position,
to find if without calculating many only one managed to find the solu-
other lines ij,rst - without explor- tion on his own. An interesting
ing the jungle of multiple attacks point is that even one of the latest
 Wandering in the Jungle 45

'Mephisto' model computers (ad- like any method of organizing one's

mittedly not the strongest) failed to thinking - it can yield good results
find the right continuation, though once it is absorbed into a player's
you might think this would not be subconscious and he follows it
difficult for a computer. automatically. However, this can
I am far from imagining that my only be achieved by special train-
proposed algorithm is applicable ing, a matter to which too few
to all complex positions. Like any players, alas, give sufficient atten-
other algorithm, however - in fact, tion ...
3 Visual Imagination and Chess
Analysis
Beniamin Blumenfeld

Note: This article was first published in 1936

The prime concern of chess theo- improve their thought-processes

reticians is the study of openings. may also be significant, even if only
In practical chess, opening knowl- as material for future research.
edge undoubtedly plays a major My first example illustrates the
role - but not a decisive one. Even theme of visual imagination.
in the recent Alekhine-Euwe match
(1935), for all Euwe's massive
opening preparation and Alekhine's
dubious experiments, we can point
to very few games in which defeat
resulted solely from a bad open-
ing. At any rate, in the vast major-
ity of games the result was - or
could have been - affected in the
later stages by errors or superfici-
ality on the part of one (or often
both) of the players.
From this it follows that cor- B
recting the faults in our processes
of thought is no less significant Blumenfeld - Zhivtsov
than perfecting our opening Moscow Championship
knowledge. Semi-final
I hope the chess press and or-
ganizations will consider the pos- In this position Black put his
sibility of investigative work on this bishop en prise with 1...~e4?.
subject. The-ideas of practical play- This blunder astonished me all
ers on how to eliminate errors and the more since my young opponent
 Visual Imagination and Chess Analysis 47

had made a good impression both Conceptually, the movements of

in this game and in the tournament the bishop could be formulated
as a whole. He was not short of like this. From c2 the bishop is di-
time on the clock, and had taken rected against h7, where it then
quite a long time over his faulty lands; all the while you are aware
move. As I discovered from talk- that on its way to h7 it has a stop-
ing to him after the game, he re- ping place on e4, but this halt at e4
jected 1...Jlh3 ( 1....i.b7 2 lbg4) 2 is not actually visualized by your
'ii'xh3 'ii'xe5 3 '6'xh6 'ii'e 1+ 4 l:txel mind's eye. Similar errors are quite
gxh6, when the advantage is rather common in practical play.
with White, who has bishop for An especially frequent occur-
knight and a queenside pawn ma- rence is this: when pondering a
jority. When making the move variation, you forget to move a piece
1...Jl~ he imagined that White or take away a captured one in ac-
was unable to win a piece in view cordance with one of the moves; in
of 2 Jlxe4 'ii'xe5 3 .i.h7+ <;tixh74 your mind, the piece mistakenly
'ii'xe5 lbd':3+, regaining the queen remains on its original square.
with a knight fork. He overlooked It should be observed that in the
that after 2 Jlxe4 the a8-rook is at- majority of cases, such errors re- ,
tacked. sult not from weakness of the vis-
At first sight it seems incompre- ual imagination, but either from
hensible that the player with Black nervous haste or else from an in-
should see comparatively far ahead sufficient effort of concentration.
in the two variations quoted, and at When mentally playing a move, it
the same time miss the perfectly is tiresome to have to register the
obvious attack made by his oppo- change immediately in your mind's
nent's very next move. eye. Often you will take a short cut
As far as I can judge from my by making the move on the basis of
own experience in analogous cases, an internal dialogue. This may
the reason for this mistake was as take the form of mentally record-
follows. When Black was calculat- ing the notation of the move; or a
ing the line 1...Jle4 2 Jlxe4, etc., spatial movement may be traced
he didn't mentally place the bishop out in your mind, without however
on e4 but kept it so to speak in yielding a firmly fixed image of
mid-air, pointing at h7, ready to the resulting position.
give check with a discovered at- Let us move on to another ex-
tack on the queen. ample of a similar theme.
48 Visual Imagination and Chess Analysis

Black's castled position by ad-

vancing his f-pawn and, if appro-
priate, his g-pawn. Not knowing
what to do about it, I played
1...lbd4 to postpone the decision.
In view of the threat to win a
pawn with ... lbxf3+ followed by
...lbe5, etc., I was convinced when
I played my move that White
would reply 2 'ith2 - and I thought
to myself, "Too bad that after
B 1...lbd4 2 '.t>h2 lbxf3+ 3 .i.xf3
lbe5 4 Jlg2 Jlxh3, he takes on h3
Sergeev - Blumenfeld with his king (not his bishop), and
Moscow Championship I can't make anything of his ex-
Semi-final posed king position."
In the event White answered
This position was reached in my 1...lbd4 with 2 lbdl?, and now it
last-round game in the same event. was a full five minutes before I re-
In view of the tournament situa- alized I could win a pawn with
tion, a draw was enough for me. As 2 ...t2Jxf3+, etc. These five minutes
we all know, playing to draw is of thinking-time were spent in
not so easy: the urge to simplify dithering over what plan to adopt,
and the fear of complications and as I couldn't make up my
have a harmful effect. mind, I took a breather from these
In this position I played 1...~4. oppressive thoughts and came
I had used up more than half an back to my previous theme: "Too
hour on it, since I was in a quan- bad, after 2 ...lbxf3+ 3 .i.xf3 lbe5 4
dary. White was threatening to ex- .i.g2 .i.xh3 he takes with his king",
change my fianchettoed bishop etc. Then suddenly I saw that he
with .i.h6. I was not interested in couldn't, as his king was on gl, not
wasting time preventing this with h2.
1...l:te8, especially since the rook In my mind, then, during those
might come in useful on the f-file five minutes, the white king was
to support ...f5. On the whole I not on gl - where I could have
came to the conclusion that White seen it was, by simply looking -
had a clear plan for attacking but on h2, the square where I had
 Visual Imagination and Chess Analysis 49

mentalJy placed it in anticipation that the movements in the mind's

of lhy opponent's move. Suppose eye when pondering one variation
that after 2 ~d 1 I had found it easy hinder the correct visualization of
to take a decision, and hadn't come a position arising in a different
back to my old idea - "Too bad", line. It is clear that the more
etc. Then it is quite possible that I variations and the greater their
wouldn't have played 2 ... ~xf3+ length, the greater is the possi-
winning a pawn. bility of error.
The present example is all the A further point should also be
more interesting since the threat to borne in -mind.
win a pawn was in my mind when I Each time you mentally execute
played 1...~d4, and yet, after men- a move in a long variation, the re-
tally moving the king in the course sulting position differs more and
of my analysis, I forgot to put it more from the one currently on the
back; the image created by my board; hence the image in your
mind's eye prevented me from ob- mind becomes paler and paler.
jectively registering the king's lo- Here and there, you may find a
cation. chess player with a particularly
However, the explanation I have strong visual imagination, who is
just given is not the only one. The confident of correctly 'seeing' a
following is also possible. When I position reached at the end of a
was thinking about 1...~d4, I de- long variation; yet there can be no
cided that since White was going assurance that his evaluation of
to reply 2 <i.ih2,I could gain noth- that position will not be adversely
ing from 2 ... ~xf3+; this ready- affected by the paleness of the im-
made conclusion stuck in my mind age. This is something that happens
even though the premise (2 'iit>h2) to every chess player - after work-
was lacking. ing out a variation correctly, you
Of course it is hard to decide cannot decide whether it is in your
which explanation is correct in this favour or not. The chief explana-
particular instance. At any rate, as tion for this, as far as I can judge
far as I can judge from my own ex- from myself, is the inadequate clar-
perience, there are moments when ity of the picture received by the
an image created by the mind's eye mind. A chess player's thought is
succeeds in supplanting reality. bound up with visual imagin-
But if such cases are infrequent, ings. Therefore, the brighter and
it is surely a common occurrence more distinct the visual picture,
50 Visual Imagination and Chess Analysis

the more easily and precisely his even if his move was the one you
thinking will operate, and the expected, and you have your re-
richer it will be. ply ready. Your reply was, after
There is another danger en- all, prepared when the current
tailed by long variations: the position existed only in the imagi-
psychological strain of having to nation. It is quite possible that
register the changes, move by once your eyes can see the new po-
move, in your mind's eye is so sition directly, with all its peculi-
great that fatigue from this exer- arities - including your opponent's
tion can affect your subsequent last move - new thoughts will arise
play. owing to the greater clarity of the
Every practical player should picture.
clearly take stock of the role played You must exercise strict self-
by visual representation, and of the discipline when analysing varia-
hazards inseparable from tactical tions. In particular, don't let your
analysis; from this he should draw thoughts skip from one line to
the appropriate conclusions. bear- another and return to the same
ing in mind. of course, the quality line several times over. Instead,
of his own visual imagination. start by deciding the order (geared
For our own part, our conclu- to the specific context) in which
sions are as follows: the variations should be examined.
After your opponent's move Then proceed systematically from
you must begin your delibera- one variation to the next. In each
tions not from any ready-made line, at every turn, carry out and
decisions previously arrived at, register the movement of the piece
but so to speak afresh; before in your mind's eye. At the end of
anything else, let your eyes take the variation, carry out a resume.
the current position in. However Only then proceed to the next
strongly developed your visual variation.
imagination may be, it is perfectly In establishing the order of
obvious that an image in the mind analysis, your basic purpose must,
is less distinct than one received by as far as possible. be to restrict the
direct perception. Thus, when your quantity and length of the varia-
opponent moves, it never pays tions. When considering the oppo-
(except when in severe time-trou- nent' s possible replies to a move
ble, of course) to reply instantly, you are contemplating, the one to
without any further thought - examine first is the one which at
 Visual Imagination and Chess Analysis 51

first sight looks most dangerous. It few short lines which serve to
is only if you find a counter to this bring out the characteristics of
reply that you should consider the position.
other, less obvious moves for your If there is a choice between two
opponent. Similarly, if you judge continuations which carry roughly
that a clear, decisive advantage the same result (equality, edge, de-
emerges from a variation a few cisive advantage), preference should
moves deep, there is no point in be given to the line that involves
mentally extending that varia- less tactical analysis and conse-
tion by thinking about the detailed quently less danger of error. This
exploitation of the advantage. principle should be rigorously ap-
If the move you are playing is plied and should override any
completely forced, so that a branch- 'romanticism'. For instance, if the
ing of variations occurs only after choice is between reducing to a
this move and your opponent's re- clearly won pawn ending with an
ply, there is no point in immersing extra pawn and playing a mating
yourself'in analysis as yet. After combination with sub-variations
your forced move and the oppo- several moves deep, it is more sen-
nent's next one, the picture will be sible to choose the former. There
clearer and the analysis easier. The are familiar cases from tournament
same reasoning applies if you are practice where a player has an-
contemplating (for instance) a nounced mate in a few moves and
variation eight moves deep, and re- then gone on to lose, since the
alize that after the first few moves mate turned out to be illusory.
you can force a repetition, i.e. a re- Our arguments, especially this
turn to the current position. In that last one, will undoubtedly meet
case it is as well to play the first with opposition from the partisans
few moves without a long think, of 'beauty' in chess. In our view,
and afterwards calculate the varia- the analysis of variations is merely
tion to the end. If it turns out to be an indispensable technical proce-
unfavourable, you can return to the dure, and if this procedure can be
point of departure by repeating simplified or made easier, so much
move~. the better. Beauty in chess is a mat-
In situations that are not ter of the inner logic and richness
sharp, where there cannot be of ideas, which in most cases can
any forced variations, your cal- be adequately disclosed by a deep
culations should be confined to a study of the position; calculation is
52 Visual Imagination and Chess Analysis

necessary only to verify that the I know what happens at a simulta-

ideas are correct. Chess is a game neous display if the lighting is bad,
of purpose - the point is to achieve the pieces are painted in irritating
the desired result with the greatest colours, or the boards are the wrong
certainty. That is why we think our size for the pieces: the result, even
argument is right. against a weak contingent, can be
Visual imagination is so impor- worse than in a display against
tant for chess thought that physical stronger opponents in surroundings
G,Onditions which assist vision are more congenial to the eyesight. I
bound to make a difference. In par- suggest that chess organizations
ticular, this means adequate light- ought to consult specialist physi-
ing during play, a board and pieces ologists and psychologists, and
of the correct relative sizes, and a draw up a standard pattern of
colour scheme that is pleasing to equipment following their guide-
the eye. From personal experience lines.
4 How Chess Intuition Develops
Mark Dvoretsky

Chess players of universal style, To me it seems more helpful to

who handle any type of position categorize players according to
with equal competence, are ex- their predominant style of think-
tremely rare. Bobby Fischer was ing, their characteristic approach
one such. So, in his best years, was to taking decisions - the intuitive
Boris Spassky. As a rule, everyone's or the logical.
play, even that of the top grand- Grandmasters of an intuitive
masters, has weak points of one bent - Capablanca, Tai, Petrosian,
kind or ano'ther. It is very impor- Karpov are examples - have a deli-
tant to eradicate them in good cate feel for the slightest nuances
time, to bring the backward areas of a position, while possessing a
of your game 'up to scratch' - keen eye for combination. They are
without of course renouncing your relatively weaker when it comes to
creative individuality. planning and strategy; they are
Traditionally, chess players are none too fond of working out
classed as combinative or posi- variations, and commit errors in
tional. At one time it was relatively their calculations.
simple to categorize them on these At the opposite pole there are
lines, but today things are different players like Rubinstein, Botvinnik
- there are hardly any purely po- and Kasparov. They conceive pro-
sitional or purely combinative found plans in the opening and
players left. In any case, this clas- ensuing phases; they think in a dis-
sification refers only to the out- ciplined manner and calculate
ward manner of play, not to the variations precisely. Yet now and
underlying qualities of thought. It again they miss unexpected tacti-
gives you too little to go on when it cal ideas; they sometimes prove
comes to choosing the form and too single-minded and insufficiently
content of a training course suited alert to the critical moments of the
to a particular player. struggle.
54 How Chess Intuition Develops

Of course, all this is no more the present phase of Capablanca 's

than a rough-and-ready sketch. To career." (From Alekhine' s famous
'diagnose' a player I am coaching article The I 927 New York Tourna-
- whether a young candidate master ment as a Prelude to the World
or a grandmaster - I usually apply Title Match in Buenos Aires.)
a much wider set of parameters. Chess players of the intuitive
Nonetheless, the classification I cast of mind do well to train by
have suggested does seem to me to solving strategic problems (such
have a good deal of practical use. as the choice of plan in the transi-
tion stage from opening to middle-
It is a great boon for a chess game ). They can usefully apply
player to be naturally blessed with themselves to exercises in the calcu-
a highly developed intuition. Yet lation of complex variations, de-
there is also a serious psychologi- manding concentrated, painstaking
cal danger lurking here, as Alexan- attention. I once suggested to Al-
der Alekhine observed: "Speed of exander Chemin that he should try
comprehension; the ability to see, working along these lines. He soon
almost instantaneously, a whole made substantial progress, quickly
range of the tactical possibilities graduating from a run-of-the-mill
inherent in any complex position; master to a strong grandmaster and
economy of thought, and resulting participant in the World Champi-
self-confidence - the obvious ad- onship Candidates Tournament.
vantages which these factors bring
are almost inevitably associated For a player whose intuition
with temptations. A player can eas- needs to be developed, things are
ily adopt a false attitude: on ac- more complicated. Some chess
quainting himself with a position, players, and even their coaches,
he may assume that the good moves have no idea how this problem can
he sees instantly, or almost in- even be approached. In this chap-
stantly, must be the very best. In ter I shall offer some ·thoughts on
consequence, his play loses in it, based on my own coaching ex-
depth as much as it gains in facil- perience.
ity. Unfortunately (for the art of Chess intuition is the ability to
chess), this gradual renunciation perform easily and swiftly - some-
of the search for the absolute, this times instantaneously - the mental
tendency to be satisfied with moves act of grasping the character of a
that are merely good, characterizes position, identifying the main ideas
 How Chess Intuition Develops 55

in it, assessing how promising expression it is apparent that these

some particular continuation is. are all various manifestations of
Intuitive insight helps us to dis- intuitive perception as applied to
pense with lengthy and compli- chess. In principle it would be use-
cated calculations; it facilitates our ful to discuss each of them sepa-
search for the right move; it sug- rately, but that would be the
gests where a solution might be subject of a special investigation.
unearthed. In chess literature, oddly enough,
Serious study of chess, of the intuition tends to be equated sim-
kinds of rationale governing the ply with the capacity to embark on
struggle - intensive analysis of material sacrifices which don't
specific concrete situations - this lend themselves to exact calcula-
significantly develops and en- tion. Essentially, two concepts are
riches our intuition. I am not go- confused: the risk which goes with
ing to prove this assertion; it is the impossibility of calculating
illustrated in Part One of my book variations to the end, and intuition.
Secrets of Chess Training - see the
chapter 'On the Usefulness of
'Abstract' Knowledge'. I also ad-
vise you to consult Eduard Gufeld's
articles 'Intuition and Inspiration'
and 'How to Develop Intuition',
which are included in his book of
selected games, My Life in Chess.

Throughout the whole course of

a game we rely on our intuition
(with a greater or lesser degree of
frequency and success). It reveals
itself in the most diverse forms.
Consider some of the concepts Suetin - Bagirov
which we constantly employ: 'po- USSR Championship,
sitional.ft.air', 'the spirit of the po- Leningrad 1963
sition', 'an eye for combination',
'a sense of danger', 'a feeling for Vladimir Bagirov has just played
the initiative (the dynamics of the 17 ... .i.e7-d6!. 'Normal' continua-
game)' - from the very forms of tions lead to exchanges, and Black
56 How Chess Intuition Develops

obtains an excellent position. He certain that after sizing up some

has nothing to fear from either 18 variations Mikhail Tal would quite
il..xd6 lbd6 (19 'it'xd6?? il..xg2+) quickly have decided that the sac-
or 18 lhf6 il..xe5 19 lhc6 'it'xb2 !. rifice was promising (that is the
Alexei Suetin wrote: "What was I point - not provably correct, but
to do? I didn't want to go in for promising) and that he ought to go
simplifications. Suddenly I was in for it. Or, instead, he would have
attracted, indeed thrilled, by the assessed its consequences as fa-
prospect of a queen sacrifice. I fe- vourable to Black, and played dif-
verishly calculated variations. The ferently.
hands of the clock kept turning re- 18 'it'xg7+!? ~g7
lentlessly, while the calculations 19 il.xf'6+
grew more and more complicated. Suetin wrote: "Now it was my
All I could do was either settle for opponent's tum to cogitate. As later
a draw or take the risk, relying on became apparent, this may well
my intuition." have been the decisive moment of
It is clear from his comment that the struggle. Where should the king
Suetin spent a large amount of time go - to h6 or g6? Bagirov thought
trying to work out the sacrifice for a whole hour, and he too made
precisely, but failed. Of course, the his move chiefly on the basis of in-
bold decision that he took con- tuition ..."
tained an element of intuitive judge- A strange conclusion, isn't it? He
ment, but no more than an element. "thought for a whole hour", and
At heart, he was proceeding not in- played by "intuition"! In actual
tuitively but analytically - and he fact Black attempted to calculate
was certainly right to do so, given everything precisely, but was un-
the large material plus which Black able to do so, and made a mistake.
receives (a queen for just one What has this to do with intuition?
piece). Any defensive resource that We can see that Suetin is talking
White had not accounted for, per- about it without having a clear no-
mitting Black to parry the immedi- tion of what it is.
ate threats, could instantly have As it happens, the sacrifice was
ended the game in Black's favour. objectively unsound, as Andrei Lil-
Nevertheless, it is likely that ienthal demonstrated. The refuta-
some players would approach the tion is 19...~g6!:
problem differently, in a genuinely a) On 20 il..d3 Black has the
intuitive mann~r. For instance, it is powerful reply 20 ...il..e7!, attacking
 How Chess Intuition Develops 57

the d3-bishop with his rook. For 20 .:an?

example, 211J..xe7 lb.d3 22 l:tg5+ White intends 21 l:th5+ c;t>g622
c;t>h623 cxd3 l:r.e8! 24 Jrl.f6l:te6. l:.h4, with the threats 23 3Ld3+ and
b) The main line is 20 l:.afl 23 Jrl.h5+.
'ii'e3 21 Jrl.d3 c;t>h6!,and now: 20 ... 'ii'e3
bl) If White follows Tal's rec- 20 ...1J..e7! was stronger; it prac-
ommendation with 22 lt:idl 'iVd2 tically forces the draw with 21
23 %15f2(23 l£if2 ltg8), Black has l:th5+ c;t>g622 l:tg5+, because 21
23 ... 'ii'xf2! 24 lbf2 l:.de8 25 Jrl.c3 Jl.xe7 f6! 22 ii.xf6 lb.f6 23 l:.xf6+
f5 ! with advantage. <iif.ig7
24 llf7+ c;t>h825 Jrl.d3 l:td7
b2) 22 Jl.xd8 l:txd8 23 l:t.xf7 (or 25 ...'ii'xb2) is hardly dangerous
(after 23 l:£6+ c;t>g724 .:Xf7+ ~g8 for Black.
25 1J..xh7+ ~h8 White's attack is 21 l:thS+
repulsed) 23 ... l:td7 24 lt:id5 !? (or 22 l:th4!
24 lllf6+ c;t>g525 l:tf5+ c;t>h4!) The only defence.
24 ...1J..xd5! and then 25 lb.d7 23 l:thxf4
1J..xg2+! 26 ~xg2 'ii'g5+ 27 c;t>f3 Not 23 llfxf4 'ii'cl+ 24 Jl.fl,
'ilff4+ 28 c;t>e21i'g4+, or 25 lllf6+ which fails against 24 ...h5 ! 25 lt:ie2
c;t>g526 l:£5+ <iit.ih4 27 :Xd7 'ii'c 1+ 'iVxc2.
28 JJ..fl (28 :n 1J..xg2+ 29 ..t>xg2 23 ... hS
'ifg5+) 28 ... 1J..xg2+! 29 <iif.ixg2 23 ...'ifh3 24 lt:ie4!.
'ii'xc2+ and Black wins. 24 ii.xd8 lhd8
19 ... c;t>b6?(D) 25 1J..d3+ ltxd3!
In time-trouble Black hastens to
simplify the position. A weaker
line is 25 .....t>g726 :xn+ 'iirg8 27
1J..c4'ith8 28 l:.7f5 1J..e8 29 lt:id5,
with dangerous threats.
26 cxd3 'ii'xd3
21 :t6+ c;t>gs
28 lhf7 h4
29 'itgl 'i¥e3+
30 '1:.7r2(D)
30 'i\)b5?
White now obtains a decisive ad-
vantage. Black had to open up his
opponent's king position: 30 ... h3!
58 How Chess Intuition Develops

perfectly suitable for training in

analysis; it all depends which men-
tal approach you are adopting.)
When exercising your intuition,
you should not be trying to cal-
culate everything 'to the end'.
Instead, looking at the indispen-
sable minimum of variations,
you should try to reach a definite
conclusion as quickly as possi-
ble. Then compare your verdict
B with the 'right answer', and you
will see whether you were search-
31 gxh3 ~f3!, with a probable ing along the correct lines, or
draw. whether from the outset you were
31 lhe2! ~h6 32 lbf4 a5 33 missing some relevant ideas -
J:tdl a4 34 h3 ~h7 35 lbdS ... cS either points of judgement or of
36 lbf6+ ~g7 37 a3 ~g6 38 lt)g4 concrete tactics.
~h7 39 J:.el ... d6 40 tl)e3 ~g6 41 In the same way, you could try
q)f51id8 42 l:te6+ (42 lt)e7+ @g7 to select the right square for the
43 J:.e6 is even stronger) 42 ... ~7 black king on move 19.
43 ltld4+ ~g7 44 l:te4 ~d7 45 Some exercises of a similar kind
lt)f3 ~f5 46 l:d4 'ife8 47 l:xh4, ( classified by theme) are to be
and White gradually exploited his found in the book, Secrets of Chess
material plus. Training which I mentioned ear-
Let us return to the position lier.
where White was faced with his
problem. Pondering this kind of Why did White's attack succeed
obscure position is one way to in the example we have examined?
develop your intuition. You would The key factors were not purely
think about the position for a while technical (objectively the queen
and try to 'guess' whether the sac- sacrifice was incorrect), but psy-
rifice was sound, and whether it chological - the kind you must
was worth going in for it. Obvi- talce into account when intuitively
ously you could not entirely do weighing up the prospects offered
without calculating
/
some variations. by this or that solution to your
(Incidentally, this same position is problem. The effect of surprise
 How Chess Intuition Develops 59

was important (Bagirov had stud- Now, however, 15 ...e5? is met

ied the position after 17 ...it.d6 in by 16 d5 ~d4?! 17 loel!. Another
his home analysis, but hadn't no- bad line is 15 ... f5? 16 d5, when
ticed the queen sacrifice), but the 16...~e5 is met by 17 ~g5! (or 17
peculiarities of Bagirov's style were lbel!) 17 ...f4 18 it.d4! (but not 18
the main thing. He is a strong posi- it.xf4 l1xf4 19 ~e6 'ifb8 20 lbxf4
tional player but usually performs ~f3+ 21 gxf3 'ifxf4), while 16 ...f4
much more weakly in obscure tac- is no help due to 17 l:xd3 ! ! cxd3
tical situations. 18 dxc6 fxe3 19 cxb7 exf2+ 20
I once managed to profit from Wfl (now we see why White gave
this myself: up the exchange) 20 ...l1b8 21 'Wc4+
~h8 22 c6.
Dvoretsky - Bagirov 16 '6'xa5 fua5
USSR Championship 17 loel J..rs
(First League), Tbilisi 1973 18 l:tacl
Alekhine Defence Not as strong as 18 d5!, which I
later played against W. Martz at
1 e4 ~6 Wijk aan Zee 1975.
2 eS ~dS 18 lt)c6!
3 d4 d6 19 g4 it.d7
4 c4 ~b6 20 dS lob4 (D)
5 exd6 cxd6
6 ~ g6
7 h3 it.g7
8 lbf3 0-0
9 ~e2 ltx6
10 0-0 .i..fS
11 ~e3 dS
12 c5 lbc4
13 hc4 dxc4
14 'ifa4 it.d3
This is a well-known variation
of the Alekhine Defence, of which
Bagirov is a connoisseur. Subse-
quent games have convinced me
that Black equalizes with 14...e5!. If now 21 a3, then after 21...~6
15 l:.fdl .. a5! and 22 ...l1ac8, the c5-pawn is very
60 How Chess Intuition Develops

weak. Overall, Black has an excel- 29 lta3 hS (D)

lent position. With this - and my It was worth considering 29 ... a5
opponent's style - in mind, I de- 30 lba2 l:txcl+ 31 lt)dxcl a4. The
cided to complicate and provoke pawns would be blocked, but at
Bagirov into a piece sacrifice. least they would be that bit further
21 b3!? ~a2 advanced.
2 I.. .cxb3 22 axb3 e6 ! was sim-
pler, with an approximately equal
position; however, the temptation
proved too strong.
22 lt)xa2 cxb3
23 lt)c3 .J:uc8?!
This is the kind of thing I was
banking on - Bagirov already com-
mits a serious inaccuracy. He hopes
to prevent 24 Ji.d4, but his move
fails to do so.
23 ... a5 was stronger; then on 24
.i.d4, Black has either 24 ... a4 25 w
/;.xg7 r:J;;xg726 l:bl l:tfc8 27 lt)d3
.:ta5 ! (intending ...:Xc5 or ...Ji.b5), 30 gxhS
or 24 ...e5 25 dxe6 .1Lxe626 Ji.xg7 30 f3 !? is more solid.
'3;xg7 27 l%d4 l:tfd8 28 ©f3 h5!? 30 ... gxhS
29 gxh5 iLf5 30 ©a4 gxh5, with 31 h4 iLfS?!
chances for both sides. It is incomprehensible why
24 Ji.d4! Ji.xd4 Black refrains from 31...a5!? 32
25 lhd4 lhcS lt)a2 lucl+ 33 lt)dxcl a4 (threat-
26 .:.b4 ening 34 ... l:tc4) 34 lt)d3 .i.f5
I considered 26 lt)d3 less accu- (34 ...J:lc4? 35 lt)e5). With 35 lt)e5 !
rate on account of 26 ... b2! 27 White would preserve some win-
lt)xb2 l%ac8. ning chances, but nothing more.
26 ... • :acs We can see that in a complex posi-
Or 26 ...b5 27 .:.Xb3 a5? 28 lt)d3 tion Bagirov is unsure of himself.
and, when the rook moves, 29 He usually avoids such situations
l2Jxb5. and has inadequate experience of
27 lhb3 bS them. Hence his intuition fails him
28 lt)dJ / :Sc7 here.
 How Chess Intuition Develops 61

32 lbe2
After this, Black's position is
hopeless. The indicated line was
32 ...~xd3 33 l:.xc7 lhc7 34 l%xd3
a5. In the endgame, passed pawns
must be pushed!
33 lbdxcl .:c,
34 lbb3 ~e4
35 lbbd4 ~xd5
36 lbxb5 l:.c4
37 tl:\bd4! .:c1
37 ...e5 38 :.as!.
38 'it;>h2 eS?!
39 l:ta5! exd4 has 48 ...f6! 49 ltg6 (49 h6 l:l.d7!)
40 lhd5 l:tc2 49 .. J::td7! (but not 49 ...<it>f7?50
41 lbxd4 l:txf2+ l:tg7+ <it>e651 h6!). The rook ex-
42 <it.>g3 l:td2 change after 50 lhf6+ J:tf7 leads
43 l:tgS+ <it>f8 to a draw, but otherwise Black ob-
44 lbfs a5 tains counterchances with 50 ...l:.a7.
45 lhhS <it>g8 Incidentally, it is essential for Black
46 l:.gS+ to insert 47 ...l:tdl !, because the im-
The sealed move. It was a good mediate 4 7 ...f6 48 l:tg6 l:.d7 allows
moment to adjourn. The position is White an easy win with 49 l:txf6+
definitely won for White, but at :t7 50 lhf7+ ~xf7 51 lbd4 (or 51
this stage he has to figure out the lbd6+) 51 ...a4 52 lbb5 ~f6 53 'it>g4.
right plan for realizing his advan- 47 'it>g8
tage, and this is best done 'at 48 lth6 a4
home'. 49 l:.a6 J:.a2
46 ... 'iti>fS(D) so ~g4 a3
47 .mis 51 ~h5 f6
At first it seemed to me that 4 7 51...l:tal 52 'at>h6leads to the
h5, was the simplest way to win, same thing. The most stubborn de-
e.g.: 47 ....;tdl 48 'wtg2 .:d2+ 49 fence was 51...<it>h752 l:ta7 l:.f2!,
'it>h3l:td 1 50 tl:\g3 (now the a-pawn but even then White would win by
is attacked) 50 ... a4 51 h6 l:r.d6 52 53 l:.xf7+ 'it;>h854 '.tg6 l:.g2+ 55
h7. However, I then discovered ~h6 ~g8 56 l:.a7 a2 57 h5 l:r.b258
that after 47 ... l:.dl 48 ~g2 Black tl:\h4 ~f8 59 lbg6+ <it>e860 tl:\e5 !
62 How Chess Intuition Develops

(preparing 61 ~g5) 60 ...l:.g2 61 On c6 the knight attacks the a7-

~d3 and 62 ~b4. pawn. It restricts the mobility of
52 ::ta7! :tat Black's rook, and of his minor
53 ~g6 l:.gl+ pieces too. On the other hand, from
S4 ~6 :tat f5 the knight would control d6 and
SS ~h6+ ~h8 prepare the advance of the passed
56 lbr7+ 1-0 d-pawn. What is more important?
To work out the variations accu-
The type of inscrutable position rately over the board is quite im-
which can serve to test and refine possible - after going through a
your intuition is often, but not al- few tentative lines, you are bound
ways, bound up with sacrifices of to fall back on intuition.
material. In his book The Test of Time
Kasparov repeatedly points out
how, in complex situations, his in-
tuitive appraisal of a position would
tum out to be right. He is evidently
proud of his intuition and consid-
ers it his strong point. Clearly,
though, any top-rank chess player
can boast of plenty of cases where
he solved complex problems cor-
rectly. To form an objective judge-
ment of how well developed a
chess player's intuition is, it is
w more important to trace how often
it lets him down. The young Milch-
Kasparov - Karpov ail Tai, for example, would almost
World Championship match (6), always hit on the strongest course
Moscow 198415 in sharp positions, finding the most
dangerous attacking resources. Nev-
Black is a paWQup, but the ac- ertheless a careful study of Kaspa-
tivity of White's pieces more than rov's games has convinced me that
compensates for the small material his flair is by no means impecca-
deficit. Clearly White must now ble. Even in his best games he
move his knight forward. But would often 'lose the thread' at
where to - f5 or c6? some point and give his opponents
 How Chess Intuition Develops 63

extra chances (which, to be sure, 21 ... .i.f8!?

they didn't always utilize). When I made this move I was
So it was in this example. Kas- very much hoping that the ex-
parov 'guessed wrongly' and missed World Champion would be tempted
the win. Afterwards he failed to to start an attack on my king with
sense when the moment had come 22 ~xe5 .i.xa2 23 .i.a 1! ( threaten-
to force a draw, and ended up los- ing not only to take the bishop but
ing. You will find the game in the also the deadly 24 'ifc3) 23 ...'tib3
Supplement to this chapter. (the only defence) 24 'ii'd2. The
variations would seem to turn out
Correct intuitive appraisal of a in his favour, for instance 24 ...hc5
situation helps a player to appor- 25 'ii'g5 (with threats of 26 'ii'xc5
tion his thinking time rationally; it and 26 'ii'f6), or 24 ...l:.xe4 25 .i.f3
. tells him when he needs to concen- l:txe 1+ 26 l:.xe 1 (threatening 27
trate and examine the variations 'ii'd4) 26 ....i.g7 27 .i.xg7 q;xg7 28
thoroughly, and when, on the con- l:tal, and Black loses a piece. Such
trary, for orle reason or another, an attack would be wholly in Tal's
there is no point in going in for de- style.
tailed calculations. After a little vacillation, how-
ever, the grandmaster chose a quiet
continuation in preference to sharp-
ening the struggle.
22 l:tbl! 'ir'd7
23 l:tedl .i.xd6
24 cxd6!
White went on to realize his po-
sitional advantage by technical
means.
The grandmaster's feel for the
position had not deceived him. On
22 ~xe5? I had prepared the unex-
B pected stroke 22 ...~b3!!, equaliz-
ing. Even though Tai didn't see
Tai - Dvoretsky this, he intuitively made the right
USSR Championship, choice.
Leningrad 1974 His decision was based on a true
assessment of the situation on the
 64 How Chess Intuition Deyelops
,

board. Tai writes that he didn•t who were deep in analysis of a dif-
want to "give the black pieces free ficult position and asked Vasily
rein". Why indeed should he go in Smyslov for his advice. After a
for complications. at the risk of short think the latter said, "You
making mistakes in calculation, should place the rook on the fifth
when his opponent's pieces are al- rank." This recommendation ap-
ready condemned to passivity and peared all too abstract, yet after a
White's obvious advancage can be while he repeated it: "Come on,
preserved by simple methods? rook to the fifth!" They started
analysing along these lines and
I dare say it is now tJ.meto sur- soon understood that Smyslov was
face from the stormy ocean of absolutely right.
combinative complexities and talk You can find examples of this
about the placid positional tasks kind of judgement by studying
and relatively simple tactical prob- game annotations, especially those
lems that we encounter almost at written by the intuitive type of
every step. player.
In some books you can read that
the process of evaluating a position
consists in isolating and weighing
up all the positional factors that
play a part in it. Nonsense! In ac-
tual fact, most of this task is per-
formed subconsciously. The art of
evaluation lies in understanding
the essence of a position- identi~
fying the crucial problem (either
positional or tactical) tl!at needs
solving - sensing the right direc-
tion for our investigations and B
detecting the desirability or oth-
erwise of a particular operation. Nimzowitsch - Capablanca
It is clear that a well-developed in~ New York 1927
tuition will enhance the speed and
accuracy of our perceptions. Capablanca wrote: "White has
You may have heard the old finally prepared the freeing ma-
chestnut about some grandmasters noeuvre b2-b4 followed by .i.b2.
 How Chess Intuition Develops 65

Black has occupied both open files (counting on 28 ....:.Xb2? 29 lhf8+

with his rooks, and is ahead of his with perpetual check, or 28 ...'i¥xb2?
opponent in development. He now 29 if d6) because of 28 ...Wb1+ ! 29
has to realize his advantage one 'it>g2'ffxb2. The right continuation
way or another, before White fin- is 27 l:.d2, and if 27 ... ..i.xa3, then
ishes developing his position." 28 'ifa6!: 28 ...l:tf8 29 'ii'xa5 'iWbl+
20 ... 'ife5! 30 'it>g2.i.xb2 (30 ... W'e4+? 31 f3)
"A subtle move with the pur- 31 'i!Vb4 or 28 ...l:tb8 29 'ii'xa5 ! and
pose of gaining time to bring the the b2-bishop is invulnerable.
queen into the fray. Black wants to 24 ... 'ih>3
occupy the seventh rank with one 25 .i.d4?
of his rooks, and the queen's help Amazingly, Nimzowitsch doesn't
is essential for this. The tex.t-move understand that he must try to save
prevents the immediate 21 b4, as himself by exchanging rooks. 25
there would follow 21 ....i.d6 22 g3 l:1ac1 ! was essential.
'ife4, and Black gains control of 25 ... l:tc2
the seventh. 26 'ifa6? (D)
As you see, Capablanca has for- White could defend more stub-
mulated his chief goal: seizure of bornly with 26 'iffl or 26 Vdl,
the seventh rank. (It is also clear preparing 27 l:te2. After the move
what White wants: to complete his played, Black gives a text-book
development and start ex.changing demonstration of the power of
rooks.) Without concrete analysis seventh-rank control.
it is hard to foresee who will be
more successful in carrying out his
plans. But what the players must
aim for is at any rate clear.
21 g3 'ifd5!
22 b4 .i.f8
23 .i.b2 'ifa2!
24 :tat
Alekhine suggested 24 l:r.bdl !?
l:txd 1 (if 24 ... a5 at once, then 25
:Xd8 :Xd8 26 ..i.d4!) 25 .:Xdl. Af-
ter 25 ...a5 26 bxa5 bxa5 (26 ....i.xa3
27 'ifa6 !) White cannot, unfortu-
nately,' play 27 'ii'a6? l:c2 28 lld8
66 How Chess Intuition Develops

26 eS! a good game by a grandmaster,

27 ~xeS l:.dd2 with his own detailed notes. After
28 'iVb7 playing through the opening, start
If 28 l:[fl, then 28 .. .'ifxe3! 29 trying to guess his moves, one by
~f4 J:.xf2! is decisive. On 28 'ii'fl, one. Give yourself a very short time
Black has 28 ...1i'd5 29 .i.d4 'ifh5! (for instance half an hour) for the
(more precise than 29 ...'ii'f3 30 whole game. Then compare your
l:.acl) 30 h4 (otherwise 30 ...lhf2) guesses with the grandmaster's ac-
30 ...'iff3. tual moves and comments.
28 lht'2
29 g4 'ife6 I once played an interesting
30 ~g3 lhh2! kind of game with Sergei Dolma-
Better than 30 ...'iVxg4 31 :n. tov. It was aimed at developing his
31 'iVf3 , feel for the endgame. I would open
Or 31 .i.xh2 'ifxg4+ 32 ~h 1 a volume of lnformator at the End-
'ifh3, forcing mate. ings section, pick a number at ran-
31 .:hg2+ dom, take the position with that
32 'ifxg2 l:.xg2+ number and set it up on the board.
33 <ati>xg2 'ii'xg4 Dolmatov would play the side
34 l:adl hS which had to achieve something -
35 l:td4 'iVgS a draw from the worse position, or
36 <tih2 a5 a win from the better position. To
37 l:.e2 axb4 begin with he would think for five
38 axb4 ~e7 minutes, then we would play the
39 l:[e4 .i.16 position as a blitz game, with the
40 %tf2 'ii'dS trainer making use of the analysis
41 l:.e8+ <tih7 in the book. In some cases we would
0-1 introduce an extra rule whereby
Dolmatov had the right to stop the
Various training exercises which clock once in the game, in a posi-
force you to take decisions quickly, tion he considered critical, and
without thorough reflection, are think for another five minutes. If
very useful for developing your in- you like, you can compete against
tuition. a friend at this game. Use two
At the first session of our chess separate volumes of lnformator,
school, Grandmaster Yusupov rec- and talce it in turns to play the role
ommendeda 'guessing' g~e. Talce of trainer.
 How Chess Intuition Develops 67

However, the most effective form happens, I have played with Dol-
of training for the fast appraisal of matov, Yusupov and other chess
positions was a game I shall now players I have coached. Your over-
describe to you. Unfortunately you all thinking time is slightly in-
cannot play it without a coach or creased, to 20-25 minutes (only
partner, and without a ready-made grandmasters are restricted to 15
stock of special positions. (This minutes). You play the same way
need can be met, however, by a • as before, except that if you get an
computer using the coaching pro- answer wrong, your clock is put
gram I have developed.) forward by one third of your origi-
Give yourself (let us say) 15 nal thinking time (i.e. by 5 minutes
minutes on the clock. Within this if you started with 15, by 6½ min-
time you have to find the right utes if you started with 20, etc.). To
move in five different positions. 'succeed' in the series, you have to
The first position is set up on the get through the five positions with-
chessboard and the clock is started. out overstepping the time-limit. As
On reaching your decision, play you will have grasped, under tht:;se
the move on the board and stop the rules you can 'win' even if you
clock. Then the next position is set make one mistake. You can hardly
up - and so on. All five positions win with two mistakes (you would
have to be solved before your flag be left with too little thinking
falls. The tasks (whether positional time), and with three mistakes a
or tactical) should not be too com- win is logically impossible.
plicated or demand deep calcula- The game ends the moment
tion, yet some should be simpler your time runs out. On the other
and some more difficult. You will hand it is possible to win 'prem-
need to economize your time to the aturely' - if the thinking time you
full, to avoid getting into bad have in hand for the last one or two
time-trouble over the final posi- positions exceeds the penalty that
tions. But playing too quickly is you would suffer for giving wrong
dangerous too - you could easily (but instantaneous) answers. In
make a silly mistake. You have this case you are not required to
'won' if you solve all five positions complete the solutions.
correctly. Otherwise you have suf- You may also, of course, carry
fered a greater. or lesser defeat. on playing if your flag falls before
There is also another version of you finish the series. It makes sense
this game - a version which, as it to do this if you have a rule ( aimed
68 How Chess Intuition Develops

at increasing the seriousness and on the number of excess minutes

responsibility of your decision- you require.
taking) which lays down a 'pen- Now let us solve one such
alty' for losing on time, dependent 'series'.

1. White to move 3. White to move

2. Black to move 4. White to move

 How Chess Intuition Develops 69

for an exhaustive scrutiny of

variations - you have to have
confidence in yourself and
take decisions boldly.
• It helps you to combat time-
trouble, since you continu-
ally have to regulate your
time expenditure.
• It gets you into good com-
petitive shape before a tour-
nament. It enhances your
5. White to move reflexes and speed of thought;
there is no chance of emo-
Solutions are given on pages tional fatigue setting in, as the
273-5. game is lively and exciting.

Practice J}as shown that this In conclusion I would like to

form of training is extremely use- draw together the basic ideas ad-
ful if you take it seriously. It devel- vanced in this chapter, to form
ops a number of important chess something like a set of rules for in-
skills at once: dependent work along the lines I
• It improves your intuition, have indicated.
your capacity to grasp both
the tactical and the strategic Aide Memoire:
details of a position quickly Recommended Precepts
and reliably. for Developing Your
• It cultivates your thought
Intuition
processes. It nurtures the habit
of instantly picking out the 1. Pay careful attention to
available •candidate' moves as your impressions; try as often as
well as the opponent's main possible to predict your conclu-
threats. Without this you can- sion in advance. To learn to guess
not achieve success - with right, you have to practise guess-
limited thinking time, mis- ing constantly.
takes will be inevitable.
• It increases your resolute- 2. Don't just remember your
ness. There is simply no time first impressions - observe how
70 How Chess Intuition Develops

your feelings change as you look 6. Consider not just the purely
more deeply into the position. technical factors, but competi-
You may hit on the truth at any tive ones - the situation in the tour-
conceivable stage of your investi- nament, your reserves of time and
gation into a position - but endeav- energy, your opponent's personal-
our to do so as early as possible. ity, the likelihood of errors on his
part, etc.
3. On discovering the 'right an-
swer', don't forget to compare it 7. 'Meta-intuitive' considera-
with your hunch. Ascertain what tions are very important. For ex-
ideas and themes prove to be the ample: in a given situation, can
most important and influential in a intuition be trusted? Does the posi-
particular position, and ask how tion on the board lend itself to pre-
far you took them into account in cise calculation, and how much
your deliberations. point would such calculations have
- how much time would you be
4. A very wide spectrum of in- likely to need for thinking about
tuitive perceptions is possible. your move?
They may not always tell you the
best move; they may relate to some 8. Analyse your own perform-
specific points of evaluation, the ance; if necessary, modify the rec-
desirability of some particular op- ommendations laid down here, and
eration, a feeling for danger, etc. work out fresh rules.

5. Relative evaluations are gen- 9. 1i-y to find the types of chess

erally more use than absolute exercise that have the maximum
ones. Verdicts such as 'the posi- effect on developing your intui-
tion is drawish' or 'the opponent's tion. Try to guess the right move
position is hopeless' are rather quickly in relatively simple situa-
crude and by no means always tions and also, conversely, in posi-
conducive to a solution. Of much tions that defy exact calculation.
more importance are specific in- Devise training exercises and
ferences which have to do with games which require you to take
comparing various moves, plans, intuitive decisions. It may be use-
ideas and prospects, or assessing ful to play games at fast time-con-
the dangers and difficulties that trols and study the games of
face you./ intuitive players, etc.
 How Chess Intuition Develops 71

10. Don't expect immediate re-

sults, but have firm confidence in
eventual success. Purposeful ef-
forts along the lines I have indi-
cated are certain to develop your
intuition. As a result your play will
be more relaxed, more assured,
quicker and sounder.

Supplement

Kasparov - Karpov w
World Championship match (6),
Moscow 1984/5 16 cxb5?
Queen's Indian Defence The first (and by no means the
last) occasion when Kasparov's po-
1 d4 lDf6 l,.c4 e6 3 lDf3 b6 4 g3 sitional flair lets him down. White
.i.a6 5 b3 .i.b4+ 6 .i.d2 .i.e7 7 could have gained the advantage
.i.g2 0-0 8 0-0 dS 9 lDeS c6 10 by 16c5!: 16...lt'.la417'iic2(threat-
.i.c3 lDfd7 11 lDxd7 lDxd7 12 ening 18 e5!) 17... e5 18 lt'.lf3, or
lDtl2 .:cS 13 e4 bS 16...b4 17 .i.b2 tbc4 18 lDxc4 .i.xc4
The consequences of 13...c5 19 'ii'c2 .i.b5 20 a3.
were examined by Artur Yusupov 16 ... cxbS
at the second session of our chess 17 :et .i.a3
school, in his comments on the en- 17...b4!?.
counter Yusupov-Sax, Rotterdam 18 :c2 lt)a4
1989. See the chapter 'Unexpected 19 .i.al lbcl
Moves in the Opening', in our 20 1i'xc2 'ili'aS?
book Opening Preparation. 20 ...'ili'e7! was stronger, not only
Incidentally, our discussion of preparing 21 ...l:tc8 but also pre-
this Kasparov-Karpov game will be venting the d4-d5 break.
based on Yusupov's deep analysis 21 'iVdl!
published in Sovietsky Sport within 21 d5 is premature on account
a day of the end of the game. of 21. ..lk8 22 li'd3 (22 'ii'dl .:et)
14 .:Cl dxc4 22 ...lDb2!. White wants to play
15 bxc4 lDb6?! (D) ~b3 and only then d4-d5.
' ...bxc4 1s
15 . better. 21 ... l:tc8?!
 72 How Chess Intuition Develops

21...lbc3 is strongly met by 22 d) Thezwischem:ug 27 ...Jlb4!?

ltib3! 'iib4 23 -.c2 (23 ... ltixa2? 24 is more tenacious, but White sim-
J:te3). Instead Black should play ply replies 28 :e2:
21 ....i.b2 22 it)b3 1i'b4. d 1) Black cannot save himself
22 tbb3 'ir'b4 by the exchange sacrifice 28 ...tbc3
23 d5 exd5 29 .i.xc3 .i.xc3 30 Ci::le7+~f8 31
24 exd5 ltic3 tbxc8 Jlxc8, since he loses a pawn
25 ~d4! ~xd4 as well: 32 d6 .)1.b4 (32 ...b4 33
26 lbxd4 fua2 (DJ :e7; 32 ...~e6 33 ~d5) 33 .i.c6
26 ....i.fS and 27 ... b4 was more .i.e6 34 d7 ct;e? 35 ..i.xb5.
cautious. d2) All that remains is to try
28 ...lbcl 29 J:le4:
d21) 29 ...it)b3 30 Jlxg7! J:tcl+
31.i.fl r!xfl+ 32 'ii;,g2!.
d22) 29 ...lbd3 30 .i.fl .i.f8 31
l'iJe7+ .i.xe7 32 %txe7 b4 33 d6
<it>f834 lte3 ! ltJcS 35 .i.d4 .i.xfl
36 Jlxc5 lbc5 37 d7.
d23) 29 ...~f8 30 d6 b4 (or
30 ...ltib3 31 d7 l:k 1+ 32 .i.fl l:tdl
33 ~d4 l'iJxd4, and now not 34
ltixd4? b4 35 d8'il .i.xfl 36 f3
Jld3+ 37 ci.>t2Jlxe4 38 fxe4 b3,
w but 34 .D.xd4!l:.xd4 35 ltixd4 .i.e7
36 ltic6) 31 lbh6+! (31 d7?! is
In this position, Yusupov has weaker: 31..ltdS 32lbh6+ gxh6 33
shown that 27 lDf5 ! gives White a l:tg4+ .iJ..g7 34 JJ..xg7 h5 35 ltg5
decisive plus: lbe2+ 36 @hl h6 37 il.xh6+ ~h7,
a) 27 ...g6? fails to 28 d6 gxf5 when 38 .i.g7? fails to 38 ... f5 ! 39
29 d7. .i.c6 .:txd7! 40 i.xd7 .i.b7+)
b) If 27 ... ..i.fS, then 28 d6 is 3 l...gxh6 32 l:tg4+ .i.g7 33 l'l.xg7+
again very strong. q.>f8 34 JJ..d5 ltJe2+ 35 ct;g2 tbc3
c) On 27 ...l:tcl, White has 28 36 :Xf7+ 'i!i>g837 Jlxc3 bxc3 38
:Xcl ilxcl 29 d6 .i.g5 30 h4 .i.c8 lua7+@f8 39 :Xa6 c2 40 d7 cl 'ff
(30 ... .i.d8 31 lbe7+ @f8 32 4Jc6 41 dxc8'1i'+ 'ii'xc8 42 l1a.8.
@e8 33 ~h3!) 31 hxg5 .i.xf5 32 Now let us look at what hap-
il.c6 ..t>rs 33 Jld4 ltlb4 34 .i.xb5. pened in the game.
 How Chess Intuition Develops 73

27 ©c6? ~c5! 31 l:txa7) 30 h4! lbc3 31 d6 (31

27 ...il.d6? would be a mistake in lha7?! il.c4 32 d6 ©b5) 31...il.b5
view of 28 ii.e5 ! l:.e8 29 l:.al .1..xe5 32 l:..xa7 l1d8! (32 ...~fS? 33 l::tb7!
30 lixa2 il.b7 (30 ...il.c8 31 1:.e2f6 with the threat of 34 d7 and 35
32 f4 Jtg4 33 lle4) 31 lha7 ~xc6 ii.xc3; 32 ...J:lb8? 33 d7 l:.d8 34
32dxc6 ~fS 33 jlli3! :_e7 34.i.d7, ii.h3 with advantage to White) 33
and Black will have to give up his l:.b7 .ile8 34 l:txb4 ©b5 35 il.e5
bishop for the rampaging pawn. f6! (35 ...~f8 36 il.d4! gives Black
28 ii.h3?t a difficult position) 36 il.dS+ ~f8
Though natural, this move is in (stronger than 36 ....i.f7 37 llxb5
fact dubious. It will soon become fxe5 38 ii.e4) 37 ~b2 and now
clear that the bishop has switched 37 .. J:txd6! with a draw, but not
to a worse diagonal, while the 37 ...©xd6? in view of 38 .ila3 ! in-
black rook withdraws to a more tending 39 l:tb8 !.
advantageous square. However, 28 :as
White no longer had a win: 29 il.d4 Jtxd4
a) 28 iLQ4 ii.xd4 29 ©xd4 (29 30 ©xd4 ~f8
©xa7+ ~fS 30 ©xc8 .i.xc8 31 d6 31 d6 (DJ
.i.f6! 32 .i.c6 .i.e6) 29 ...l:tcl !? 30 31 l:C.allbb4 32 lbc6 l2Jxc6 33
lhc 1 ©xc 1 31 d6 <ifr'f832 d7 (32 dxc61Lc8 is in Black's favour.
lbc61Lc8!; 321Lh31Lb7!; 32 .i.c6
g6!) 32 ... <.te7 33 ©c6+ c;;t>xd734
©b8+ ~c7 35 ©xa6+ ~b6, and
Black's two united passed pawns
fully compensate for the piece lost.
b) 28 il.eS l:r.e8!(unsuccessful
alternatives are 28 ... ©b4? 29 d6
©xc6 30 d7 lbxe5 31 dxc8'if +
ii.xc8 32 l:txeS il.f8 33 l:txb5, and
28 ...1Lb7? 29 d6 il.xc6 30 .i.xc6
lhc6 31 d7 ~e7 32 il.f6 gxf6 33
l:.xe7 l:td6 34 l:te8+ ~g7 35 d8'ff
•l:.xd8 36 ltxd8 a5 37 ~fl) 29 lLil B
©b4 30 ~b4 il.xb4 with equality.
c) 28 lbe7+ (probably the most 31 ... ltlc3!
dangerous try) 28 ...1Lxe7 29 lhe7 Black could already have made
b4! (not 29 ...l:tcl+? 30 ii.fl c;;t>fg certain of the draw with 31...%td8
74 How Chess Intuition Develops

32 d7 .i.b7 33 ~xb5 .i.c6, but Kar-

pov astutely senses that the posi-
tion has turned in his favour as a
result of his opponent's unsure play,
and he decides to try for the win.
32 ltk6?
At this stage Kasparov was
oblivious to the danger. He should
have forced a draw by 32 .i.g2 !
l:td8 33 .i.c6 (threatening 34 d7)
33 ...~cS 34 ~xb5.
There was also another, less ob-
vious, method, which was pointed
out by Vadim Zviagintsev: 32 d7 l:xb8 36 Ji.xb7 b2 (36 ...l:d8 37
.i.b7! (defending against 33 ~c6 .i.c6) 37 .i.c8, and White wins.
or 33 .i.g2), and now not 33 ~f5 However, Karpov is fully alert to
J:d8 (since on 34 ~xg7 or 34 ~d6 the danger.
Black has the simple 34 ... .i.c6), 34 toes?!
but 33 l:tal!. The point is that 34 l:al .i.xc6 35 .i.xc6 l:te6 36
33 ... a6 is answered by the pretty :Xa7 was a more resilient line, al-
34 loc6! .i.xc6 35 lha6. The inter- though it isn't at all simple to hold
esting try 33 ... a5?! is met by the the endgame after 36 ...lhd6.
zwischenzug 34 l:a3 ! (stronger 34 ... f6!
than 34 l:txa5 rJ.e7) 34 ...b4 35 35 d7
:Xa5. Hence Black's best course is White can no longer save him-
to settle for the draw with 33 ...rJ.e7 self. 35 .i.xb7 :Xe5 ! 36 l:tal b4
34 J:el+ ~8 (but not 34 ... rJ.d6? and 35 lnd7+ 'iPf7 36 l:al .i.xg2
35 l:e8 rJ.c7 36 ~6!) 35 l:tal. The 37 rJ.xg2 ~e6 are equally bad con-
moves could also be transposed, of tinuations.
course: 32 :lal Ji.b7 33 d7. 35 J:d8
32 -· .i.b7! (D) 36 .i.xb7 .fxeS
A draw would result from 37 .i.c6 'l;e7? (D)
32 ... b4 33 d7 b3 34 d8 ... + :Xd8 35 A time-trouble error. There was
~xd8 .i.d3. an easy win by 37 ...e4 ! 38 ltal
33- ~g2 .:eS! ~e7 (38 ... ~e2+ 39 'iPfl ~d4 is
Kasparov may have been count- also possible) 39 l:txa7 'l;d6 40
ing on 33 ...M? 34 d7 b3 35 ltJbS! .J:.a6<lilc741 'iPfl b4.
 How Chess Intuition Develops 75

40 lhb5 ~c6
41 lthS?!
41 lte5 ! was stronger, and if
41.. .l:ta8, then 42 l:te6+ 'iPc5 43
l:te7 a5 44 lbg7, attacking the h-
pawn too.
41 ... h6
42 .:te5 :as
The sealed move; 42 ....:tdS was
not bad either. The endgame is
quite interesting, but we will break
off here,. since from this moment
on the contest depended on ad-
38 hb5! lbxb5 journment analysis rather than the
38 ... <it>d6!?39 ~d3 l:txd7! 40 ability to find the strongest moves
.i.xh7 a5. over the board. Black won on the
39 ~eS+ 'iPxd7 70thmove.
5 Chess in the Style of Jazz
Sergei Dolmatov

I want to show you some games of to develop this quality in yourself

mine in which a fierce fight erupted by analysing the early games of fa-
literally in the first few moves - in mous players who made their mark
the opening itself, or the early stage at a young age - such as Boris
of the middlegame. The struggle to Spassky, Mikhail Tai and Alexei
seize the initiative was conducted Shirov. Their ideas, conceived over
with sharp and for the most part the board rather than in quiet pre-
highly unconventional means. game study, have an air of naivety
The games were all played and directness. Sometimes they fail
years ago, when I was taking my to stand up against dry mathemati-
first steps in big-time chess. This is cal scrutiny, but refuting these
no accident. The characteristics of ideas proved so complicated that
youth are freedom from routine, their opponents missed their way.
optimism, and faith in our own Allow me to make a comparison:
powers (a faith which is some- this light, improvising manner has
times excessive and comes from the same place in chess that jazz
underestimating the opponent). has in music. And jazz, of course,
Our thoughts take flight unbur- continues to be played today.
dened by experience and knowl-
edge; no dogma stifles our inward Dolmatov - Lerner
freedom; interesting, striking games National Qualifying Tournament,
are often the result. With the pass- Daugavpils 1978
ing of years, alas, this 'nonch- Philidor's Defence
alance' wanes.
A capacity for original thought 1 e4 eS
is of paramount importance for 2 lt::if3 d6
beating a strong opponent. This is 3 d4 exd4
something difficult to learn and 4 lt::ixd4 lt::if6
probably impossible to teach. Try 5 lt::ic3 Jl.e7
 Chess in the Style of Jazz 77

I suspect my theoretical knowl- with the simple retreat 12 .i.f2. Af-

edge ran out at this point, but that ter 12...ltJe6 13 'ti'd2 White would
scarcely bothered me. After all, in bring the queen's rook to dl and
this position common-sense moves only then start to think about fur-
are not hard to make. Just stick to ther plans - whether to break in the
sound principles - develop your centre with e4-e5 or prepare a
pieces, fight for the centre - and kingside pawn advance with h2-h3
you have nothing to fear. andg2-g4.
6 .i.e2 0-0 My opponent wasn't in the
7 0-0 l:Ce8 mood for patient defence; he de-
7 ...lDc6!?. cided to give battle in the centre at
8 f4 .us once.
9 .i.f3 11 ... d5?!
White has obtained a powerful 12 eS cS?
centre. Today I still cannot fault It was better to settle for the
my opening strategy. modest retreat 12...lDd7.
9 .... l2Ja6 Do you believe Black's action
10 :tel c6 can succeed, with his bishop on c8
If 10... l2Jc5, then 11 l2Jb3 l2Jxb3 and his knight on a6? You don't?
12 axb3 is not bad. Then the refutation must be found!
11 Ji.e3 (D) 13 exf6 l::txe3
The 'point' of Black's play!
14 l::txe3 cxd4 (DJ

I played the rook to el so that on w

ll ...ltJc5 I could defend my pawn
 78 Chess in the Style of Jazz

At the moment I am the ex-

change up, but two of my pieces
are attacked. If either of them is
taken. the material advantage passes
to Black.
Konstantin Lerner was expect-
ing only 15 'ibd4 .i.c5; he hoped
to exploit the pin on the gl-a7 di-
agonal. (Actually, after 16 'iVd2, it
is not clear that he can.) He obvi-
ously underestimated my reply.
15 .bd5! B
Now both pieces are immune
due to 16 ilxf7+. Also 15 .. .'ffxf6 even if there is no mate, White can
16 tl)e4 is bad for Black. To protect regain his material by capturing
his queen he needs to develop his the b-pawn with his bishop.
queen's bishop - but where? You 19 ... .i.xe4
can see at a glance that any move If 19...fxe5 20 'ii'xe5+ <it>h6,
then
with this bishop has its drawbacks: 21 tZ:lf6wins on the spot.
15... .i.d7 16 'Wh5 g6 17 .i.xf7+, 20 'Wxe4 ~
15....i.e6 16 .l:C.xe6 or 15....i.g4 16 My opponent was relying on
'ifxg4 dxe3 17 .i.xb7. this zwischenzug. Instead, 20 ...fxeS
15 ... Ji..fS 21 'fixe5+ 'fif6 (21. ..~g6 22
16 .l:C.e5 .i.g6 1We6+)22 'iVxf6+ ~xf6 23 .i.xb7
I have removed one piece from is hopeless for Black.
attack, but extricating the other is 21 'iVf3 fxe5
more complicated. 17 tZ:le4is pos- 22 1Wg4+
sible, when 17 ...gxf6 fails to 18 Black has won a piece after all,
.:CS, but I chose another route. but his king can't escape a mating
17 fxg7 <hg7 attack.
After l 7 ....i.xg7 18 'ifxd4, eve- 22 ... ~h6
rything is suddenly clarified. 23 .l:C.el!
18 tZ:le4! f6 All White's pieces must take
19 'ifxd4! (D) part in the assault! He threatens
A picturesque position! The both 24 lb:e5 and 24 .l:C.e3.
centre is wholly occupied by white 23 tZ:ld7 •
pieces. If the rook is taken, then 24 ~hl!
 Chess in the Style of Jazz 79

In such cases you need to calcu- 26 J...xb7 1-0

late the variations right to the end.
To make this task easier, I recom- The next example, like the fore-
mend that you start with moves going one, is a specimen of what
against which your opponent has a you might call the lightweight
single forced reply. For instance, genre - it features a quick overrun-
the queen check on e6 looks invit- ning of the black position. Inciden-
ing, but you would have to exam- tally, it shouldn't surprise you that
ine not only 24 .. .citg7 but also I am not showing any games I lost.
24 ... l2:)f6 and 24 ... °iVf6; in some Of course you must carefully study
lines you might miss the retort your losses to detect the reasons
...'iVb6+.The king move, renewing for your errors. But right now, why
the threat of l:te3, gives the enemy should I spoil my pleasure by re-
no choice. calling my own failures?
24 ... .i.cS (D)
24 ...exf4 25 'iVxf4+ is wholly Dolmatov - Franzoni
bad for Black. World Junior Championship,
Graz 1978
Sicilian Defence

1 e4 c5
2 l2:if3 e6
3 d4 cxd4
4 l2:)xd4 l2:)f6
5 l2:)c3 l2:)c6
6 .i.e2
This is rarely played (the usual
continuations are 6 l2:)db5 and 6
l2:ixc6 bxc6 7 e5). While offering
to transpose into the Scheveningen
(6 ... d6), White allows a bishop
25 .:r.dl! sortie to b4, after which he will
Again White threatens to bring have to sacrifice a pawn. I had ana-
his rook' to h3, but now there is the lysed this sharp variation with my
additional possibility of taking on coach Mark Dvoretsky, and then
b7 with tempo. used it a couple of times with suc-
25 ... lot's cess. I don't know why no one
 80 Chess in the Style of Jazz

plays this way today; in my opin- there was a way - you can try to
ion the line gives White a highly find it yourself. I later used my im-
promising position. provement successfully against
6 .itb4 Sergei Gorelov, but unfortunately
7 0-0 hc3 I haven't kept the score of that
8 bxc3 lbxe4 game.
9 .i.d3 (DJ 10 cxd4
It was worth considering 10
.i.xe4!?, but I was hoping to trans-
pose back into my analysis after
10... dS 11 .i.a3.
10 ... lbr6
White is a pawn down. As yet he
has no attack, but he possesses two
bishops and a certain advantage in
space and development. In addi-
tion, as I recall, I had a healthy op-
timism and confidence in my
powers-which is of considerable
B importance in this kind of situa-
tion. Incidentally, I would still en-
9 ... ibxd4 joy playing such a position today.
Here my knowledge ended. I White's initiative is of an enduring
knew that 9 ...lbxc3?! was danger- nature and not easy to neutralize.
ous for Black due to 10 'ii'g4 or 10 11 .i.gS 'ifaS!? (D)
'iiel, and had only analysed 9 ...d5. 12 f4!
I was acquainted with only one An unorthodox decision (it isn't
game in which that move had been usual to put a pawn on f 4 with your
played. This was Geller-Khasin, bishop on g5), but evidently the
USSR Championship, Riga 1958, correct one. White shouldn't be in
which went 10 .i.a3 'ii'a5 11 1Wc1 a hurry to exchange on f6. By ad-
lbxd4 12 cxd4 .i.d7 13 libl .i.c6 vancing his f-pawn, he brings his
14 .i.b4 'flc7 15 'ifa3 a5 16 .i.xe4 king's rook into the attack. In an-
dxe4 17 c4 f6 18 A.d6 'ifd7, with swer to 12... 'ifb4 I would have
about equal chances. I don't re- given up a second pawn with 13 f5.
member exactly how I was going 12 b6
to improve on White's play, but 13 Jlxf6
 Chess in the Style of Jazz 81

yet imperceptibly his game dete-

riorates to the point of hopeless-
ness. Why this happens, where his
play could have been improved - I
don't know myself!
19 :t'2 l:tg5 (DJ
How is White to continue the at-
tack?

w
But now it is the right time to
exchange. This enables White to
gain a tempo with 14 'ii'f3.
13 ... gxf6
14 'iif3 l:tb8
After14 ...'i!Vd515 'ifxd5 exd.5 16
l:.ael + 'iifi>f817 l:.f3, Black would
have the worse endgame. He isn't w
yet ready to fight such a patently
rearguard action. The pressure on g2 is fettering
15 fS .i.b7 my pieces. I need to deflect the en-
16 Ae4! emy queen from the long diagonal,
It is important to dash the oppo- even if only for a moment.
nent's hopes based on counter- 20 c4! 'iixc4
pressure against g2. With heavy 21 fxe6 dxe6
pieces on the board, Black's posi- 21...'ir'xe6 was a more stubborn
tion is difficult - his king is under defence.
attack and his rooks are disunited. 22 'ii'f4!
16 Axe4 A double attack on f6 and b8.
17 'iixe4 'iidS But couldn't it have been carried
1s 1i'h4 .:.gs out without the diversionary pawn
To me, this game contains a sacrifice?
kind of riddle. It looks as if Black 22 l:tb7
k~ps making good, logical moves, 23 l:tcl!
 82 Chess in the Style of Jazz

This is the point! All my pieces which is not at all a bad opening
are now in the attack. White's move? I can't understand it myself.
threats are irresistible. 7 dxc4
23 'iWdS 8 .i.xc4 0-0
24 'ifxf6 %te7 9 0-0 :lxc3?!
25 'ifb8+ 1-0 10 bxc3 'ilc7 (D)

Dolmatov - Flesch
Bucharest 1981
Caro-Kann Defence

1 e4 c6
2 d4 dS
3 exdS cxdS
4 c4
Against the Caro-Kann I always
play the Panov Attack, and quite
successfully too - it has brought
me a large number of wins to date. w
4 lbf6
5 lbc3 e6 At the end of the 1970s, Worlcl
6 lbt'3 .i.b4 Champion Anatoly Karpov suc-
7 .i.d3 cessfully employed a plan based
The encounter Dolmatov-Khari- on exchanging his dark-squared
tonov, World Junior Champion- bishop followed by developing his
ship qualifying tournament, Sochi knight on d7 and bishop on b7.
1978 went 7 cxd5 exd5 !. At that Janos Flesch is aiming for a similar
time the theory of 6 ... .i.b4 was in set-up, but goes about it inaccu-
its infancy. and Black's recapture rately: the premature exchange on
on d5 with his pawn took me by c3 increases White's options. The
surprise. I went on to win a good plan has to be initiated with either
game. but acquired nothing from 9 ...b6 or 9 ...lbbd7. The latter move,
the opening. Since then I have incidentally, was played against me
started playing 7 Ad3, transposing by Jonathan Speelman in a game I
to a variation of the Nimzo-Indian. demonstrated at the second session
The resulting positions suit me very of our school (see the book Open-
well. Why, tben. do I avoid l d4, ing Preparation).
 Chess in the Style of Jazz 83

11 ~d3! same weakness later, with the ma-

A natural and logical move, noeuvre ..slg5-h4-g3 !
withdrawing the bishop from at-
tack. I confess that at the time I
didn't even look at the reply
11...'ibc3. (Today I am not quite
such an optimist - I'm sure I would
examine it.) After 12 .i.f4! (but not
12 ~gS ~bd7, transposing to the
Speelman game), White has a huge
lead in development, ensuring him
more than sufficient compensation
for the sacrificed pawn.
11 ... ~bd7
Black hopes to arrange his B
pieces on the Karpov model after
12 c4 b6 13.~gS ~b7, but I don't 12 ... :es
give him the chance. The usual square for the rook in
12 .b3! (DJ this variation (this is where Kar-
An unconventional development pov used to put it). It would have
for the bishop in this opening sys- been better to move it to d8, but my
tem. In his youth a chess player has opponent didn't figure out my in-
fewer dogmas and more energy - tention.
it's easier for him to think up fresh 13 lbd2!
ideas. At a more advanced age he What is he to do now? The
knows exactly how players have knight is heading for d6, and after
played before in similar cases, and 13...'ifxc3 14 lt)c4 Black may well
this knowledge sometimes pre- lose his queen. I am sure he should
vents him from approaching the coolly play 13...b6, although after
position without limiting precon- 14 ~c4 ~b7 15 ~d6 White has an
ceptions. obvious plus.
Here I succeeded in pinpointing 13 ... J:.d8 •
the main weakness in the oppo- 14 'ii'f3
nent's camp - the vulnerable d6- On 14 ~c4 ~f8, the d6-square
square. Actually, even after the would be covered. Therefore I acti-
nonpal bishop development on g5, vate my queen, guarding the c-
White often tries to exploit this pawn at last, and stopping the
 84 Chess in the Style of Jazz

black bishop from developing to In particular, I wanted to make

b7. development difficult for the black
14 ... lbrs(DJ bishop. However, knowledge of
Now what would you play for general laws such as the one just
White? mentioned tends to make the deci-
sion easier - it gives us an idea of
what to look for.
15 ... lt)xe4
16 ~xe4 ltld7
A sorry spectacle - the black
pieces scurry back and forth. He
evidently wants to play 17 ... ©f6,
but of course I don't allow it.
17 iLe7! l:te8
18 Jlh4 f5??
Anyone would surely have
played this - it's hard to endure
w such powerful pressure for long.
Nevertheless it would have been
15 lbe4! better to show patience and refrain
A typical stratagem! In such from weakening the position.
cases it's useful to exchange some 19 ~c2 lcib6
of the opponent's developed pieces 20 ~b3 .i.d7
- then your own lead in develop- 21 ~g3
ment can be utilized more easily. I Before taking the f5-pawn, it
suggest an analogy with hockey: if helps to improve the position of
one player is sent off, the advan- the bishop. It is amusing that this
tage of five against four is palpable piece has, after all, made its way
but not decisive. Remove another round to g3, the customary square
pair (one from each side), and de- for it in this variation.
fending becomes much more diffi- 21 'ir'c6
cult with three against four. With 22 'ii'xfS 'ir'xc3
two against three it is all but im- 23 .i.e5!
possible. White has protected his d4-
Of course, when White played pawn and parried the threat of
this move there were also consid- 23 ... 'Wxb3, which would now be
erations of a more concrete kind. decisively answered by 24 'ii'g5.
 Chess in the Style of Jazz 85

23 ... 30 l:[(3 lbxf6

The queen returns to the defence. He has to give up his queen,
24 1i'hs :Sc8(D) which is tantamount to resigna-
How should the attack be con- tion.
tinued? 31 l:tg3 rl;;g7
32 :n :.rs
33 :Xg6+ h:xg6
34 'f!IgS l:r.c6
35 'ii'e5! l:tb6
36 g4 l::tbS
37 dS! 1-0
Take note: having achieved a
material plus, White didn't relax
the pressure but looked for the
most direct and energetic path to
victory. In such situations a player
sometimes feels that his work is
w essentially done, he eases off and
starts to play carelessly. Conse-
25 f4! quently his opponent succeeds in
I usually find it hard to decide organizing a defence or even ob-
on a change in the structure of the tains counter-chances.
position - I prefer piece-play. How-
ever, this pawn move is very The following game ( one year
strong. The threat is not only to earlier) was played in a similar
bring the rook into action via f3, vein: the same opening, the same
but also to play f4-f5. energetic exploitation of the oppo-
25 ... g6?! nent's opening inaccuracies. I was
25 ....i.c6 is bad due to 26 W°g4! then an international master, tak-
"fie7 27 .i.xe6+, but after the move ing part in a strong grandmaster
played White can still demolish his event for the first time. Bent Larsen
opponent's defence by force. The was one of the favourites. Experi-
most tenacious move was 25 ...lb<:4. enced grandmasters are prone to
26 'ii'h4 'f!le7 underestimating young and ambi-
27 .i.f6 'fllf7 tious opponents. We had already
28 rs ~d5 crossed swords in the first round,
29 fxg6 W'xg6 when I managed to crush him
86 Chess in the Style of Jazz

quickly with Black. As this was a it makes sense to prevent ...b7-b5

double-round tournament. our next by playing 10 a4!?, along the lines
encounter soon came - again with of the Queen's Gambit Accepted,
a favourable result for me. but at the time I didn't feel like
weakening b4, and thought up a
Dolmatov - Larsen different idea.
Amsterdam 1980 10 a3!?
Caro-Kann Defence This too is prophylaxis against
...bS, but of a more sophisticated
1 e4 c6 kind. In reply to 10 ....i.e7 I planned
Of course, Larsen didn't suspect to withdraw my bishop to a2 first,
bow dangerous it was to play this and then answer the flank thrust
opening against me. 11...bS with a counter-blow in the
2 d4 d5 centre: 12 d5 !. On the other hand if
3 exdS cxdS 10....i.xc3 11 bxc3 bS, then after
4 c4 ~f6 12 .i.d3 the threat of 13 a4 is un-
s ~ e(i pleasant.
6 ~f3 .i.b4 All the same, this last variation
7 .i.d3 dx:c4 looks like the most logical reaction
8 .i.xc4 0-0 to White's idea. In lines where
9 0-0 a6 (D) Black swaps on c3, White's a2-a3 is
a wasted tempo: the pawn should
either stay on a2 or be advanced to
a4. At the 1982 Zonal Tournament
in Erevan, Lev Psakhis had done
some good preparation for his
game with me, and headed for the
position in question. There fol-
lowed: 12...WdS 13 a4 .i.b7 14
'ii'e2 l:.c8 15 axb5 axb5 16 l:.xa8
.i.xa8 17 .i.d2 ~e4 18 .i.;xe4 'ifxe4
19 'ii'xb5 .i.dS 20 l:.el 'ii'g6 21
We2 ll)c6, and Black had sufficient
w compensation for the pawn sacri-
ficed. The game was soon drawn.
I didn't know anything about Of course. a clear-headed ex-
this move. Later I discovered that amination of the position is much
 Chess in the Style of Jazz 87

more difficult at the board, with Black would have to swap queens,
the clock ticking away, than in as 14 ...l:ta.7is too risky: 15 .i.f4 (15
home analysis. So if you manage 'ii'h5!?) 15 ...l:td7?! 16 'ii'h5, with
to think up a sensible idea like 1O strong pressure on Black's king-
a3, its practical chances of success side. After 14...'ii'xd.5 15 .i.xd.5 .=.a1
are very considerable - even if 16 Ji.f4, Black can't play 16...Ji.b7?
there is a solution to the problem 17 Ji.e3, but 16 ...l:td7 17 Ji.xf7+
facing your opponent. l:txf7 18 Ji.xb8 is also bad for him.
10 ... Ji.e7 He can only settle for a permanently
11 Ji.a2 b5?! (DJ inferior ending with 16... Ji.e6 17
11... l2k6 is better. .i.xe6 f xe6. That was the least of
the evils, though. Objectively, Black
would be justified in counting on a
draw. "But why should I deal with
a little boy so timidly?" the grand-
master must have thought.
14 tl)xe7+ "flxe7
15 Ji.g5
The two bishops in an open po-
sition guarantee White an over-
whelming plus. All I need to do
now is play natural attacking moves
and make sure my opponent does
not manage to jump out of the trap
he has landed in.
12 d5! 15 ... tl)bd7
What is Black to do now? He 16 l:tel 'ii'cS
doesn't want to allow the capture 16 ... 'i~'dS would have offered
on e6 - he would have to spend the more resistance.
whole game defending a clearly 17 Ji.e3
inferior position. Not, of course, 17 l:tcl? .i.xf.3.
12 ... exd5 11 ... 'it'r5(DJ
13 l2Jxd5 Ji.b7? 17 ... 'ilh5 was relatively better,
Larsen had to exchange with although after 18 tl)g5 'ii'xdl 19
13 ...~xd5. He was afraid of the l:taxdl Black has a difficult end-
answer 14 'ii'xd5 (stronger than 14 ing.
Ji.xd5 l:ta7, threatening 15...l:td7). 18 ~4!
88 Chess in the Style of Jazz

w
Black's queen is almost trapped. The bishop has come across to
Of course, the routine 18 ttld4? is guard f7. Of course White has vari-
weaker on account of the reply ous ways of winning, but I recom~
'1s ...'Wg6. mend that you always look for the
18 'iVe4 kind of solution that I chose in this
19 Jigs 'iVc6 game.
20 :Cl 'iVb6 24 'ild4!
21 .ile3 Total domination! There is no
The game takes a highly amus- hurry to pluck the apple - it will
ing tum. My dark-squared bishop fall of its own accord. First deprive
moves back and forth, gaining the opponent of all reasonable
tempi all the time. moves, then finish him off. I was
21 ... 'iVd8 even sorry to play the remaining
22 ttlfS few moves; I just wanted to gloat
Since move sixteen, as you can over the ideal placing of the white
see, only White has been playing. pieces, since I could no longer do
My opponent has roamed all round anything to improve it.
the board with his queen and fi- 24 ... 'ifb8
nally brought it back to its starting It was not for nothing that I had
square. In the meantime I have trained myself in 'prophylactic
brought all my pieces into the at- thinking•. I understood at once that
tack. Black was planning 25 ....:ds. The
22 Jle4 winning variation had to be calcu-
23 t2Jd6 / .ilg6 (DJ lated to the end (which is very easy
 Chess in the Style of Jazz 89

when the opponent's possibilities 6 .i.e2 .i.d6

are so limited). It was now time to 6 ... c6 may be a little more pre-
pick up my point and go home. cise, so as to recapture on d5 with
25 f4! ltd8 the c-pawn.
26 rs .i.hs 7 cxd5!? exd5
27 h3 ©b6 8 ©d4
28 'ifxb6 'ifxb6 A typical device in Reti's Open-
29 .i.xb6 lh.d6 ing - the knight heads for f5.
30 .i.e3 8 ... .i.xe2
It is usually after this kind of 9 'ifxe2 (D)
precise move that your opponent What would you play now?
capitulates. (If the bishop had gone
anywhere else, Larsen would have
started thinking about 30 ...l::td2.)
1-0

No doubt you have gained the

impression that I only win with
White. Then let me show you a
game in which I had Black.

Van der Sterren - Dolmatov

Amsterdam 1979
Reti Opening

1 ©f3 dS Of course Black could simply

2 b3 .i.g4 castle (9 ... 0-0), but after 10 ©f5
3 Ji.b2 ©d7 White would be a little better. I
3 ....i.xf3 would lead to a wholly didn't want to concede the initia-
unexplored situation - the kind I tive to my opponent, and chose the
try to avoid. Black's plan, which most thematic continuation - al-
had brought me success a few though a slightly risky one.
times already, is simple: ...e6, ...c6, 9 ... .i.eS!
... ©gf6, ... .i.d6, ... 0-0, ...l::te8and 10 f4
at some point ...e5. I also had to take 10 .i.a3 into
4 c4 e6 account. If then 10...c5, White has
5 e3 ©gf6 11 f4! .i.c7 12 ©f5. I planned
90 Chess in the Style of Jazz

10...loe4!, with 11 f4 .i..f6 or 11

'ifb5 .i..xd4 12 exd4 'ii'g5 to follow.
10 ... hd4
Once again I may remark nos-
talgically that young players are
apt to play without prejudices. You
might feel reluctant to leave your
opponent with a strong bishop on
the long diagonal. A mature player
might not have risked this ex-
change (and would therefore have
refrained from 9 ....i..e5). In actual B
fact, Black's sturdy position in the
centre and the slight vulnerability First sec if you can solve your
of White's set-up (weakened by problem (the d5-pawn is en prise)
f2-f4) ensure me adequate coun- with some normal, useful move.
terplay. ' 15 ... l:tad8!
11 .i.xd4 All my pieces are now in play.
After the anti-positional move After 16 lZ'lxd5 lZ'lxd5 17 1Wxd5
11 exd4+? Black would gain the ltJf6 Black wins his pawn back
advantage by l I...1Ve7 12 1i'xe7+ with the better game.
<i;xe7 13 .i..a3+ ~d8 14 0-0 .l:.e8 16 l:tabl
15 loc3 lob8 !, to be followed by Van der Sterren has decided to
16...loc6. guard against any tries with ...c4.
11 cS However. it is not only my c-pawn
12 .i..b2 0-0 but also my d-pawn that is ready to
13 0-0 l:te8 advance. You can see how useful it
14 'ifd3?! is to have a number of strategic
It was better to place the queen threats in reserve at the same time.
onf3. while not hurrying to put any of
14 ... 'ii'b6 them into action! .
White will soon have to worry 16 ... ._.c6
about either ...d4 or ...c4. 17 l:f3 d4!
15 loc3 (D) The moment has come! With his
Now what would you play? last move, Van der Sterren revealed
It's no good trying to come up his wish to start a flank attack; I re-
with a brilliant'idea at this stage. spond with a central counterstroke,
 Chess in the Style of Jazz 91

which, by the way, required pre- defend the d-pawn indirectly: 21

cise calculation. .i.xd4 is strongly answered by
There was also another tempt- 21.. .J:r.xd4! 22 l2:)xd4l2:)e623 .l:td3
ing possibility: 17 ... lbe4!?. White .l:td8. At the same time I avoid the
can't reply 18 lbxd5? because of exchange of queens which might
18 ...'iVd6 with a lethal pin on the otherwise have occurred after 21
d-file, but after 18 lbxe4! dxe4 19 .l:tcl.
1i'c3 f6 20 .l:tg3 he would maintain 21 :Cl l2:)e6.
a playable position. I wanted more. 22 d3
18 exd4 cxd4 On 22 f5, Black has the very
19 lbbS strong 22 ... lbgS 23 .l:td3 a6 ! and
Not 19 lbe2 l2:)c520 'ii'c4 'ii'e4 then 24 lbxd4 lbg4! or 24 lbc7
21 .l:te1 d3 22 'ii'xc5 1Wxe2!. .:te4. The move White plays weak-
19 ... l2:)cS ens the e3-square; a black knight
20 'ifc4 (DJ immediately heads for it.
22 ... l2:)g4
23 fS! (DJ
Now what would you play?

My pieces are excellently placed,

but the d4-pawn is under attack. In
, this sharp position Black had al-
ready foreseen a purely positional 23 ... l2:)f8!
solution. • A sober response. In answer to
20 .•. 1!t'b6! 23 ...tbe5? White had prepared to
By, placing my queen on the give up his queen with 24 fxe6!
same diagonal as the white king, I l2:)xc4 25 exn+ ~h8 26 fxeS'ii'+
 92 Chess in the Style of Jazz

and wins. Nor is 23 ... lng5?! con-

vincing: 24 l%g3~e3 25 'ii'c5 'ii'xc5
26 l::txc5'iJdl 27 h4!. But why
plunge into unnecessary complica-
tions when your opponent's posi-
tion is sufficiently compromised
already? The threats of 24 ...~e5
clild 24 ... ~e3 are very dangerous.
Black only needs to make sure that
capturing the d4-pawn will not get
White out of serious trouble.
24 ..bd4 'ikh6! B
White's h2-pawn is attacked. In
addition his back rank is weak; the 30 ... l':td6!
position of his rook on c 1 is shaky. 31 lhb5 b6
On 25 l%h3,Black would not play Black has given up a third pawn,
25 ...lid2?! 261i'c3, but25 ..J:txd4!! but now his b-pawn can be de-
26 ~xd4 'ii°d227 l:tfl J:el 28 'ifc2 fended from d7 by his knight.
(28 lU3? J:.xfl + 29 lhfl 'fi'e3+) 32 lte4 l':ted8
28 ...lhfl + 29 'it>xfl 'ii'f4+. 33 .:tees ~g7
25 h3 a6! 33 ...ltJd7 can be answered by 34
The overloading of White's lled5. There is no need to hurry
pieces makes itself felt. ff 26 hxg4, with the knight move; the best
then 26 ... axb5 271i'c3 J:xd4. Nor thing just now is to ~ring the king
is 26 .t.b2 any help: 26 ... itJe5 ! 27 nearer the centre. In the endgame
he5 axb5 28 it'f4 lhe5 !. you should utilize any breathing-
26 .bg7! 1Wxg7 space to strengthen your position
27 :g3 axb5 to the maximum.
28 li'xg4 'ifxg4 34 <ili>f2 ~f6
29 lbg4+ 'ii7h8 35 J:te3 lbd7
30 J:.c5 (D) 36 g4 l:tc8
White has only two pawns for 37 a4 %tc5
his piece. Nonetheless Black still 38 J:.b4 l:tcdS
has some difficulties in realizing 39 'itie2 ltJeS?! (D)
his advantage. His pawns are all So far my play has been logical
shattered and his knight is out of enough, but at this point I weaken
play. and start to play carelessly. There
 Chess in the Style of Jazz 93

was an easy win with 39 ... ~c5 40 Another inaccuracy! The cor-
d4 .l:.xd4. rect 50 ....l:.ld4! would have forced
the rook exchange in more favour-
able conditions - on the fourth
rank.
51 .l:.b5 l:td5
52 l:txd5 .:Xd5
53 b4 l:td4!
54 l:te4 l:td1
Of course Black must retain one
pair of rooks. After an exchange
on e4 the two pawns would be no
weaker than the knight.
55 h4 · Wf6
s6 «iti>r4 no+
57 <itig3 ~e5
40 d4'- ~c6?! 58 l[d4 l:gl +
40 ...:Xd4 was stronger. 59 «ili>h3 @e7
41 .l:.c4 ~xd4+ Black has achieved a great deal;
42 ~f2 ~c6 he has stopped the white rook from
43 ~g3 ~e5 penetrating, activated all his pieces
44 :ce4 :c6 and cooped up the enemy king on
45 ~h4 ~d7 h3. The threat is 60 ...l:tal.
46 @h5 cj;g7 60 as bxa5
The impression is that Black has 61 bxa5 l:al
dithered a little, and made the win 62 <itig3 .:Xas
more complicated. The important 63 g5 (D)
thing in such cases is to keep calm 63 h5!
and try to regain the thread of the When exploiting a material ad-
game, without being in a rush to vantage you should avoid exchang-
force events. ing pawns. The fewer pawns on the
47 ~h4 l:tcl board, the greater the drawing
48 l:tb4 h6 chances!
49 ~g3 l:cdl 64 @f4 'itrf8
I now understood that I had to Again Black allows the game to
exchapge a pair of rooks. drag on. Bearing in mind the prin-
50 @a :Sd4?! ciple I have just stated, I didn't
94 Chess in the Style of Jazz

73 l%a7 l:b3+
74 Wf4 l:b4+
75 'itg3 lt:Jes
76 :as l:g4+
77 «.t>f2 :.C4+
78 <ite3 lt:Jc4+
0-1

Dolmatov - Lerner
Tashkent 1983
Sicilian Defence

1 e4 c5
want to play 64 ...f6. and yet this 2 lt:Jf3 d6
would have led to a quick win: 65 3 d4 cxd4
gxf6+ <it>xf666 l:d6+ <it>g7 67 l:e6 4 lt:Jxd4 lfil6
lt:Jf7 68 l:g6+ <it>f869 f6 li::Jd8!. 5 lt:Jc3 loc:6
65 lld6 q;g7 6 .i.gS e6
66 l:b6 7 'iid2 a6
On 66 l:h6, I intended to im- 8 0-0-0 h6
prison the rook with 66 ...lt:Jg6+! 9 jle3 (D)
67 fxg6 fxg6, after which the game Today I usually choose the more
is decided by zugzwang: 68 <it>f3 cautious 9 jlf4.
t[a4 69 <it>g3l:r.b470 <it>h3J:r.g4.
66 -· lt:Jg4
67 J:r.b4 l:a3
68 J:r.c4?! :b3 ?!
We both missed 68 ...l:r.f3+! 69
<it>xf3(69 We4 :Xf5!) 69 ...lt:Je5+.
69 l%a4 00! •
Black threatens 70 ...:b4+J.
70 l%a7 :b4+
71 ~ lt:Jg4
72 l:e7
Or 72 g6 lt:Je5+.
72 ... l:bl!
·Now the threat-is 73 ...:n+.
 Chess in the Style of Jazz 95

9 ile1 same defence would not be avail-

10 f4 lcixd4 able owing to 17 ll:ix/6+ 1lixf6 18
11 Jlxd4 1lib6 - Dvoretsky.]
In our day this system is very 14 0-0
popular, but at that time it was still 15 .i.xf6 .i.xf6
in its infancy. The next phase of the 16 eS .i.e7 (D)
game, then, is a case of improvisa- After 16... dxe5?! Black would
tion at the board - by both sides. I not obtain enough for the queen,
should point out that today White but 16 ....i.h4!? deserved serious
more often places his pawn on f3, consideration. There could follow ·
rather than f4. 17 'ii'e3 b4 18 lcie4 dxe5 19 fxe5
11 bS 'ii'a5 20 lcid6 iLe7, though after 21
12 Jid3 .i.b7 ile4 White's position would still
13 ~bl .i.c6?! be preferable.
Black is preparing ... b4, and first
prevents the knight from . side-
stepping to.. a4. The immediate
13... b4? would be met by 14 lcia4
(after 14...lcixe4 15 'ii'xb4, not only
the bishop but also the g7-pawn is
attacked). Nonetheless the move
Black plays has a serious draw-
back, which is underlined by my re-
ply. It was better simply to castle.
14 1Wel!?
Now 14...b4 can be answered by
15 .i.xf6 .i.xf6 16 lcid5. Further-
more there is e4-e5 hanging over
Black's head, utilizing the opposi- 17 fS!
tion of the white rook and black The standard method of attack
queen on the d-file. when Black has not been able to
[As Grandmaster Kindermann exchange pawns on e5. Of course,
has indicated, after 14 ... M 15 .i.x/6 the move required exact calcula-
iLx[6 16 lcid5 a5 Black would tion.
maintain a defensible position. How does White continue the
The-r:eforeinstead of 14 1lie 1! ? he offensive if Black now takes the e-
reco'mmends 14 1lie3!, when the pawn? I hardly looked at 17...dxe5
 96 Chess in the Style of Jau.

18 f6 ii.xf6 19 3i.h7+. Black's rook, approaches the key point f5 with

bishop and two pawns are stronger tempo.
than the queen. The correct move 19 ... 'i!ic7
is 18 fxe6!. Then giving up the After 19...ii.e4 20 ii.xe4 fxe4 21
queen (with 18...fxe6 19 ii.h7+) tl)f5, the pin on the d-file is deci-
would be unfavourable, but other- sive: 21...d5? is refuted by 22
wise Black would face serious dif- lhd5. If 19 ...3i.d7, then 20 ..i.xf5 is
ficulties. strong.
17 ... b4 20 il:}xfS dxeS
This is the move Lerner was 21 'i!ig3 g6
counting on. What now? 18 tl)e4 21 ...ii.g5 22 h4 ..i.f6 23 ltixh6+
exf5 is not good for White. I con- Wh8 is a more resilient defence.
fess I am proud of my next move. 22 tl)xb6+ 'it>h8(D)
When conducting, a sharp fight
you need a high degree of alertness
and ingenuity to exploit all your re-
sources. If at some point your play
is not energetic enough, your at-
tack may come to a dead end and
your opponent will seize the initia-
tive. In the present case, Black has
the bishop-pair; all he needs to do
is beat off the immediate threats
without suffering too much dam-
age ...
18 tl)e2!
An unpleasant surprise. The
fearsome f5-f6! is threatened, for Black's king is vulnerable; I have
example: 18...'ifa5? 19 f6! gxf6 20 excellent attacking prospects. How
11i'g3+ 'it>h821 exd6, or 18 ...ii.d5? do I keep up the pressure? Don't
19 f6! gxf6 20 'i!ig3+ 'iti>h821 'iif4 imagine that a stroke of genius is
'i;g7 22 ltig3. If 18...dxe5, then 19 called for. Sometimes you do need
fxe6 is strong, as before. to look for difficult, concealed so-
18 ... exfS lutions, but more often you simply
19 tbd4 have to play logical, precise moves
Here is a consequence of Black one after the other, without making
having his bishop on c6! The knight any mistakes.
 Chess in the Style of Jazz 97

23 lihfi everything. Let me explain how this

The hitherto inactive rook joins happened. In one sense, I was get-
in the attack. The position is not ting carried away - I thought I must
yet ripe for combinations like 23 be close to a beautiful finish to a
l£ixt7+. game where everything had gone
23 ... JidS so well. On the other hand I didn't
Black had to protect his f-pawn. quite feel confident enough, and was
How should White continue now? on the lookout for some conven-
A good idea that suggests itself ient way to go into a safe endgame
is l£if5 !. But if we are playing it, let with an extra pawn. I fell between
us do so with tempo! two stools. My conflicting motives
24 'ii'h3 ~g7 took their toll; they made me use
25 l£if5+! gxfS up an inordinate amount of time,
If 25 ... ~g8, White has the deci- and at the crucial moment (when I
sive 26 'ifh6 ii.f6 27 tbe3 (simul- was already in time-trouble) they
taneously attacking d5 and f6) stopped me from selecting and cal-
27 ...ii.g7 2~tbxd5. culating the right line.
26 'iVg3+ ~h6 (D) The first possibility is 27 l1xf5.
After 26 ...Wh8 27 l1xf5, mate is After 27 ...l1g8 the black king aims
inevitable. to escape to f8. To carry on the at-
tack, White must sacrifice a rook
with 28 l1h5+.
The second way is 27 ii.xf5.
Black's only reply - 27 ... 'ii'c4 -
leads to an endgame where White
has an extra pawn: 28 "iVh3+ "iVh4
29 'i!i'xh4+ii.xh4 30 l:txd5.
Finally, White can give an inter-
vening check with 27 'fih3+ Wg7,
and only then play 28 ii.xf5. This
time 28 ... 'itc4 is no good, and in
addition to 29 'ii'h7+ White is sim-
w ply threatening 29 l:.xd5. Black
has only one defence: 28 ... ii.xa2+!
Here White has three continua- 29 Wxa2 'i!i'c4+ and 30 ...'ifh4.
tions, two of which win. Unfortu- I saw all this clearly, but still
nately I chose the third, and botched couldn't make the right choice.
98 Chess in the Style of Jazz

Undoubtedly the simplest solution prevent me from making aggres-

(and the one most in keeping with sive use of my rooks. Besides, the
my style at the time) was 27 il.xf5, position of the white king is by no
heading for an ending where there means secure, especially after the
would only be a few technical dif- e5-pawn is captured.
ficulties to cope with. 31 ... l:.ad8!
Yet I was sorry to stop attacking. With opposite-coloured bishops
At the same time, I failed to calcu- you shouldn't always cling to ma-
late the rook sacrifice fully. This terial; the initiative is more impor-
was a pity. The sacrifice was a di- tant. Black is quite happy to offer
rect and beautiful way to win: 27 his e-pawn - it is only getting in his
.:Xf5 ! :g8 28 lth5+ ! ! <iiitxh5 29 way.
'ii'h3+ ~g5 (29 ... il.h4 30 'i!r'f5+ is 32 'ii'xeS+
no better) 30 'fifS+ '.t>h6 31 'ili'h7+ What else?
~g5 32 llil ! il.e6 33 h4+ ~g4 34 32 il.f6
'ii'e4+ '.t>h5 (34 ... c;t,g3 35 'ii'f3+ 33 'i!i'c7 llxdl+
'iti'h236 g4) 35 g4+! 'iti'xh4 36 'ili'hl+ 34 l:.xdl 'ii'f.2!
with a quick mate. Active defence! Black not only
I couldn't decide on the rook attacks the bishop but also threat-
sacrifice, but didn't want to swap ens 35 ...b3 !. This explains my next
queens. That is why I plumped for move.
the third possibility. 35 il.e6 il.xb2!
27 'ifh3+ <iiitg7 I had foreseen this counter-
28 il.xfS? il.xa2+! stroke, but thought (quite rightly, I
29 ~a2 am sure) that I had to allow it.
Not 29 c;t,a1?l:th8. 36 ~b2 'iff6+
29 'ifc4+ 37 ~bl 'i!i'xe6
30 c;;t,bt 'ifh4 38 ltd3 (D)
31 'iVe3 38 J:.c8?
I thought it would be hard for :es
With 38 ... ! Black could have
Black to defend, given his exposed maintained the balance. The threat
king and the presence of oppo- to exchange queens (39 ...'ii'el+ 40
site-coloured bishops, which in the <i1i>b2
'ili'e5+)would have tied White
middlegame ought to strengthen down and given him no time to de-
the attack. However, this verdict is velop an attack.
false; I didn't consider that by post- My opponent's moves just be-
ing a rook onthe d-file Black could fore the time-control prove weak,
 Chess in the Style of Jazz 99

This is where you can't help los-

ing your head. It's so easy to blun-
der, for instance with 40 ...'ii'd5? 41
~g4+ and 42 'iixc8.
40 l:tc4?
41 ~g3+ ~g4
42 ~d6
It's aJl over! With your flag dan-
gling, you only look at checks and
captures. Lerner, of course, simply
had no time to assess the conse-
quences of my quiet move.
The game was adjourned. It
and land him in a lost position didn't last long after resumption.
again. 42 :tc3
39 'ii'f4 f6?! 43 'iie7+ <it'g6
40 lle3 44 ~e8+ <it'fS
The right tactics! With the oppo- 45 'iie6+ <it'g5
nent in time-trouble, avoid forced 46 'ii'xg4+ <it'xg4
variations; play them only if they 47 lhc3 bxc3
win. Of course Lerner was expect- 48 <it'a2 <it'f4
ing me to check on g3, and would 49 <it'b3 ~e3
have moved his king instantly in re- • 50 ~c3 rs
ply. But how should he now react? 51 h4! 1-0
 6 Practical Chances in Chess
Beniamin Blumenfeld

Note: This article was first published in 1934

Purely specialized factors (posi- being at all convinced that it is

tional understanding, the ability to the best of all those available; he
calculate far ahead, etc.) are not in merely judges that it gives the
themselves sufficient for success most chances in practice.
in chess; you also have to apply
those faculties which make for
success in other forms of sporting
contest and in real life: quick-
wittedness, the habit of speedily
finding your bearings in a new
context; the ability to take
weighty decisions that have im-
mediate effects, in circumstances
that defy analysis; total dedica-
tion to a given aim, whether vic-
tory or salvation from defeat; in
a bad position, self-control and
endurance; in a good one, re-
fusal to let your achievements go Kmoch - Nimzowitsch
to your head. Niendorf 1927
In the chess press, games are
mainly discussed with a view to de- From the diagram, play pro-
termining the theoretical correct- ceeded 44 ... b4 45 ~a4 (not 45
ness of the play. Yet this kind of cxb4 ~xd4 46 bxa5+ ~b5, when
elucidation does not always faith- Black emerges with two strong
fully reflect the actual course of the passed pawns). At this point, the
struggle. An experienced chess obvious-looking continuation was
player will sometimes opt for a 45 ... ~xe5 46 1'.xd7 (no better is
particular continuation without 46 dxe5 1Lxa4 47 cxb4 .i.b3 48
 Practical Chances in Chess 101

bxa5 llxa5, which leaves Black away from the queenside and carry
with an extra pawn and winning out a prepared combination.
chances in spite of the opposite- SO Wg3 (D)
coloured bishops).
[After45 ... lbxe5 46 Ld7fud7
47 cxb4 a4 Black wins without
trouble. A more tenacious line is
46 Le5 La4 47 cxb4 ilb3 48
bxa5 l:.xa5 49 ilf4, preparing for
~h4-g5 or ile3 followed by @f4. -
Dvoretsky]
Instead of this, Nimzowitsch an-
swered 45 .i.a4 with 45 ... bJ, and
there followed 46 .i.xc6+ @xc6.
The position now looks dead drawn;
the black passed pawns on the a-
and b-files..are easily stopped, and
a kingside breakthrough is impos- 50 ... %1b4!
sible. Black aims to obtain passed
The game continued: pawns which will advance with
47 g5 l:.a7 tempo, owing to the awkward plac-
48 l:.b2 ing of the white rook on b2. We
White blockades the black pawn. now see why he didn't do the obvi-
There was danger in sticking to ous thing and push his a-pawn at
purely waiting tactics, for instance any time in the last few moves.
48 @f3 ltb7 49 ..t,gJ a4 50 Jla3 51 cxb4 a4
b2 ! 51 l:.xb2 l:tb3 ! 52 l:.xb3 cxb3 52 b5+
53 @f3 'it;>b554 c.1i>e3 b2 55 Jlxb2 White gives up a pawn to open a
'iPc4 56 ~d2 'iPb3, and Black wins path for his bishop; yet the rook
the bishop. and bishop prove helpless.
48 ... l:tb7 52 ... Wxb5 53 Jla3 c3 54 l:tbl
49 'ifi>f4 .i.c8 'it'c4 55 f 4 'it'xd4 56 @f2 'it'c4 57
Black's aim seems to be to try to 'it>el d4 58 @e2 @d5 59 @f3 Jlb7
penetrate on the h-file with his 60 :e1 'iPc4+ 61 Wf2 b2 62 f5
rook, so White's next move is exfS 63 e6 .i.c6 0-1
natural. Yet Black provoked this From the combination that
reply so as to draw the white king Black carried out,. we can see what
102 Practical Chances in Chess

dangers were lurking for White in a Instead of this simple variation

position that looked hannless. Nim- that leaves Black with no chances,
zowitsch was therefore right to White devised a combination. The
think that the continuation he chose game continued: 32 ~e7 Jlxe6 33
gave the best practical chances. ~xg8 ilxc4! (White was evidently
counting on being able to meet
33 ....ilxg8 by 34 f4).
Thanks to his combination
White has come out the exchange
up, yet Black has acquired definite
counter-chances in the shape of
mobile queenside pawns sup-
ported by the two bishops. Black
even went on to win.
The conclusion from this is
that given a sufficient advan•
tage, we should select those con-
tinuations which enable us to
achieve the win without allowing
Kmoch - Yates the opponent any counterplay.
San Remo 1930 We may also draw a further con-
clusion which is less of a platitude.
White has obtained a decisive Suppose there is the choice be-
advantage by fine play. He should tween two continuations: the
now continue simply with 32 l16xe5 first gives a decisive positional
dxe5 33 l:txe5 [33 l"i:Je7I? is sim- advantage with a balanced dis-
pler- Dvoretsky ], and if 33 ...J:.g5, tribution of material; the second
then 34 ~e7 !. In this variation gives' roughly the same amount
both players have approximately of advantage, but with a mate-
equal forces. Although doubled rial imbalance (as in our exam•
and isolated, the white queenside ple with rook and knight against
pawns are fulfilling their function two bishops). In this situation it
perfectly well - holding up Black 's pays to select the first continua-
pawns on the same wing. On the tion. With balanced material the
kingside, White can create two devices of attack and defence are
united passed pawns. The game more familiar; there is less scope
would win itself. for the unexpected.
 Practical Chances in Chess 103

victory. In such cases you often

find that the whole character of
the battle is altered; pieces that
were well positioned for con-
ducting the attack_turn out to be
on the wrong squares once the
specific goal_is attained; the play
shifts to another sector of the
board where the opponent's
forces are more numerous or bet-
ter placed. You should therefore
exercise particular caution at
critical moments when the win
Yates - Ahues of material is possible, and care-
Hamburg 1930 fully consider whether gaining a
material plus is worth a deterio-
White CQuld have decided the ration in your position.
game immediately with a fairly
simple combination: 41 lbg6+ hxg6 In the foregoing example White
(41 ...'ii'xg6 loses a piece) 42 'fih4+, overlooked a line that was immedi-
etc. Instead, probably without giv- ately decisive. Quite often, however,
ing it a thought, he played 41 lbh5, a player will deliberately reject a
which at first sight looks very strong simple winning line because he
too. wants to win 'brilliantly'.
There followed 41 ... ireS! (not The following game (see dia-
41...1i'xg3 42 3'.Xd4+ 'fie5 43 lb.e5, gram overleaf) was a particularly
etc.) 42 'itihl hc3! 43 :Xe5 he5. sorry case.
White now has queen against rook
and minor piece, but his kingside In this position Black played an
attack has evaporated and Black interesting combination:
can work up active play. White 31 ... lbgS 32 lbxb? f3 33 lbxf3
eventually suffered defeat. (the only way to avoid mate or loss
In connection with this exam- of a piece) 33 ... lbxf3+ 34 .1'.xf3
ple, we •can make the following .1'.c6! 35 .1'.xc6 l:td3 36 'ifb2
general observation. When an at- All Black had to do now was
tack culminates in material gain, gain a clearly won position with
it is too early to be celebrating the natural 36 ... lbxc6, which• is
104 Practical Chances in Chess

aim. Striving for dramatic effects

- which stems from a false under-
standing of chess beauty - often
has lamentable consequences.

If simple, clear solutions are

what you should seek in a won
position, the converse applies: in
a lost ( or considerably worse)
position you should try to stir up
complexities. In a situation where
B natural continuations condemn
you to defeat, you shouldn't
Sergeev - Grigoriev shrink from material sacrifices;
Masters' Tournament, the main thing is to obtain active
Moscow 1932 counter-chances.
In particular, it is worth noting a
what Grigoriev would surely have characteristic feature of Alekhine's
played in a blitz game. To the gen- play: in inferior positions he doesn't
eral amazement of the spectators, allow his opponent's advantage to
however, he sank into thought, and grow, but seeks to disrupt the natu-
after some reflection played the un- ral course of events; he steers the
expected 36 .. .1!i'xg4+. There fol- game into a new channel and con-
lowed 37 .i.g2 lUT3 (aiming for a jures up sharp play, if necessary by
'pretty' mate by bringing his knight sacrificing. This hallmark of Alekh-
to f 4 or h4) 38 lbcS ibfS 39 l:.e4 ine 's style remains particularly
(the refutation Black had missed) clear in my memory from the large
39 ...'t!i'gS40 lt)xd3, and White won. number of games (mostly off-hand)
In his quest for beauty Black which I played against him when
forfeited his well-earned win. This he had yet to scale the summits of
example should be a lesson to many. chess fame.
The best continuation is the one Similarly in Aron Nimzowitsch's
which leads most surely to the games you may observe that he
goal - of victory. The inward doesn't go to pieces in lost posi-
beauty of chess lies in purpose- tions and sometimes saves hope-
fulness and in choosing the most less situations by unearthing some
economicaf means to achieve the practical counter-chance.
 Practical Chances in Chess 105

The following example is char- 23 ... l:.xcJ?, and after 24 l:.afl e2

acteristic. 25 l'hf7+ ,l;'.txf726 °ifxf7+ ~h6 27
'ifif8+ he had to resign.
Some would say that Nimzo-
witsch scored an undeserved win,
but I find it impossible to agree.
Even the strongest master cannot
calculate everything to the end;
hence he will sometimes obtain in-
ferior or even lost positions, not
only against his peers but also
against very weak opponents. Keep-
ing calm and finding counter-
chances in a difficult position is a
w distinct virtue. In such circum-
stances, a mistake on your oppo-
Nimiowitsch - Euwe nent's part is highly probable:
Karlsbad 1929 having achieved a decisive plus
in a long and tiring struggle, he
White looks doomed. His posi- will have exhausted most of his
tion is undeveloped. Black has a energy and will be in a mood for
strong pawn on e3, which threat- demobilization, imagining that
ens to queen. White cannot take all his difficulties are behind him.
this pawn: 23 l:.xe3 is answered by Thus it is ipcorrect to speak of
23 ... 'i!.Vxd4,winning a piece. An- luck and chance when a player
other bad line is 23 .l:xn+ :.Xf7 24 saves a lost game. Luck favours
'i!.Vxc8e2. Yet in this gruesome the strong!
situation Nimzowitsch kept his
head and played 23 lbc3, putting Not all top-class players have
his knight en prise. their wits about them at all times,
If Euwe had reacted to his oppo- as we see from the next example.
nent's desperate stroke with due
attention, he would of course have Janowski resigned in the dia-
found the winning continuation: gram position overleaf. Tarrasch
23 .. .'li'xd4 24 l:tafl 'ir'e5+ 25 ~hl rightly remarked: "Demoralized,
l:tc7, etc. Not suspecting any danger, Jan ow ski laid down his arms too
ho~ever, he played the obvious early. Instead he should have
106 Practical Chances in Chess

Em. Lasker-, Janowski Orlov - Blumenfeld

World Championship match (7), Semi-final,
Paris 1909 Moscow Championship 1932

played his last card, by no means the game. White played the strik-
such a bad one. With 63 ...c5 64 ing move 36 lbc4.
lt::ld5?(the obvious-looking move, If Black takes the knight he is
as Black seems to have ho defence blocking the c-file for-his rook, and
against mate) 64...l:.xf3+65 c.fi,xf3 there is nothing to stop the white
.i.xe4+ 66 ~xe4, he might have pawn from queening. Of course
achieved a pretty stalemate. His Black can play 36 ...l:.c3, but then
opponent could of course have White replies 37 ltlxa5 and the win
avoided this with 64 l:tb7.Still, in is a matter of fairly simple tech-
the heat of battle the stalemate nique. Similarly, 36 ...l:a4 is hope-
might have escaped him." less after 37 ltle5 or 37 lbb6.
In any of these lines it is hard for
My own experience has re- White to go wrong- everything is
peatedly confirmed that practi- simple and clear. I t!terefore de-
cal chances can be found in the cided to let him queen, at the price
most hopeless-seeming position. of allowing Black some practical
Here is quite an interesting case. counter-chances.
--
Play continued:
--Black'sposition is hopeless; the 36 .. Jla2+ 37 ~dl .i.b4 38 c7
strong pawn on c6 should decide dxc4 39J:.b8
 Practical Chances in Chess 107

It seems to be time for Black to Black threatens to give perpet-

resign, because White will queen ual check on el and e2. White must
next move. Black has, however, sacrifice the exchange.
one concealed chance to make the 41 lhb4 axb4
win difficult for his opponent. 42 c8'fi'
39 ... .:ld2+ White has to give up his e-pawn,
There is no hope in 39 ... c3 40 as 42 e4 is met by 42 ...l:tel+, and
c8'ii' c2+ 41 'iixc2 .:.xc2 42 'i>xc2, Black draws either by perpetual
when White has a technically easy check or by the pawn advance ...b3.
win. 42 ... :Xe3
40 'iii>cl? 43 f4 (D)
After this natural reply, the win Now Black can force a draw.
is doubtful. White should have White should have gone for 43
played his king to el without being 'ili'c5 .:.xf3 44 'ii'xb4, when Black
afraid of the discovered check, for plays 44 ...h5 followed by .. JU5; in
instance: 40 'iii>el .:lxh2+ 41 '.tfl this line White does have winning
c3 42 c8'ii' c2 43 'ili'h8+ 'ifi>h644 chances, though they are difficult
l:k8 c 1'ii'+ 45 .:.xc 1 .:.h 1+ 46 '.f.,g2 to exploit.
l:[xcl 471i'f6, and the win is guar-
anteed. However, to decide on this
line, White would need to perceive
what complications arise after the
'natural' move.
40 ... l:te2! (D)

43 ....:.c3+ 44 'iitb2 .:.b3+ 45 '.tc2

.:.c3+ 46 <ati>d2 .:ld3+ 47 ~e2 l:[c3!
(not 47 ....:.xd4 48 'ili'c5) 48 '.td2
l:td3+ 49 <ati>c2 l:r.c3+50 @d2 l:td3+
w lh,.lh,
108 Practical Chances in Chess

Indeed White can achieve noth-

ing, for example: 51 @e2 lk3 52
'ir'c5 lk2+ 53 Wdl b3 54 d5 exd5
55 'ii'xd5 J:txh2, and if 56 ~xc4
then 56 ... b2, while 56 ~c 1 is met
by 56 ...%1c2+57 ~bl c3!, and it is
White who must seek salvation in
perpetual check.

The ability to realize one's ad-

vantage coolly in a won position
and keep one's presence of mind
in a lost position are common to
most experienced chess players; Rotlewi - Fahrni
yet in clearly drawn positions, a Karlsbad 1911
loss of concentration and weak-
ening of the will to win can be of time for White to continue the
observed relatively often. In his fight.
tournament book of New York Yet Rotlewi did play on. He had
1927, Alekhine reproached Spiel- spotted a practical chance based on
mann for agreeing a draw in some the very fact that Black has a pawn.
games where he had practical Subsequently, the position in
chances - however minimal - of the next diagram arose.
winning.
To illustrate how the dogged
pursuit of victory can bring the
desired result even in a drawn posi-
tion, I append the following exam-
ple.

As is well known, endings with

rook and knight against rook (with-
out pawns) are drawn. You would
not think that having a pawn could
harm Black's cause. Fahrni is a
sufficiently experienced master to B
avoid bluncfering. It looks a waste
 Practical Chances in Chess 109

Black played the natural move It would be wrong to conclude

79 ... a3, whereupon a study-like from the foregoing that we ought
finish ensued: to _play in anticipation of our oppo-
80@f7 '.th6 (if 80 ...l:.hl, then nent 's errors. A reader who care-
81 tbd5!, followed by 82 tbf6+, is fully ponders the examples will
decisive) 81 Wg8! 1-0 realize that practical chances can
The interesting point is that be created only by accurately as-
without his pawn, Black could sessing the position and taking due
save himself by playing for stale- account of its characteristic fea-
mate with ...l:t.gl. tures.
7 Is it Worth Complicating?
Vladimir Vulfson

This is a question we often have to 2 d3 d5

face, but there is no ready-made 3 .!bd2 .!bf6
formula for answering it. Every- 4 .!bgf3 .!bc6
thing depends on the concrete cir- 5 g3 dxe4
cumstances. We shall learn to cope 6 dxe4 JlcS
with this problem better if we look 7 ~g2 eS
at how other players have solved it 8 0-0 0-0
and elucidate their decisions criti- 9 '1ie2 ~e6
cally. It will be interesting to trace In this scheme of development
the influence of a player's charac- Black usually plays ...a5, ...b6 and
ter and style on his choice of moves, ...~a6. By placing his bishop on
and to detect those situations where e6 Tal practically condemns it to
owing to individual bias he fails to being e,cchanged for a white
find the objectively best path. knight.
Let us start by analysing two 10 c3 as
games by Mikhail Tai. The first In this kind of position Black
was played in his heyday, the sec- must not allow b4, which would
ond many years later (I hope you guarantee_ White the initiative on
will perceive the difference). In ex- the queenside.
amining the games, we shall train 11 .!bc4 tbd.7
ourselves to search independently 12 lbgS (D)
for answers to the difficult ques- Black must part with his bishop.
tions which will inevitably arise. If it is exchanged on e6, he can re-
capture with a piece or with the f-
Vasiukov - Tai pawn. A third option is to give up
USSR Championship, Baku 1961 the bishop for the c4-knight. Which
King's Indian Attack would you prefer?
The white knight on c4 is the
1 e4 e6 more dangerous of the two. It
 Is it Worth Complicating? 111

14 ii.d2
White's plan is clear: lladl,
with ~c 1 to follow.
14
15 l:.adl
16 "fie2
Not 16 'ifb5? 1:xd2!.
16 ... l::td6
At this point, quite a good line is
17 .i.cl (17 ~f3!?) 17 ... l::tfd8 18
.:Xd6 'ii'xd6 19 llel with 20 .i.fl to
B follow, gradually taking control of
the queenside squares. With such a
might transfer itself later via e3 to build-up, the two bishops would
d5 or f5. It is true that with a dou- ensure White a slight but lasting
bled pawn on e6 Black would be positional edge. A good textbook
defendinguhese squares and the f- example of how to handle such po-
file would be opened. Nevertheless, sitions is the game Petrosian-Sax,
after the continuation 12...'ii'e7 13 Tallinn 1979 (see the supplement
lbxe6 fxe6 Black's position looks to this chapter).
suspect. Apart from playing to 17 ii.h3 (D)
simplify with 14 .i.e3, in the hope What do you think Tai played
of subsequently exploiting the here?
weak pawns, White could also se-
riously consider 14 <j.,h1 followed
by f4. With the opening up of the
game, the power of the white
bishop-pair would tell.
12 ... hc4!
13 1i'xc4 "fie7
At this point Evgeny Vasiukov
could have returned his knight to
f3 and then played it to h4, provok-
ing ... go - after which his bishop
would obtain the excellent square
h6. However, he was reluctant to B
retreat without special prompting.
112 Is it Worth Complicating?

17 ... l:lxd2!? Black's idea is now clear. He

An unexpected combination - threatens both 21...h6 (the white
though not for Tai, who had fore- knight has nowhere to retreat) and
seen it before his previous move. 21...ihf2+ 22 ~hl .i.e3.
Here is his own comment: "Look- Without doubt, Tal's ingenious
ing at the position, you get the im- combination is highly attractive.
pression that the game must carry But is it sound? After all, even af-
on in the same quiet strategic chan- ter capturing on f2 Black will only
nel. There seems to be no place have one pawn for the exchange.
here for combinations. But that is Tai gives the variation 21 ~g2
false. When I started thinking about 'ii'xf2+ 22 Wh 1 ~e3 23 tbh3 ~xc 1
my 16th move I automatically 24 lt)xf2 ~xb2, with the better
turned my attention to 16 ... :xd2, ending for Black. Another try - 21
but very quickly realized that Black ild7 'ir'xf2+ 22 Whl ~e3 23 ~xc6
achieves nothing with this move, - is refuted by 23 ...ilxg5 ! with the
since after 17 °W'xd2l%d8 18 'W'c1 deadly threat of 24 ...°W'f3+.
llxdl 19 llxdl he remains the ex- Yet White has a stronger de-
change down. fence. The players and annotators
''Then a flight of fancy inter- all overlooked the fairly simple
vened. I saw that with the white 21 l:td3 !, taking control of the im-
bishop on h3 instead of g2, the portant squares f3 and e3. After
combination would be playable. I 21...ti'xf2+ 22 Whl lt)c4 (22 ... h6?
spent 40 minutes on move sixteen. 23 l:tf3 'i!i'e224 ~fl. and the queen
At first I wanted to play 16 ...l:td7, is trapped), White has the pleasant
provoking 17 .i.h3 (with tempo!), choice between 23 l:tf3 'ir'xb2 24
but I decided that that would be too 'ir'xb2 lt)xb2 25 l%xt7 (25 lt)xt7!?)
superficial. White could simply 25 ... h6 26 ~e6 hxgS..27 .:t2+, and
continue with 17 .i.cl, preserving 23 'ii'fl 'ii'xfl + 24 ~xfl lt)xb2 25
a minimal positional plus. l%d7.In either case Tal could hardly
"In the event I chose 16 ... l:.d6, have saved the game.
whereupon to my astonishment Vasiukov thought of~ defensive
Vasiukov quickly played 17 .i.h3. idea that was not bad either, but de-
The retort 17...lhd2! was instanta- cidedly inferior to 21 l:.d3!.
neous ..." 21 ~fS?! g6
18_ 'ifxd2 22 b4! (D)
19 °iVcl The play suddenly takes on a
20 .:Xdl sharper character. Tal now has the
 Is it Worth Complicating? 113

choice between several possibili- t't)xa5 28 .i.xb7 ! (28 ... t't)xb7 29

ties. Which is the strongest? llbl).
[With 28 ... c6! 29 .i.c8 (threaten-
ing 30 'gb]) 29 ... ile7! Black
would retain the better chances,
since the rook is prevented from
penetrating and the passed a-
pawn is securely blockaded.
Dvoretsky]
22 ... axb4
23 cxb4 .i.e7
[I suggest it was worth seriously
considering the simple 23 ... hb4,
with good compensation for the
exchange. - Dvoretsky]
24 .i.d7 lbd4
The ga{Ile continuation was Tal writes: "Black doesn't want
22 ... axb4 23 cxb4 .i.e7. But why to simplify the position. He rejects
not place the bishop on d4? The 24 ... 'ii'xg5 25 .i.xc6 "ilfxcl 26
answer is that after 23 ....i.d4 24 b5 1:1.xclbxc6 27 l:txc6 .i.d6, and con-
lba5 25 .i.g4!- 'ii'xf2+ 26 @hl tinues to focus his attention on the
.i.e3, White has the defence 27 k.ingside."
lbh3 !. White's position hangs by There was no point in going in
this tactical finesse. for 24 ... lbxd7 25 :xd7 'it'xg5 26
Is it better to withdraw the bish- Wxg5 .i.xg5 27 l:1.xc7 (stronger
op to e7 at once, or after exchang- than 27 b5 4)a5 or 27 a3 .i.cl)
ing pawns on b4? What is the 27 ... lbxb4 28 a3!, when White ob-
difference? Either way Black can tains a dangerous passed a-pawn.
obtain two pieces for a rook if he 25 V/ixc7 .i.d8
wants, but the important thing is to 26 V/ixb7 "NixgS
stop White from activating his Objectively Tai has taken a risky
forces. After 22 ....i.e7!? 23 .i.d7 the decision. He has allowed his oppo-
c-file is not yet open, and Black has nent not one, but two passed pawns
the excellent resource 23 ... t't)b8!. on the queenside. He has, however,
Not that the consequences are by kept the queens on the board. In an
any means clear: 24 .i.g4 'ii'xg5 25 ending with rook and pawn against
'i!ixg5.i.xg5 26 bxa5 lbc4 27 .i.c8 ! two minor pieces, the presence of
114 Is it Worth Complicating?

an outside passed pawn becomes With both players in time-

the key factor for assessing the po- trouble, White blunders away his
sition; but in the middlegame there best pawn - the a-pawn. After 30
are chances for an attack on the l:tbl ©xb5 31 axb5 'ii'd3 32 l:lcl
king - after all, Black has one piece 'ii'xb5 331%c8Black would have no
more (though it is by no means a advantage. The simple 30 c;.t,g2
simple matter to bring the b6- (30 ... 'ii'xb4 31 'ii'b8 'ii'e7 32 a5)
knight and the dB-bishop into the looks even stronger.
attack). Moreover, Tai always han- 30 ... ©xa4!
dled his strongest piece with great 31 -.f.ig2
dexterity - suffice it to recall his 31 .i.xa4 ©e2+.
famous game against Oscar Panno 31 ... ©b6
in the 1958 Interzonal Tournament 32 l:.cS 'ii'f6?
at Portoroz. Black in tum goes wrong. Here
27 .i.e8 'iVf6 is Tal's explanation: "At this point
28 a4 ~! White unexpectedly offered a
It is important to drive the draw. Distracted, I forgot about the
bishop away and free the queen for immediate 32 ...~g7, which would
active operations. have ensured an easy win, and in-
29 .iLbS "ird6 (D) stantly played 32 ...1i'f6?."
We could draw an obvious con-
clusion about the importance of
being 'distraction-proof'. By the
way, take a good look at the posi-
tion after 32 ... ~g7!. See how ef-
fectively_ the black knights are
placed. Together with the other
pieces they control all the entry
squares!
33 'ii'b8!
White acquires the.counterplay
to save himself.
w 33 1i'f3+
34 ~gl 1idl+
Attacking b4 and threatening 35 ~g2 'iVf3+
29 ...©f3+. 36 ~gl lbe6
30 1%cl? 37 .:.c6 'ii'dl+
 ls it Worth Complicating? 115

38 c;;>g2 ~d4
39 l:d6 ~xe4+
40 c;;>gl 'iVl>l+
41 wg2 •h.-¼

Throughout the game Ta! was

constantly taking risks. To begin
with, he upset the balance by con-
ceding the advantage of the bish-
op-pair in the interests of rapid
development. Then he went in for
a decidedly dubious combination.
Finally, in his search for attacking
chances, he allowed his opponent queen's knight (it was on c6 be-
to obtain two united passed pawns. fore, whereas now that square is
Such was his style of play at that occupied by a pawn) and the white
time! dark-squared bishop. Both these
differences rather count in Black's
Ribli-Tal favour.
Candidates Tournament, Once again Tai will probably
Montpellier 1985 have to exchange his light-squared
Reti Opening · bishop for a knight, conceding the
advantage of the bishop-pair to his
1 ~f3 dS opponent - not that that is of much
2 g3 jlg4 significance here.
3 jlg2 c6 11 h3
4 b3 ~d7 11 'ii'e2, followed by ~c4-e3,
5 jlb2 ~gf6 was preferable.
6 0-0 e6 11 jlxf3
7 d3 jlcS 12 'ii'xf3 'iile7
8 ~bd2 0-0 13 :Sdl?!
9 e4 dxe4 Another routine move, after
10 dxe4 eS (D) which some difficulties already
The structure of the position is arise for White. He should have
more or less the same as in the put his a-pawn on a4, preventing
last game, isn't it? The only differ- not only the exchange of dark-
enc~ is in the placing of the black squared bishops with 13 ...jla3, but
116 ls it Worth Complicating?

also the move 13 ... b5, which robs

the knight of its rightful square c4.
13 ... b5!
14 h4?!
One error often leads to another.
Evidently Zoltan Ribli has thought
about his light-squared bishop and
decided to bring it out to h3, but in
so doing he weakens g4.
·white's top-priority task is to
re-deploy his knight that is poorly
placed on d2. For this, there is only B
one route: via fl to e3. Hence the
right move is 14 nfel !. by Tal, who never seems to forget
14 ... a5 about the enemy king.
Immediate occupation of g4 is 16 'ii'e6!
not feasible; 14 ... 'iVe6 can be met 17 'iVfS ltJg4
by 15 ii'f5 !. For the moment, there- 18 ::te2 l:.ad8
fore, Tal gives his opponent some Already White's position is dif-
worries on the queenside. He plans ficult. On 19 iLh3, Tal was intend-
15 ... a4. ing 19 ... .:t.d3! (with the threat of
15 c3 ..Jbg3+) 20 ~g2 'ii'xf5 21 exf5
So as to meet 15 ... a4 by 16 b4. ltJxf2 22 :xf2 iLxf2 23 ~xf2
However, the opening of the a-file J:tfd8 24 ~e2 e4.
would surely have been the lesser 19 iLf3 lld3!
evil, as another important square- 20 _ ~g2 itJxf2?
d) - has now been weakened. It must be said that Tai likes sac-
15 ... itJb6 rificing two pieces for a rook. In
16 l:tfel? (D) other words, the previous game -
Better is 16 'ii'e2 or 16 ~h3. where the distribution of forces
White's whole attention has been was the opposite - ~as rather an
fixed on the queenside in response exception.
to his opponent's last few moves. 21 lhf2 ~xf2(D)
Ribli is oblivious to possible diver- Can White somehow organize a
sions on the other side of the board, defence?
and incautiously weakens the f2- 22 'ii'xe6? is no good at all:
point. ThiS"is promptly exploited 22 ... fxe6 23 c;tixf2l:fxf3+ 24 ltJxf3
 Is it Worth Complicating? 117

w
l:txdl. After 22 @xf2? 'i!Vd6White 26 @g2?!
succumbs to a lethal pin on the d- If 26 @el or 26 ~e2, then
file. The best defensive chances 26 ...f4 27 gxf4 l:txf4 is strong. Per-
lay in 22 .i.e2 ! l:txd2 23 J:r.xd2.i.e3 haps it was worth trying 26 It.e 1 !? .
(23 ... .i.c5 24 'Wxe6 fxe6 25 1'.g4) 26 ... J:r.xf3!
24 ltd3 .i.c5 (24 ...'ii'xf5!? 25 exf5 27 lt)xf3
.i.c5 26 .i.f3 l:l.c8) 25 'ii'xe6 fxe6 In this case, if White continues
26 .i.f3 (or 26 l:tf3). Black has an 27 @xf3 'i!id3+ 28 ~f2 fxe4+ 29
extra pawn, but the win is still a @gl, then Black has 29 .. .'ii'xg3+
long way off. It is astonishing how 30 ~hl ltf2.
resilient a chess position can be; 27 ... 'it'xdl
for all White's numerous mistakes, 28 l£ig5 0-1
he could still have held on! White resigned as after28 ...'ifih5
22 'it>xn? °ifd6 or 28 .. ,\jid7 he is the exchange and
23 .i.cl g6 a pawn down.
23 ... 'ifc5+, with 24 .. .'ii'xc3 to
follow, is also strong. As you can see, the mature Tal
24 °ifg5 f6! (D) had lost none of his imagination,
Black must deflect White's queen and still had the same penchant for
before opening the f-file. Theim- attacking and playing combina-
mediate •24 ... f5 is unconvincing: tions. Yet he did so on a sound po-
25 @g2 J:r.xf3 26 @xf3 'ilid3+ 27 sitional basis - he tried not to take
@f2 fxe4+ 28 ~g 1 e3 29 'it'xe5. the same liberties that had been
, 25 'ii'h6 f5 characteristic of his youth.
118 Is it Worth Complicating?

The next game is quite different 10 ... 0-0

in character from the ones already Black also has other options.
examined. It has a direct bearing The move played clearly reveals
on the problem of prophylaxis. No his intentions - to prepare the ad-
wonder - the white side was played vance ...f5.
by Tigran Petrosian. 11 lbn (D)

Petrosian - lvkov
Nice Olympiad 1974
Nimzo-Indian Defence

1 d4 lbr6
2 c4 e6
3 lbc3 .i.b4
4 e3 c5
5 Ji.d3 lbc6
6 ~3 hc3+
7 bxc3 d6
8 e4 eS B
9 dS l'/:;Je7
10 lbd2 Petrosian realizes what his op-
If 10 0-0, Black has the prophy- ponent is planning, and takes
lactic reply 10...h6 !, so that a knight counter-measures in good time by
move can be met by 1l ... g5, stop- transferring his knight to g3. It
ping White from opening the game might have paid Black to switch
with f4. Spassky-Fischer, World plans with 1 l. ..lbg6, so as to jump
Championship match (5), Reykja- to f4 with his knight if the occasion
vik 1972 continued 10 l2Jh4 h6 11 arose. White in turn could respond
f4 (reckoning on l l...exf4 12 .i.xf4 to this move flexibly by placing his
g5 13 e5 ! with complications fa- knight on e3 and his pawns on g3
vourable to White) l 1...~g6! 12 and f3, then pushing his h-pawn -
l2Jxg6 fxg6. Here Spassky com- profiting from his delay in cas-
mitted a serious strategic error by tling. The opening move-order that
exchanging pawns on e5, leading Petrosian selected contains a fair
to a static position where the white amount of poison, yet for some
bishops had no scope. Fischer out- reason it has not been used lately.
played his opponent and won. 11 ... 'fiaS
 Is it Worth Complicating? 119

Black has decided to play on the power of the two bishops. Perhaps
kingside, so it is incomprehensible White was put off by 17... ltJxg3 18
that he brings his queen out on the hxg3 e4 !? (18 ...exf 4 19 ..i.xf4, in-
other wing. If he wanted to develop tending l:.ael and at some point
his queen on a5 he should have done ..i.xg6) 19 ..i.xe4 ~g7! (19 .....i.f5
so last move, when for one thing 20 g4!?; 19...ltJf6 20 ~xg6!? hxg6
the reply ..i.d2 was not possible, 21 'ii'xg6+ <if.ih8
22 f5 with a formi-
and for another Black still had the dable attack). However, Black's
possibility of castling queenside. compensation for the pawn is
12 ..i.d2 ltJe8 scarcely adequate.
13 ltJg3 rs Petrosian wants to preserve his
14 exf5 ~xf5 knight from exchange and is there-
Or 14.....i.xf5 15 ~xf5 ~xf5 16 fore in no hurry to start decisive
'iic2. operations. His opponent, however,
15 'ii'c2! g6 is given time to strengthen his po-
16 0-0 ..i.d7 (D) sition.
17 ... ~f6
18 ltigS
The knight is very strongly po-
sitioned here, as any attempt to
drive it away with ...h6 will weaken
the black kingside.
18 l:ae8
19 f3! (D)

w
An important point is that in
contrast to the King's Indian, the
black knight cannot reach d4 - that
square is covered by a white pawn.
17 4Je4
17 f4! looked attractive, so as to
open up the game and exploit the B
120 Is it Worth Complicating?

A typical Petrosian move. Tak- 22 1'.c2! 'iYa5

ing control of the e4- and g4- 23 a4
squares, he prepares to play g4 at a White has thus succeeded in
suitable moment, to deprive his limiting his opponent's possibili-
opponent's pieces of the f5-point. ties on the queenside. Yet even
19 t'i:Jg7 now the struggle is not over.
20 g4! (D) 23 ... 'iVc7
24 h3
White needs to bolster his g-
pawn, so as to prepare f3-f4.
24 ... a6
25 a5 (D)
Otherwise he would have to
reckon with 26 ...b5 - followed by
... c4, if White took on b5 with his
c-pawn.

The ex-World Champion of

course prevents the bishop ex-
change with 20 ... .i.f5, which his
opponent intended. All Black's
minor pieces are now shut out of
play. Despite this, the position still
remains unclear. To be completely
happy, White would have to clamp B
down on the queenside too, with
a4. 25 ... b5
20 ... 'iVa4 26 axb6
Borislav lvkov misses an excel- Which piece do you think Black
lent chance to confuse the issue, as should recapture with? Who bene-
indicated _by Petrosian: 20 ... b5 !? fits from a queen exchange?
21 cxb5 c4 22 .i.xc4 Jlxb5. The answer, of course, is that
21 'ifb3! l:tb8 Black does. In a cramped position
 Is it Worth Complicating? 121

you need to exchange pieces! I be- 38 :.Xf8+ wxrs

lieve that after 26 ...'i!fxb6! (271Wa2 39 il.xe4 l:tb3
'ii'b2) Black would preserve excel- 40 il.xc5+ We8
lent drawing chances. 41 110 1-0
Now recall the situation before
White's 17th move. The question In conclusion, I venture to show
is, have Petrosian's refined ma- you a game of my own.
noeuvres been justified? Wouldn't
it have been simpler to play 17 Tsariov - Vulfson
f4!?, 'cutting the Gordian knot' at Moscow 1989
once? Sicilian Defence
26 lhb6?
27 'illa3 1 e4 c5
White will bring his queen 2 lt.Jc3 lbc6
across to the kingside for the at- 3 f4 e6
tack, while the black queen lacks 4 lt.Jf3 d5
all scope. 5 d3
27 °ifd8 White was evidently happy with
28 'ifcl Wle7 the ending after 5 ... dxe4 6 dxe4. I
29 li'el ltb2 was aiming for a more complex
"A lone man on the battlefield is game.
no warrior." White will soon drive 5 lt.Jf6
the rook off the second rank. 6 e5 lt.Jd7
30 il.d3 il.c8 7 g3 b5!?
31 il.cl l:tb3 Usually this advance needs to be
32 .i.c2 l:tb6 prepared, but here I had the chance
The work of prophylaxis is suc- to carry it out at once.
cessfully completed. What now 8 il.g2 b4
follows is, essentially, the first ac- 9 lt.Je2 g6?!
tive move of the game. It was better to carry on in the
33 f4! h6 same spirit with 9 ... a5 and then
34 fxe5 Wlxe5 ... lbb6. However, I thought it was
35 'illxe5 dxe5 important to restrain White's king-
36 lt.Je4 h5 side pawns.
37 .i.a3 10 0-0 h5 (D)
The harvest commences. White's standard plan in such
37 ... lt.Jxe4 positions involves a kingside pawn
122 Is it Worth Complicating?

advance: h3, g4 and eventually f5. 12 ... Ji.a6

In the present case, however, he 13 ltJg5!
has another highly promising plan. Another powerful move. We
Have a try at finding it. now see that 13 ... JLe7 can be met
by 14 c4! Ji.xg5 15 cxd5!. Black
therefore guards his c6-knight.
13 ... 'ikc7
14 c4!
Think what happens if the pawn
sacrifice is accepted.
If 14... dxc4, then 15 'ii'a4. Then
after 15 ... cxd3 16 ltJc3, there is,
for example, 16... d2 17 'ii'xa6 or
16... Ji.b7 17 ltJb5 followed by 18
ltJe4. This all looks extremely dan-
gerous for Black, yet his defence
w can be strengthened. Instead of
taking the second pawn, he should
Taking into account my back- play the immediate 15... Ji.b7! 16
ward development, it wouldn't be dxc4 ltJb6 17 'ii'c2 ltJd4, with a
a bad idea to strike in the centre by tenable game.
pushing the_c-pawn. Black's pawn I preferred to close the position.
on b4 makes it a little harder to im- Alas, this failed to get Black out of
plement this idea, yet White could serious trouble.
very well play 11 c4 bxc3 12 bxc3, 14 ... d4? (D)
intending 13 c4. My opponent found I had pi:rµied my hopes on the in-
a more cunning way to carry out activity of two white pieces: the
the same plan. e2-knight and the cl-bishop. Yet
11 a3!? bxa3 after the positional pawn sacrifice
It was worth thinking about 15 f5, these pieces come to life.
11... a5 12 axb4 cxb4. Still, after 13 Another way for White.to develop
~e3 (stronger than 13 c4 bxc3 14 his initiative is 15 'ii'a4 Ji.b7 16
bxc3 ii.a6) White would be better. l:.bl.
12 bxa3! So White has two continuations
I had only reckoned on 12 l:.xa3 to choose from. Whic~ would you
°ifb6 13 c4 d4, with obscure conse- have preferred? Calculating the
quences. variations to the end would be
 ls it Worth Complicating? 123

bogged down in the mass of varia-

tions, and eventually didn't risk it.
"I had the feeling I ought to play it,
but I couldn't work it all out," he
explained after the game. "Why
should you work it all out?" I asked
in amazement. "If an idea like that
came into my head, I'd be sure to
play the sacrifice."
Shying away from the sacrifice,
my opponent persuaded himself he
w could win with 15 °fi'a4.
[He was right, too. This way
fairly complicated (especially in White does indeed get a large plus,
the 15 fS line). At a certain point and he does it by simple means,
you would have to trust to intui- without taking risks. From the
tion. practical viewpoint, White's deci-
After 15 f5!? gxf5 16 ll'if4 the sion was the most expedient. -
threat is 17 lt)fxe6, when there Dolmatov]
would be no rescuing the black 15 'Wa4 .i.b7
king. The best defence is 16...l:th6, 16 l:tbl ~b6
but then comes 17 ~xn !! ~xf7 18 11 fts llb8
lt)xe6!: Black's gaping wound - the b-
a) 18 ... lhe6 19 'iixh5+ ~g7 file - needs covering up.
(19 ... ~gS 20 .i.d5 threatening 21 18 lt)e4 lt)d7
'Wg6+; 19, ..J:tg6 20 .i.d5+ @g7 21 Otherwise the c5-pawn can't be
.i.h6+ ! .:xh6 22 'Wf7+) 20 .i.d5 ! defended.
'ifxeS 21 .i.h6+! with a mating at- 19 lt:if6+! (D)
tack. The point of White's play. If
b) 18 ...@xe6 allows White two 19 ...lt)xf6, then 20 .i.xc6+ is deci-
strong replies: 19 .i.d5+ @e7 20 sive. However, he clearly underes-
.Jlg5+ ~e8 21 .Jlxh6 or 19 'ii'f3 ! timated my reply.
with irresistible threats. 19 ... ~d8!
To be quite honest, I hadn't seen It turns out that there is no
the pawn sacrifice. Tsariov did see forced win: 20 lt)xd7 '.txd7 (in-
it, and made a conscientious effort tending 21.. ..Jle7) 21 .i.xc6+ 1Wxc6
to c~lculate it. :flowever, he got 22 'Wxc6+ ~xc6. Black threatens
 124 ls it Worth Complicating?

play? White was threatening to

strengthen his position with ~h2
and liJgl-J3, or 1:1b5and ~cl-b3.
The J6-pawn is seriously cramping
Black, whose king feels highly un-
comfortable in the centre. It must
be acknowledged that White has a
distinct plus. - Dolmatov]

20 ... iLa8, and the e2-knight is out

of play just as before.
20 ild2 iLa8
Not 20 ... ~xf6? on account of
21 iLxc6!.
21 'ifa4
22 .:.Xbl
23 exf6 B
24 h4!
Increasing the pressure is not Again we face a dilemma. If
simple, but my opponent sets a cun- 25 ...'Wc8, White will strengthen
ning psychological trap. He de- his position with J:1b5and ~cl-b3.
fends against ...h5-h4, which might The alternative is to give up the
in some circumstances be play- queen with 25 ... l:b8. From the
able, and thereby seems to demon- practical standpoint it _is usually
strate that his queenside initiative best to choose the most active con-
has dried up. I rose to the bait and tinuation. Let us see: 25 ...1:1b8!?26
incautiously made the natural iLxc7 1:lxbl+ 27 'itiih2! (preparing
move. lbgl-f3) 27 .. .cJ;;xc7.It would now
24 ... ~d7?! of course be absurd to win the a7-
25 ila5! (DJ pawn with 28 iLxc6?. After 28
[Allowing a blow like this is un- lbgl [28 'fi'c2 first is technically
pleasurtt, of course, but how else better - Dolmatov] 28 ...l:b2 the
was Black to,,bring his rook into advantage is certainly on White's
 ls it Worth Complicating? 125

side, but in mutual time-trouble l:tadl ¥.i'c811 ¥.i'cl l:td8 12 lhd8+

Black should have played this way ¥.i'xd8131:tdl 'iff8 14 h3 .i.xf3 15
all the same. ii.xf3 (D)
25 ... 'iVcS?!
26 'ii'b5?
A picturesque position would
arise after 26 l:tb5! 'it'a6 (I was
banking on this pin when I played
25 ... 'ifc8) 27 ltJc 1 l:tb8 28 t2Jb3
'it>c8. Now 29 ltJxc5? fails against
29 ...l:txb5 30 ltJxa6 l:txa5. How-
ever, what move is Black to make
after, say, 29 'it>h2... ? He turns out
to be in zugzwang in the middle-
game ! For example 29 ... l:tb6 30
.l:txb6 axb6 31 ii.xc6, or 29 ...ii.f8
30 :Xb8+ 'it>xb$ 31 ii.c7+.
26 °fVb8! 15 ... a6?!
27 .i.el a6 A move recommended at that
28 'iVxb8 !hb8 time by theory, on the basis of
29 lhb8 ii.xb8 Timman-Matul<;,vic, Wijk aan Zee
By forcing the queen exchange 1974, which continued 16 lDd5?
Black has attained equality. Subse- t2Jxd5 17 l:txd5 t2Jd4 18 .i.xd4
quently my opponent became too exd4 19 'ir'f4 'flie7 20 e5 l:.e8 21
obsessed with searching for win- 'ii'xd4 J..xe5 with a quick draw.
ning chances - which have already The straightforward 15 ...ltd8
melted away - and actually lost. leads to an ending which is favour-
able for White: 16 1:txd8 'ii'xd8 17
Supplement 'iidl 'iixdl+ (17 ...l!fe7!?) 18 ii.xdl.
Andersson-Hazai, Pula 1975 went
Petrosian - Sax 18... ltJd4 19 f3 ltJe8? (19 ... a6 20
Tallinn 1979 ltJa4 t2Jd7) 20 ltJd5 c6 21 ltJe7+
Pirc Defence ~f8 22 lcic8 a6 23 c3 t2Jb5 24 a4
t2Jbc7 25 ii.b3 J..f6 26 t2Jb6 t2Je6
1 lt:)f3 g6 2 e4 J..g7 3 d4 d6 4 ltJc3 (26 ...'it>e727 J..c5+) 27 .i.xe6 fxe6
t2Jf65 .i.e2 0-0 6 0-0 ii.g4 7 ii.e3 28 .i.c5+ 'it>f7 29 lcid7 ~g7 30
ltJc6 8 'ii'd2 e5 9 dxe5 dxe5 10 <it>t"2
.i.h4+ 31 g3 ii.f6 32 We2 .i.d8
126 ls it Worth Complicating?

33 ilf2 ilf6 34 lbc5 lbd6 35 21 'Wc2 lbdS

lbxe6+ <"J;,;f7 36 lbc5 ild8 37 b3 If 21...ilc5, then 22 ilg5. fol-
~a5 38 b4 ilc7 39 a5 ~b8 40 ~e3 lowed by a queenside pawn ad-
1-0. vance with a2-a4 and b2-b4.
In Petrosian's view Black should 22 'i!ib3 °ii'd3?! 23 'ii'c4 °ii'd6 24
attempt to exchange dark-squared 'ife2 'We6 25 'Wd3 lbc6 (better is
bishops with 15 ...h5 !? followed by 25 ...'ii'e7 followed by 26 ...lbe6) 26
...<"J;,;h7
and ...ilh6. a3 'i!ie7 27 b4 lbd8 28 lt:ic4 lt:id7
16 lbbl! l:d8 17 J:txd8 'ii'xd8 29 ~g4! lt:ie6 30 ltJaS! b5 31 lbc6
18 c3 °ii'd3 19 lbd2 ..i.f8 20 'ii'bl 'ii'e8 32 c4! lbf6 33 cxbS axb5 34
'fi'b5 'ii'xb5 lbxe4 35 'ifc4 lbd6 36 'ii'd5
If the queens are exchanged h5 37 ilxe6 fxe6 38 'ifc5 lbf5 39
Black has an unpleasant ending, 'ifc2 .i.g7 40 b5 lbd4 41 'Wc4'Wd7
much as in Andersson-:::,Hazai. 42 a4 lt:if5 43 'ife2 1-0
 8 Thoughts about a Book
Mark Dvoretsky

There are no hopeless positions; can state with great pleasure that
there are only inferior ones that can Simagin was right.
be saved. There are no drawn posi- Secondly, as a chess trainer, I
tions; there are only equal ones in am in constant need of fresh, high-
which you can play for a win. But quality material. There is no lack
at the same time, don't forget that of interesting games played in the
there is no such thing as a won po- chess world, but once published -
sition in which it is impossible to in magazines or Jnformator - they
lose. become familiar not only to a coach
Grigory Sanakoev but also to his students. However,
the world of correspondence chess
In all my life I have only played is almost entirely ignored by over-
two games by correspondence, and the-board players - unjustly, for
I am unlikely to play any more. plenty of ideas can be gleaned from
Despite this, when I heard about it, striking and profound ones, dili-
the publication of Third Attempt, a gently worked out in home analysis.
collection of games by Grigory Books that confine themselves
Sanakoev, the twelfth World Cor- to giving the moves of games with
respondence Chess Champion, I explanatory variations may be in-
immediately acquired a copy. There structive but are not very interest-
were several reasons why. ing. Happily, the book in question
In the first place, I recall that acquaints us not only with some
when Grandmaster Vladimir Sima- fascinating duels but with the liv-
gin won the USSR Correspondence ing human being who fought them
Championship in the mid-1960s, - his experiences and thoughts, his
he spoke· with great respect about opinions, his advice. One thing I
the class of play of one of his op- find particularly impressive about
ponents - Grigory Sanakoev. Ac- the book is its use of quotations
quainting myself with the book, I (always relevant, I may add) by
128 Thoughts about a Book

famous thinkers of the past. Chess chess player, a man of wide learn-
is one of the branches of universal ing whose thinking is nonetheless
human culture, and we ought not original. (I am sure this image is
to impoverish ourselves by focus- faithful, though I don't know Sana-
ing solely on its narrowly profes- koev personally - you cannot de-
sional aspects. ceive an experienced reader!)
Contrary to the author's convic-
tion, I am sceptical about the pros- I could not help beginning by
pects for correspondence chess. setting down my overall positive
The appearance of computers ca- impression of Sanakoev's book,
pable of analysing at grandmaster but that was not at all the reason
level will inevitably tempt players why I 'took up my pen' (an out-
to use their services to attain good dated cliche in the computer age!).
competitive results: We know that The theme of this chapter lies in
many leading over-the-board grand- certain fundamental problems of
masters, including the thirteenth chess intelligence which my read-
World Champion, have succumbed ing of the book prompted me to
to the temptation and make every think about.
possible use of powerful comput- I chose what seemed to me the
ers for opening analysis. (The ad- most notable encounters in the
verse effect of this process on the book, and gave them to Grandmas-
popularity of chess is obvious-af- ter Vadim Zviagintsev for study.
ter all, the fans are interested in a His task was to look at the critical
contest of personalities, not of ma- moments of the games and work
chines.) And in correspondence out the difficult decisions for him-
chess the computer can be used self (without moving the pieces on
throughout the entire game. the board, of course) - or some-
Yet there is no doubt that Sana- times to find the complete sequence
koev has always played independ- of moves in the crucial phase,
ently and always will. What 'attracts when the outcome of the game was
him in chess is first and foremost decided. In many ca~es the young
the creative endeavour, the single grandmaster at over-the-board play
combat of intellects; the result is came to different conclusions from
only secondary. A reading of the the experienced correspondence
book_conjures up a highly conge- player. We singled out these situa-
nial image of the author as a bright, tions for further analysis, discus-
uncompromising, self-confident sion and clarification.
 Thoughts about a Book 129

Let me state that even the most

conscientious analysis, if scruti-
nized in depth, will prove to con-
tain controversial points, sometimes
downright errors - such is the
complexity of chess. For that rea-
son, the following critical exami-
nation of certain episodes from
Sanakoev's games is by no means
intended to cast a shadow over his
book. I have adopted the same ap-
proach before, when writing about w
excellent works by Jan Timman
and John Nunn - which I had used Dvoretsky - Peev
for training purposes with Sergei European Champions' Cup,
Dolmatov. Less significant books Plovdiv 1975
would simpJ.ynot have come within
the orbit of our attention. however, is 29 .:U-4!with the terri-
ble threat of 30 l:tc4. If 29 ... ~d6,
A penchant for brilliant then 30 l:.b6 'ili'c5+ 31 l:.d4, and
moves Black has no moves; while if
29 ...lld2, there is the beautiful
If a chess player discovers a strik- variation 30 'ii'a8+ ~c7 31 l1xb7+!
ing, outwardly attractive possibil- ~d6! 32 l:.d7+! 'ifxd7 33 'ii°a3+
ity, he often falls under its spell 'itie5 34 ~xd7+ ~xf4 35 'ii'c 1 'ite3
and can no longer resist its tempta- 36 ~e5 and wins.
tion. This must have caused every Unfortunately I was enticed by
one of us to drop points at some a flashy move which I had been
time or other. I can't help showing planning but which turned out to
you a memorable example from be none too effective.
one of my own games. 29 ~d7?!
Into a threefold attack!
The _position looks won for 29 ... l:.xd7!
White. Admittedly, 29 l:tb6?'ii'c5+ The only defence; 29 ...'ir°xd7
is useless; 29 c4 'ii'a6 ! is uncon- loses to 30 'ir'a8+, while 29 ...Citi>xd7
vincing, and so is 29 'Wa8+ 'it;c7 30 is met by the decisive 30 :Ixb7+
'Wa5+ b6. A very strong move, 'ite8 31 lixf5!. If 29 ...~d6, then if
130 Thoughts about a Book

nothing else White has 30 lt)e5

'if d5 31 lt)xf7.
After Black's move, I decided
that the line I had planned - 30
Was+<j;c7 31 11f xh8 - led to an
immediate draw: 31...ii'c5+! (but
not 31. ..l:td2? 32 'ii'e5+ followed
by 33 l:tf2) 32 '.th2 'iid6+ 33 ~gl
'it'c5+ 34 l1f2 l:td2 35 lUl li)g3 36
'6'f6 li)xfl 37 'it'xf7+ ~b8 38
<j;xf} .:.Xf2+ 391i'xf2 'ii'xc3. How-
ever, my attempt to carry on the
fight with 30 J:.xfS?! gxfS 31
'ii'a8+ <j;c7 32 'ii'xh8 proved even Engel - Sanakoev
weaker on account of 32 .. JWe4!. Romanian Chess Federation
After 33 .:n l:[d2 34 'ii'g7 peace Jubilee Tournament 1976-9
was concluded, though Black's po-
sition is already slightly better. e6-point with ....i.d5, and if the
case arises it can move forward to
My examination of Sanakoev's e4. Hence 24 ...b4 seems indicated.
games showed me that one of his In Sanakoev's view, after 25 li)dI
characteristics is this same- ten- l:tc826 ~3 White has a clear plus.
dency to select pretty moves - even (The knight, indeed. looks likely to
at the ex~nse, sometimes, of qual- jump to g4.) .
ity. But why should Black allow the
knight to leave d I? Instead of
White intends 25 'ii'h3 followed 25 ...J:tcS, a much stronger line is
by 26 hxg6. How substantial are 25 ....i.c5 ! 26 'ilh3 ..i.d5. Now the
his threats? They must certainly be black rook is ready to go to c8. On
taken seriously. For instance, the 27 hxg6 fxg6, Black has every-
natural 25 ...:CS is answered by 26 thing securely defepded, and lLJe3
hxg6 fxg6 27 'ii'h3 ..i.d5 28 ~bl !, will always be met by ...Jlxe3. The
preparing 28 lt)xd5. consequences of 27 Jlxb4 .i.xb3
In general terms. the knight is (27 ....i.xb4 28 J:txb4 :CS also mer-
w~ll .placed on c3. From here it its attention) are uncertain. Black
deprives Black of the convenient would clearly be justified in play-
possibility of strengthening the ing this way.
 Thoughts about a Book 131

The above considerations are It would be wrong for the king

prosaic. The solution found by Sa- to go forward (29 ...~f5? 301i'h3+
nakoev is, on the contrary, highly ~f4 31 .l:th4+ q;e3 32 l:te4+! .i.xe4
dramatic. 33 fxe4+ and 34 "ifxc3), so the
.24 ... ~-~"
.JiiLa,;J •• game ends in perpetual check .
Black aims to answer 25 bxa3 Sanakoev, of course, saw this
with 25 ...l:.xd2 ! 26 @xd2 l:.d8+ 27 variation and- mentioned it in his
'it>cl 'ifxc3 28 ~bl J:r.d229 'ifcl book. He is a highly combative
i.xf3 30 J:Uh3 l:.e2, when Black player, and in other circumstances
has the advantage. White does even he would hardly have given his op-
worse with 25 1:tlh2 l:.xd2! 26 ponent the chance to 'bail out' like
Jhd2 'ii'xc3. this. In the present case, though,
In the game, there followed 25 the magic of the beautiful move
lobl ?? .i.cS! (this time Black over- must have influenced his conclu-
came the temptation to play for a sion that other continuations gave
'brilliancy' with 25 ... lud2? 26 White the advantage.
loxd2 'iVc3~counting on 27 bxa3?
:CS! 28 'if d 1 J:.d8 ! - the attack can
be repulsed by 27 'Wd3 !) 26 'iVh3
1Wc6!27 hxg6 'iVxf3! 28 gxh7+
'it>h829 'ifxf3 .i.xf3 and Black had
an obvious endgame plus, which
he duly converted into a win.
Zviagintsev too thought about
24 ... .i.a3, but had doubts about
playing it, because he saw that
White could force a draw. He was
unsure whether to go into the com-
plicated position that results from
24 ...b4!?.
25 hxg6! J:.:xd2! A. Zaitsev - Sanakoev
26 lhb7 'ibc3! 6th USSR Championship 1963-5
21 mis+ <li/g7
23· l:tlb7+ 25 ... .i.xh4
28 l:t8h7+ gives the same result. 26 J:.hl .i.xg3!
28 ~g6 The exclamation mark was added
29 Ah6+ ~g7 by me. Sanakoev himself considers
132 Thoughts about a Book

this move dubious. Here is what he total correctness of Black's formi-

writes: dable attack.
"The temptation was too great ... . In situations like this, postal
I recalled that 'the wise man un- players are probably intent on ana-
derstands it is simpler to deny lysing the variations as deeply and
himself a passion than to struggle precisely as possible. An over-the-
against it afterwards' (La Roche- board player, on the other hand,
foucauld), but with the chance of a with neither sufficient thinking time
sacrificial attack against Zaitsev, I nor the right to move the pieces
thought 'No! I can't chicken out.' about, is obliged to break off his
"If you assess the manoeuvre calculations at the earliest suitable
... .i.xh4xg3 in purely objective moment and apply his positional
terms, you have to conclude that judgement. That is why the ability
26 ... .i.g5 was stronger. After the to judge correctly is perhaps more
modest reply 27 Jhh.5, Black would weakly developed in correspon-
continue 27 ....i.xd2 28 .i.xd2 bxc4, dence chess specialists than in or-
with a queenside initiative; White dinary players; the former simply
would have no serious counter- have less practice, as they solve
play. This would have guaranteed most of their problems by analyti-
me long months of very pleasant cal means.
analysis, with evaluations ranging Still, even if valid as a whole, no
from 'better' to 'much better'. On remark of this type can cover all
the other hand, the more uncom- cases. For example, acquainting
promising 27 cxb5 would allow myself with the games of Mikhail
the pawn sacrifice 27 ... h4 ! 28 bxa6 Umansky, another World Corre-
"fle7 followed by :..hxg3, with a spondence Chess Champion, I was
dark-square attack that would not impressed precisely by their deeply
be too simple to resist." · conceived strategy.
To me, the emphasis here seems 27 <i?i>xg3 lof4
wrong. After 26 ... .i.g5 27 .:r.xh5 28 1Ve3 "i!le7
~xd2 28 .i.xd2 bxc4 29 bxc4 29 <i?i>f2 "i!lgS
Black has a good game, but noth- 30 ~n (D)
ing more. By contrast, the piece 30 .:r.h2 is bad on account of
sacrifice is not only tempting but 30 ...h4 31 lofl bxc4 32 bxc4 .i.xc4
~ery strong. You only have to look 33 ~xc4 .:.xc4 34 'ii'b3 l:tec8 35
at the position after two or three .ilxf4 exf4.
more moves to be convinced of the 30 ... h4
 Thoughts about a Book 133

An excellent move, cutting the

white queen off from the important
square e3. If 34 'iVxd6, there fol-
lows 34 ...°iVh4+35 ..ti>d
1 Wf2, and
then ...lkc8-d8.
34 .i.a3 (D)

Sanakoev makes no comment on

this move, though it is open toques-
tion - for after 31 J:[g1 the queen
must retreat\. the check on h4 being
unavailable. However, ...h3 pres-
ently follows, and h4 becomes ac- B
cessible to the queen again.
Black had another attractive at- In such tense situations you
tacking possibility, suggested by can rarely succeed in conducting
Zviagintsev: 30 ... bxc4! 31 bxc4 the attack by simply making one
.i.xc4 32 lbxc4 :Xc4 33 'ifb3 .:CC8. common-sense move after an-
In my view, White would have no other. A moment comes when you
real saving chances after either 34 need to exert yourself, to discover
itxf4 .:c2+ 35 3te2 exf4, or 34 and calculate a clear-cut path to
il.xc4 'ti'g2+ 35 ~e3 'ii'xhl 36 .i.b2 your goal. Doing this in a corre-
'Wg2(h2) - and if 37 l:tcl, then spondence game is of course far
37 ...l:.xc4!. easier than in a normal one. Sana-
31 l:tgl 'i!Vh6 koev saw the correct solution, but
32 'ifb6! then Zviagintsev found it too:
White must prepare to withdraw 34 ... lbg2+! 35 ~di (35 .i.xg2?
his king to el, which is no good at loses quickly to 35 ...'ti'e3+ 36 ~fl
present owing to 32 ...lbg2+. hxg2+ 37 J:[xg2? 1'.h3, while 35
32 h3 l:txg2 hxg2 36 .i.xg2 'ife3+ 37
33 ~el .:tcS! ~dl amounts to a transposition)
134 Thoughts about a Book

35 ... 'ii'e3 36 l:.xg2 hxg2 37 .i.xg2 intuitively reject 34 ...lbd5 and look
bxc4 38 .i.xc5 dxcS 39 bxc4 for something else.
bc4! (39 ...li'd4? 40 l:[cl l:[d8 41 37 h2
'ir'a5) 40 lbxc4 'ii'd4+! 41 lbdl (41 38 exf7+ <:J;xr1
'iite2 or 41 'iitc2 would be very bad 39 'ii'xd6 'ii'd4
due to 41...'iixc4+) 41 ... 'ili'xal+ 42 Concrete analysis convinced
'iite2 'ii'xa2, and Black undoubt- Black that after 39 ....:CCS40 'ii'd5+
edly has a winning position. his king would be exposed to a de-
Why, then, did Sanakoev reject cisive attack.
this line? For one thing, he was not 40 .i.xcS 'ii'xd6
entirely sure how to assess the end 41 .i.xd6 hl 'iii'
position. But the main reason was 42 'iii'c2 'ii'h6
the alluring prospect of playing for 43 c5 'ii'e3
a 'brilliancy'. 44 a4!
34 ... ~dS? White now has an advantage
35 exd5! which is both material (three mi-
"White has no reason to plunge nor pieces for the queen) and posi-
into a jungle of variations such as tional, and which he duly conducted
35 Wxd6 'We3+ 36 'itdl b4 37 to victory.
cxd5 Wxgl 38 ..i.xb4 l:[cc8, or to
try to clear up the position with the Limits of calculation
dubious 35 ..i.xc5 lbxb6 36 ..i.xb6
bxc4 37 bxc4 .:r.b8 38 .i.f2 l:.b2. In As already noted, it is natural for
either case it is not exactly clear over-the-board players to want to
what he does about the pawn on conclude their analysis of varia-
h3." (Sanakoev) tions as soon as it is feasible. That
35 'ii'e3+ way they save time and energy, yet
36 ~dl 'ii'xgl sometimes they fail to probe into
37 dxe6 the position deeply enough; they
Black has no time to recapture on will fail to notice some concealed
e6, e.g. 37 ...fxe6 38 ..i.xc5 dxc5 39 tactical or strategic _resources, and
~c2! h2 40 l:[bl ! hl"ii' 41 .i.d3. hence miss the strongest continua-
Otherwise, however, White takes tion. This cannot be helped; "Real
on f7 and Black's king becomes life is, to most men, a long second-
highly insecure. On reaching this best. a perpetual compromise be-
position in his analysis, surely, an tween the ideal and the possible ...
over-the-board grandmaster would (Bertrand Russell).
 Thoughts about a Book 135

•Razuvaev analysed this far and

rejected 18 J:.cdl. He was wrong!
21 lbxcS lhcS
22 lhg7+! (D)

Razuvaev - Beliavsky
USSR Championship, Minsk 1979

Yuri R~uvaev settled for the B

immediate win of a pawn with 18
lbxe4?!, which allowed Black to 22 ... ~xg7
equalize with 18 ....i.xe4 19 'ii'xe4 23 'ii'xa7+
'ifdS!. The remaining moves were White wins his rook back and
20 'ii'xd5 exd5 21 lUdl q.,e6 22 emerges with two extra pawns.
ffl .i.d6 23 hd6 ¼-¼. Many correspondence games
An intermediate rook move to supply us with excellent training
dl looked attractive, but in the case material for overcoming this psy-
of 18 l:tfd 1 Black has the strong re- chological barrier and widening
ply 18...'ii'e8! 19l:td7+ ~g6, when the analytical horizon. After all,
20 lbxe4? fails to 20...hfl+. White postal players normally continue
must attack the queen with the analysing where over-the-board
other rook. players would be certain to stop.
18 :Ledl! 'ife8?! The important thing is merely to
18 ... 'iib6 is a better chance, al- choose examples where the diffi-
though after 19 l:td7+ followecl by culties facing the player are not
20 lbxe4 the initiative remains with analytical in nature, but conceptual
White. and psychological.
19 l:td7+ ~g6 "In the following game l pulled
'\ 20 lbxe4 e5 off what was probably the longest
136 Thoughts about a Book

and most complicated trap of my (he reaches this position by an-

career." (Sanakoev) other move-order: 29 g3 d5 30
.:e3) 30 ... ~xh3+ 31 'iitg2 ~g5 32
l2Jxg5 hxg5 33 l:.hl+.
Another dismal prospect for
Black is 29 ... e4 30 4Jd4 d5 31 h4
(31 g3 !? l2Jxh3+ 32 ~g2 lbg5 33
'ii'xd5) 31...l:tadS? 32 g3 lbd3 33
lt:)ac6 l:.d7 34 l:.dxd3. After other
moves White will drive the knight
back with g3 all the same, and ob-
tain a won position.
However, why not expel the
knight at once? The point is that
w' Black has a neat tactical resource
from which he emerges a pawn up.
Sanakoev - Sevecek 29 g3 ~xh3+
6th World Championship 30 ~g2 ~!
1968-70 Here is a psychological barrier
for an over-the-board player: on
How should we assess this posi- spotting this blow, he would be al-
tion? Sanakoev writes: "Black ap- most certain to break off the analy-
pears to have got what he wants. sis and look for a more solid
The knight remains out of play on continuation. Sanakoev continued
a5; next move Black will carry out studying the position and came to
the long-awaited advance ...d6-d5. the conclusion that White can now
fu the ensuing fight White will cer- win by force.
tainly have the better chances on That being so. 29 g3 was an ex-
account of his healthy extra pawn, cellent trap - creating the perfect
but the outcome is wholly un- illusion of a blunder. Black must
clear." have concluded tha! White simply
I would alter the emphasis. missed the capture on f2.
White has a positional plus as well 31 ~ 1ib6+
as a material one. After. for in- 32 ..tg2 lhaS
stance. the prophylactic 29 l:te3!?, 33 'Wif7 l:.aa8
Black cannot play 29 ... d5 because The only defence.
of a line that Sanakoev gives: 30 g3 34 ©114 (D)
 Thoughts about a Book 137

37 l:tc4! 1-0
Black resigned due to 37 ...'ii'xc4
38 lbxe5+ or 37 ... ~b2+ 38 l:tc2
fixb4 39 lbxf8+ ~g8 40 ii'h7+
'itxfS 41 l:tfl+ @e7 42 'ilkxg7+.
This variation is not an integral
part of the trap, since it is not com-
pulsory - White has other ways of
winning.
The attentive reader may be
wondering why I did not award two
exclamation marks to 29 g3, as Sa-
nakoev does. The answer is that I
''The end position of the trap", have doubts about the move's ob-
according to Sanakoev. That is not jective strength. Black had an extra
strictly true; the analysis is not fin- defensive possibility in 34 ... 'i!i'a7!.
ished yet. ,:Hhat is White aiming Sanakoev considers that after the
for? Not 35 lbg6+ 'it>h7 36 'ii°f5, continuation 35 lllg6+ 'ith7 36
which is harmless due to 36 ...'it>g8. lbxf8+ :Xf8 37 'iVxa7 :Xa7 38
The real threat is 35 l:tg4 followed l:[xd6 White wins easily, since the
by 36 lbg6+ 'it>h7 37 lbxf8+, but counter-attack on the second rank
Black can parry it by pinning the is useless: after 38 .. .l:ta2+39 'it>h3,
rook. if 39 ...l:.ff2 then the king escapes
34 ... 'ii'c6 the checks via g4. In fact, though, a
We now encounter a second counter-attack is also possible on
psychological barrier; this, inciden- the third rank: 39 .. .l:U-3!(D).
tally, is where Zviagintsev stum- For example, 40 l:tc6 (40 l:.d7 !?
bled, having overcome the first l:txc3 41 l:[g4 'it>h8is not convinc-
one. ing either) 40 ....:C2 (40 ...:a3!?) 41
35 lbg6+ ~h7 c4 J:.cc3 42 J:.g4 g5.
36 _.fS Is there a win here? If there is, it
It turns out that with his queen can perhaps only be established in
on c6 Black cannot defend him- correspondence play - the problem
self, since 36 ...~g8 now comes up has become purely analytical, and
against 37 'ii'xf8+! l:.xf8 38 lbe7+. everything hangs by a hair. After
A splendid conception! 43 b5 l:r.f244 J:.e4 ! :tt3 (44 ...h5 45
' 36 ... 'iVxc3 .:xe5) 45 lbe5!? or 45 @g4!?,
138 Thoughts about a Book

problems. The question of chess

styles is very important and de-
serves a few remarks, however
brief.
Logically, it is clear that if a con-
tinuation - whether positional or
tactical in nature - is indisputably
strongest, it ought to be selected by
a player irrespective of his style. If
it is not selected. then we are no
longer talking about style but about
a limitation in his manner of play.
Style comes to the fore, above all,
White seems to win. The same re- • in a situation where there is a choice
sult is reached after 43 ...l:r.fe3 44 between possibilities of roughly
c5 e4 (44 ...l:r.e245 l:r.a4!) 45 b6 equal worth (particularly a choice
l:r.e2 46 b7 :cc2 47 l:r.xg5.How- of opening strategies). Of course,
ever, there is still 43 ...lU5!? 44 c5 this is just a rough outline - in ac-
h5 45 l:r.a4l:tf7! to consider. Then tual fact everything is much more
if 46 b6? !, Black has 46 ...l:r.c2! complex. There are very many con-
(threatening 47 ...g4+ 48 q.,h4 l:tg7) troversial borderline situations, and
47 l:r.a7l:r.xa748 bxa7 l:r.a2. The anyway decisions are sometimes
play can surely be improved for taken on psychological grounds.
both sides, but all this is too com- There is nothing wrong with that.
plicated and unclear. "An experienced chess player will
We can now see why it was so sometimes opt for a particular
important to arrive at the right continuation without being at all
evaluation of the position before convinced that it is the best of all
White's 29 g3. In an over-the- those available; he merely judges
board game, a player with such a that it gives the most chances in
big advantage would hardly have practice" (Beniantin Blumenfeld).
any reason to go in for complica- You may consciously select a line
tions whose outcome he could not which you know is not strongest,
predict, however hard he tried. just so as to give the game a char-
Sanakoev's play was wholly acter that suits you and is unwel-
true to his style - he usually fa- come to your opponent. The
vours a' tactical solution to his question here is merely how far
 Thoughts about a Book 139

this psychological approach can would forfeit most of his advan-

legitimately go, where the line tage" (Sanakoev).
should be drawn. Letustrytorefinethisvariation
An analysis of concrete exam- by playing first 18 'ii'xc5 l:txc5,
ples to trace how a player's style and only then 19 cxb4 l:r.c220 'iii>f'2.
influences ~s decisions would be (20 ..i.d3 l:lxb2 21 J:tdbl l%d2),
very interesting and useful. Unfor- when after 20 ...l::txb2 21 l:tdbl
tunately, to my knowledge, no one :Ubl 22 %bbl rJ/;d723 l:'lcl White
has yet carried out such an investi- has, this time, occupied the c-file
gation. Some speculative attempts and moved his king nearer the cen-
have been made to establish a clas- tre; his advantage is quite substan-
sification of styles, but that is all. tial. However, Black can improve
his defence by sacrificing a pawn
with 20 ...c.te?!, so as to retain con-
trol of the open file and White's sec-
ond rank. For example, 21 l::tdc1? !
l%hc8 22 l%xc2 .:xc2 23 b3 d4! 24
ltd 1 l:ba2 (now we understand
why the black king didn't go to d7)
25 .Uxd4 ..i.xg2.
A dangerous plan was suggested
by Grandmaster Stefan Kinder-
mann, also starting 18 'iVxc5 l:.xc5,
but then deviating with 19 J:tacl !?
w bxc3 20 b4! l:r.c721 .:d3:
a) In the bishop endgame that
Sanakoev - Ljungdahl arises from 21...~d7 22 1:.dxc3
6th World Championship l:thc8 23 l:txc7+ l:txc7 24 l:txc7+
1968-71 '3iixc7,Black has a difficult defence
ahead of him. His only hope,
White stands better, of course. 25 ...<.t>b6(followed by 26 ....tc6 or
The only question is how he can 26 ...a5), is easily dashed by 25 a4!
extract the maximum from the po- and then ·a5 and cJ.>f2-e3-d4-c5.
sition.· The pawn ending after 25 ....tc6 26
"After 18 cxb4 ?! li'xd4+ 19 a5 ..i.b5 27 .txb5 axb5 28 <it>f2is
l:r.xd4l:r.c2 20 .i.d3 l:.xb2 21 lie 1 lost for Black (as readers may ver-
~d7 22 l:r.c2l:.xc2 23 ..i.xc2 White ify for themselves).
 140 Thoughts about a Book

b) Black does better to refrain

from the rook exchange and play
21...d4!? 22 l::txd4 @e7. It is then
unclear how significant White's
advantage is.
18 Jhbl!!
A 'mysterious rook move', to
use Nimzowitsch's phrase. Sana-
koev writes, ,.Such a continuation
is more difficult to find than a
forced combination involving sev-
eral piece sacrifices." He is right,
although it seems to me that the
main difficulty lies not in finding which promises Black real counter-
White's move but in assessing its chances even in the (highly likely)
consequences. event of his losing the a6-pawn.
18 ... 'ilfxd4+ No, Sanakoev's decision does not
I 8 ...bxc3 I 9 bxc3 'ii'xd4+ is less look convincing - it is somehow
precise, because White can choose strategically suspect!
between 20 :Xd4 (as in the game) This verdict can be amended if
and 20 cxd4. Black cannot avoid only we continue our study of the
opening the b-file; 18...a5? is- en- position and discover White's plan
tirely bad: 19 'ii'xc5 lhc5 20 cxb4 here. In itself, the plan is nothing
axb421 ll.d4!?or21 l1bcl!. unexpected; Nimzowitsch, after all,
19 :Xd4 bxc3 stressed that "breaking into the op-
20 bxc3 ltc7 ponent's camp, usually on the sev-
21 .:r.db4 il.c8 (D) enth or eighth rank, is the logical
On arriving at this position in culmination of effective manoeu-
his analysis, an over-the-board vring in an open.file". However, it
player would most probably cut is quite impossible to see in ad-
short his calculations and reject the vance how dangerous the doubling
plan beginning with 18 .:tabl (this of rooks on the eighth rankwill be.
is what Zviagintsev did). Indeed, For that, we need a detailed analy-
what has White achieved? He has sis, which is not easy to perform
seized the b-file, but there is noth- even in a postal game. Playing
ing there to attack. White's queen- over-the-board with limited think-
side pawns have been weakened, ing time, it is not worth even trying
 Thoughts about a Book 141

to work out the variations accu- settle for an exchange of bishops

rately. You have to put your faith in that favours his opponent, or else
your intuition. It is interesting to give up the c-pawn.
see what it tells you here. 23 ... .l:.e8
22 l:.b8! Let us look at 23 ...:.Xc3:
Why doesn't White defend his a) The obvious reply is 24
c3-pawn? Surely because he wants Jlxa6, but after 24 ...l:t.e8 25 l:ta7+
to give his opponent no time for (25 l:tbb8 Jld7) 25 ... <3;;f826 Jlb5
the following deployment of his (26 Jlxc8 .l:.exc8 27 l:t.bb7 <3;;g8! 28
forces: 22 l:.lb3 <i;;e,723 l:tb8 (23 Il..xf7 .l:.cl + 29 <3;;f2 l:t.lc2+ 30
<i;;f2is better) 23 ...l:td8 24 l:ta8 d4 !. ~g3? ! lt8c3+ 31 ~h4 :.Xg2 gives
The tempting 22 c4 would work White nothing) 26 ...l:td8 the posi-
in the case of 22 ... dxc4 23 l:t.xc4 tion is not entirely clear. Black's
l:t.xc4 24 .i.xc4 @d7 25 l:tb8 l:td8 pieces are tied down as before
26 l:t.a8 Jlb7 27 l:t.a7 and 28 Jlxa6. (his bishop doesn't have a single
However, exchanging on c4 is not move), but the advance of his d-
obligatory~ Black has the stronger pawn, and a kingside break with
22 ... <3;;e7!23 cxd5 exd5. ... g5 where suitable, promise him
22 ... <3;;e7(D) distinct counter-chances.
b) A stronger line is 24 l:t.a7+!
cJ;;d8 (24 ... Jld7 25 l:t.bb7 .l:.d8 26
.i.xa6 @e8 27 lhd7 :.Xd7 28 Jlb5)
25 .l:.b8! (25 l:txf7? l:tc7) 25 ...l:tc7
26 l:r.aa8 (threatening a permanent
pin on the eighth rank after 27
~xa6) 26 ... a5 27 :.Xa5 and then
27 ...l:tb7 28 l:tba8 or 27 ...l:tc 1+ 28
@f2 l:tc2 29 ~el (not 29 l:taa8?
cJ;;c7 30 cJ;;el l:be2+ 31 <3;;xe2
~a6+) 29 ...';J;;c7(the threat was 30
l:t.a7 or 30 l:t.aa8 followed by 31
w ~a6) 30 .l:.b3 and 31 l:ta7+. It
would seem that in this variation
23 l:.a8! Black cannot break free.
Sanakoev consistently pursues 24 l:t.b3 f6
his_plan. After 23 Jlxa6?! l:t.d8(or If 24 ....i.d7, then 25 J:txa6 l:tec8
23.: ..l:.e8) he would either have to 26 l:taa3. This is stronger than 25
142 Thoughts about a Book

l:.xe8+ .i.xe8 (25 .. .'~xe8) 26 .i.xa6 a-pawn, especially if the bishops

J:ta7 27 J:tb6 nc7, when 28 :t.b7 are exchanged in the process. Sec-
fails to 28 ...l:.xb7 29 .i.xb7 .i.b5! ondly, it is important to hinder the
followed by ...(j;;d7-c7. white king's advance to the centre
25 JTLd3! of the board. •
Not at once 25 ~f2? fxe5 26 Let us return to the position af-
fxe5 J:tf8+27 ~e3 l:.f5. ter 26 fxe5 (D ):
25 fxe5
26 fxeS h6
27 ~
By now White's advantage is
beyond dispute; he went on to ex-
ploit it convincingly:
27 ... J:tcS 28 'itie3 J:td8 29 'itid4
.:SS (29 ...lk7 30 J:tbb8) 30 llbb8
<it>d731 J:ta7+ ~c6 32 .i.xa6 J:ta4+
33 <it>e3 (33 «iii>d3?
.i.xa6+ 34 l:txa6+
lha6 35 Jhd8 l:txa2) 33 ... d4+ 34
cxd4 l:a3+ 35 «it>e4 .i.xa6 36
.:.Xa6+! l:.xa6 37 J:txd8.:.Xa2 38 B
l:td6+ «it>b5
39 .:.Xe6 lhg2 40 ~d5
1-0 Instead of the colourless 26 ...h6
Black went under without a it was worth trying 26 ...l:.fS!?, with
fight. What became of our impres- a view to 27 .i.xh7 d4. After 27 h3
sion that he could hope for coun- h6 (27 ... g6) 28 l:tbb8 (28 .t.xa6?
terplay, that White's decision was .i.xa6 29 l:txa6 l:tf5) 28 ...l:te8,
'strategically suspect'? Were those Black can defend successfully. It is
really just empty words? clear that White should, after all,
They were not; we were basing prefer 27 .t.xh7!? d4 28 c4! Jhc4
our evaluations on objective fac- 29 l:a7+ ~d8 30 h3.
tors in the position, and were there- Zviagintsev suggested 23 ... f6!?
fore right to suppose that Black (D) (instead of 23 ...l:.e8). Let me
had far more stubborn ways to de- show you some variations we dis-
fend himself. Here are some con- covered together.
siderations to make the search for a a) 24 l:tbb8 l:te8 25 .i.xa6 .i.d7
defensive plan easier. First, Black 26 l:txe8+ .i.xe8, and if 27 l:tc8,
should not be afraid of losing the then 27 ...l:ta.7.
 Thoughts about a Book 143

we ought to have considered first)

involves taking the c3-pawn in
some circumstances or other. Let
us go back to the position after
White's 22nd move.
In reply to 22 ...0-0!?, Sanakoev
gives 23 lh8, with the variation
23 ...Jld7 24 :bb8 ltxb8 (24 ...:Ccs
25 l:txc8 ilxc8 26 ~f2!) 25 l:txb8+
:cs, and now not 26 :b7? .abs!
27 ~xb5 axb5 28 .:xb5 h5 29 l:tb3
l%c4 30 g3 .:a4 with counterplay,
but simply 26 l:r.xc8+! .axc8 27
b) 24 .:b3 fxe5 25 fxe5 :lf8 !. <li'f2, and the penetration of the
The rook restricts the mobility of white king decides the game. How-
the king, with the additional threat ever, Black can play 23 .. J:txc3 !?
of 26 ....J:.fi. 24 :lbb8 l:tc6 (D).
c) 24 ~xa6 l:te8, and now:
cl) 25 Jlxc8 l:.exc8 26 lhc8
.:txc8 27 J:tb7+ ~f8 28 exf6 gxf6
29 lbh7 lhc3, and Black retains
genuine saving chances in the rook
ending.
c2) The same goes for the
variation 25 l:tb3 fxe5 26 fxe5
~xa6 27 lha6 .:tec8 28 l:tbb6
lhc3 29 lhe6+ ~f7.
c3) 25 Jid3 !? fxe5 26 fxe5 ~d7
(26 ...lhc3 27 J:ta7+ is dangerous
for Black) 27 l:xe8+ (it is also
worth considering 27 J:ta3l:ec8 28
l:bb3) 27 ...~xe8 28 l:b3. White's How do we assess this position?
advantage here is considerable, al- Black keeps his extra pawn and is
though· the result of the struggle in no immediate danger, but his
still remains unclear. forces are completely tied down.
,Another possible defensive ap- The attempt to untangle with ...:es
proach (which, strictly speaking, (aiming for ...~f8 and ...~d7) fails
 144 Thoughts about a Book

against .i.xa6. Black must play 24 l:a7+! - they are decidedly in

... g6 and ... 9i/g7, and then be con- White's favour. The best defence is
tent with waiting tactics. The ques- 23_.0-0! 24 l:.bb8 :C6. This brings
tion (to which I have yet to find a us once again to the position in the
clear answer) is whether White's last diagram. The objective verdict
resources are adequate to breach on White's whole idea starting
his opponent's defence. with 18 :abl evidently depends on
However, a general point is that the assessment of that position.
once Black has castled, White is Let us state our conclusions. The
not obliged to sacrifice his c-pawn; complicated (and surely not infal-
he can keep the advantage with 23 lible) analysis we have just per-
.:t.Ib3!?. It therefore makes sense formed illustrates once again that
for Black to take the pawn a move even the most difficult-seeming po-
earlier. sitions may be viable. Nevertheless
Let us look at 22 .. Jhc3!?. The it does not cast doubt on the bril-
reply 23 .i.xa6 seems obvious, on liant decision taken by Sanakoev
the grounds that after 23 ... ~d7 24 on his 18th move. The defence, af-
:Ia8 l:d8 25 l1bb8 the permanent ter all, is extremely hard, White al-
pin on the eighth rank will guaran- ways retains chances of success,
tee White a decisive plus. (He will and in any case we did not see any-
bring his king to the centre and thing more convincing for him.
then advance his passed a-pawn.)
If Black tries to untie himself with Realizing an advantage
23 ... ~e7 (with theideaof24 ....:r.e8
and 25 ....i.d7), he comes up against In examining the last two exam-
a tactical stroke pointed out by ples we _havealready broached this
Yusupov: 24 l:[lb7+!!, which wins subject, which is of immense im-
a piece after 24 ... .i.xb7 25 l1xb7+ portance to every chess player. So
and 26 l:[bS+. However, Black has is another which is closely linked
a saving line in 23 ...0-0!, and if 24 to it - finding defensive resources
.i.b5 then 24 ...l:a3, attacking the in difficult positions ..
a2-pawn and preparing to bring his Referring to the diagram over-
bishop out to a6. leaf, Sanakoev writes: "The criti-
23 l:ta8! is an improvement. We cal position; essentially, it is the
have ,already discussed the situa- play from this position that makes
tions arising from 23 ... ~d7 24 the game notable. By means of a
.i.xa6 l:td8 25 l:bb8 and 23 ... cite7 cunning regrouping, Black has set
 Thoughts about a Book 145

Not 43 ... c5 44 il.d3 l:th6 45

d6+! ~d8 46 il.xh7 with a won
rook endgame.
44 il.xd5
45 il.f3!
46 Wgl!
The point of White's play be-
comes clear - the rook is trapped.
46 ... il.f5
46 ... l:tg4 puts up more fight. al-
though after 47 il.xg4 hxg4 48 b4
f3 49 b5 .i.e4 50 b6 White should
win all the same.
Sanakoev - Engel 47 l:td4
10th World Championship It is important to stop 47 ... il.g4.
1978-84 47 ... ~e6
47 ...il.d7!? 48 b4 ~e6 is a shade
'
his opponent a concrete problem - more accurate, but this too prom-
what to do about the g- and h- ises no saving chances. White con-
pawns. The black bishop may be tinues 49 l:td 1! .i.b5 50 :d8 and 51
shut out for the time being. but will l:tb8.
White's passed pawn on the queen- 48 b4 ~e5
side outweigh his material losses 49 .l:.d5+ ~e6
on the kingside? After all, the 50 b5 il.g4
black pieces may regain their ac- 51 b6! il.xf3
tivity with an eventual ... f4." 52 gxf3 1-0
First let us see how the game Black resigned on account of
ended (with notes based on Sana- 52 ...~xd5 53 b7 l1h3 54 b8'if.lhf3
koev•s judgements and some of his 55 'ii'b7+, or 52 ... .l:.h3 53 l:t.d8!
variations). (but not 53 b7? l:tg3+ 54 ~h2 {or
42 dS! f4+! 54 @:f2} 54 ...l1g8 followed by
43 ~ ..Jlb8 and ... ~d6-c6).
On 43 ~d4 ?! l:tg4 ! 44 ~c5 cxd5 Zviagintsev rejected 42 d5 due
4 5 il.xd5 l:txh4 46 b4 l:th 1 Black to the reply 42 ...c5 l. He was doubt-
retains good chances of saving the ful about the bishop ending that
g~e. arises from 43 d6+! (we will take
43 ... cxd5 Sanakoev's word for it when he
146 Thoughts about a Book

says that "after other moves Black save himself with 46 ... f4+!! 47
is out of danger", even though 43 gxf4+ ~ 48 .i.c4 (48 Jtxh5 .i.c2)
l:ta2 deserves examination). Sana- 48 ....i.c2 (or 48 ....i.g6). As you
koev analysed this endgame and can verify without difficulty, White
considered it won. Here are his is unable to exploit his extra pawn.
variations after 43 ...lhd6 44 lhd6 He is hampered by having his b-
~xd6 45 ~xf7 ~e5 46 g3 (D ). pawn on a light square - the same
colour as his bishop - and if his
king comes over to c3, Black will
have the chance to counter-attack
on the other wing.
There is also a slightly different
drawing line: 45 ...f4+! (in place of
45 ...We5) 46 ~xf4 We7 (alterna-
tively, 46 ....i.c2 at once) 47 .i.xh5
(47 .i.c4 ~f6) 47 ....i.c2.
Zviagintsev chose another plan
for realizing White's advantage,
though he based it on the same
idea - of trapping the black rook -
which Sanakoev carried out in the
46 ...~f6 47 .i.xh5 .i.g8 48' .i.dl actual game.
~e5 49 h5 and now: 42 ~f2! l:tg4 43 .i.e2 l:txh4 44
a) 49 ...~h7 50 ~c2 .i.g8 51 h6 .i.f3 ~d6 (Black can't face losing
.i.h7 52 g4 f4+ 53 ~d2 .i.xc2 54 the c-pawn) 45 b4! (intending 46
~xc2. J:.dl and 47 ~g3; the immediate
b) 49 ....i.f7 50 h6 ~f6 51 ~f4 45 l:tdl is imprecise due to 45 ...c5 !)
Wg6 52 .i.h5+ !. 45 ... f4 (45 ...J:.hl 46 l:tc2 l:tbl 47
c) 49 ....i.e6 50 .i.c2 .i.f7 51 h6 lhc6+ ~d7 48 J:.b6) 46 d5! (oth-
~f6 52 ~f4 ~g6 (52 ....i.g8 53 erwise Black brings his bishop
.i.xf5 .i.xb3 54 g4 c4 55 g5+) 53 round to the defence) 46 ...cxdS 47
g4 ~xh6 54 gxf5 ! .i.g8 55 f6 .i.f7 l:txd5+ q;c7 48 J:.bS! .i.g6, and
56 ~e5 ~g5 57 .i.f5 .i.xb3 58 only now, finally, 49 ~gl! with
.i.e6 and the f-pawn queens. good chances of victory.
For all this, Vadim's intuition
hadnot failed him. From the posi- I hope you are convinced by
tion in the last diagram, Black can now that Sanakoev's book is rich
 Thoughts about a Book 147

in content and offers us abundant In the event of 33 lb.a5? l:lf3 34

food for thought. I have dwelt on a l:.h5 l:lxg3 35 lbh7+ r:Ji;g8,Black's
few episodes only (a further ex- passed g-pawn ensures him ade-
ample of the author's play is exam- quate counter-chances. Ilia Smirin
ined in Chapter 13, 'Virtuoso De- played 33 .i.e4?!, offering a bish-
fence'), but of course there are op exchange, which looks like a
many more games in the book, and perfectly sensible solution to the
in all of them the reader is sure to problem White faces.
discover something interesting and In the game, this decision paid
useful. off: 33 ... .i.xa4? 34 lba5 .i.e8 (or
Postscript: I would like to take 34 ....i.c6 35 .i.xc6 bxc6 36 l:r.g5)
the opportunity to make some cor- 35 .i.xb7 l:fl + 36 <t>d2.i.g6 37 c4
rections to some analysis of my l:12+ 38 <t>c3<t>g739 l:.g5 !, and in
own, published in Technique for this hopeless position Black lost
the Tournament Player (Dvoretsky on time.
and Yusupov). This will involve us How can the defence be im-
with the same problems as before: proved? The general view is that, in
realizing an advantage, and look- such situations, going into a rook
ing for defensive resources in a dif- ending offers the best chances of
ficult position. salvation (everyone knows the say-
ing 'all rook endings are drawn').
Yet after 33 ....i.xe4 34 J:.xe4 l:g8?
(34 ... h5? 35 l:e5) 35 '.t.>d2!?,with
the idea of ~e3-f4, Black is con-
demned to total passivity and should
definitely lose.
In rook endgames it is always
essential to try to activate your
rook. After 33 ...llil+ 34 'it>d2one
idea is 34 ....i.xe4!? 35 lhe4 lli2+
and then 36 <t>d3l:.f3+ 3 7 l:e3 :t2
38 b3 ~g7, or 36 'iii>dlh5! 37 l:e5
w l:.h2 38 l:xa5 ri;;g7 (with ... ~f6
and ...h4 to follow), Black obtains
Smirin-Vogt counterplay which is probably
Saltsjobaden 198819 adequate to draw. Another possi-
bility is 34 ...l:gl 35 .i.xc6 bxc6 36
148 Thoughts about a Book

l:.xa5 l:.g2+ 37 <itid3llxg3+ 38 <itie4 out my error (in 64 - Chess Review

l:C.g2(but not 38 ... :!g 1 39 lt.c5 g3 No. 12, 1996). Indeed I have not
40 'ili'f3 g2 41 b3!) 39 l:lc5 lU2! 40 managed to find a win here. In the
a5 g3 41 a6 :lf8 42 .l:tg5 l:ta8 43 last variation, for example, White
lhg3 l:txa6. might try 40 c6 .ilf3 41 'it>c4h5 42
Black may also exchange rooks .ila8 'it>e7 43 'it>b5, counting on
and defend the seemingly difficult 43 ...'it>d6(d8)? 44 'it>b6h4 45 c7 or
position with bishops of the same 43 ...h4 44 gxh4 g3? 45 c7(+)!
colour: 33 ... lleS! 34 lbe8+ .ilxe8 (now we see why the king avoided
(D). the squares d4 and c5 - after 45 ...g2
the black pawn will not queen with
check). Yet after 43 ...h4! 44 gxh4,
Black first plays 44 ...'it>d6!-forc-
ing the white king to b6, where it is
exposed to check - and only then
45 ...g3.
Instead, the grandmaster pro-
posed an ingenious breakthrough
plan: 35 b4!? b6 (Black loses at
once with 35 ... axb4? 36 a5, or
35 ....ilxa4? 36 bxa5 followed by 37
.ilxb7) 36 bS .ilf7! (Black has to
w hold up White's main threat of c4-
c5) 37 'it>d2'it>g738 'it>d3.At this
I confess I analysed this posi- point the only continuation Zaitsev
tion carelessly. I did see that after gives-is 38 ... h6 39 c4 'it>f6 40 c5
35 .ilxb7 .ilxa4 36 c4 Black avoids <itie541 cxb6 ~d6 42 ~e3 .ilb3 43
immediate defeat by means of .ilc6 .ilxa4 44 b7 'it>c745 b6+ ~b8
36 ....ilb3 37 c5 <Ji;g7(and if 38 c6, 46 .ilxa4. Here, though, I disagree
then 38 ... .ildS! 39 .ila8 .ile6!). with him - Black's resources are
However, I was either relying on not yet exhausted. Quite apart
38 .ilc8, and underestim~ting the from the plausible try 38 ... <itif6,
pawn sacrifice 38 ... <Ji;f6!39 .ilxg4 sacrificing the h-pawn to bring his
<itie5- or else I didn't notice that king to the queenside and hinder
aftet 38 •<attd2<attf639 ~c3 Black White's c4-c5, Black can play
has 39 .._..ildl ! with· the idea of more accurately within the terms
....ilf3. Igor Zaitsev rightly pointed of the plan Zaitsev examines.
 Thoughts about a Book 149

38 ... hS 39 c4 @f6 40 cS @e7! The most important thing is to

(but on no account 40 ...@eS? - the forestall 33 ...l:r.f3.
king's place is not in the centre but 33 ... 3'.xa4
in front of the white pawns) 41 34 llxaS!
cxb6 @d8! (D) An unexpected switch of fronts
- Smirin only considered 34 lbg4
3'.c6, with a probable draw.
34 ... 3'.c6
A tougher defence appears to be
34 ... 3'.e8, trying to get in ... h5 at
the first convenient moment.
35 l:tgS!
This rook is like a bothersome
fly.
35 ... 3'.f3
35 ...l:tg8 36 l:th5.
36 l:thS :.C7
w 37 'it>d2
Having skilfully tied down his
How is White to win now? He opponent's pieces, White simply
doesn't succeed in queening a intends to strengthen his position
pawn with 42 3'.f5 3'.b3 43 'it>d4 by advancing his queenside pawns .
.b.a4 44 'it>c53'.d 1 45 'itid6 3'.f3 - This kind of plan for realizing an
the bishop arrives in the nick of advantage (domination with no
time. Moreover, the position after counterplay for the opponent) is
42 'it>e3 3'.b3 (42 ... 'it>c8) 43 3'.c6 wholly in the style of Anatoly Kar-
'it>c8 44 b7+ 'ilr>b845 b6 3'.dl is pov. For an over-the-board player,
drawn. a very important point is that
It turns out that Black has more hardly anything has to be calcu-
than one way of saving himself. lated here (in contrast to the line
This casts doubt on the plan begin- with 33 .i.e4, where we had to im-
ning with 33 3'.e4, and compels us merse ourselves in both the bishop
to look for other ideas. ending and the rook ending). This
_ The· strongest line was sug- means there is far less likelihood
gested by Grandmaster Viktor Bo- of errors, such as those committed
logan. by Zaitsev and me in our analysis
33 l:tgS! of the bishop ending.
150 Thoughts about a Book

Incidentally, in concluding that two paths to victory", he once

there is only one solution, we are wrote. Another experienced ana-
wholly concurring with Zaitsev's lyst, Gavriil Veresov, expressed
views ... Many years of experience the same thought: ..In positions
in chess analysis have convinced close to the borderline between a
me that in tense positions balanced draw and defeat, we generally en-
on a knife-edge, there cannot be counter a unique solution."
 9 Missed Brilliancy Prizes
Artur Vusupov

Mark Dvoretsky's suggestion that ruining masterpieces) to observe

I should write about some spoilt my numerous blunders. At any rate,
'masterpieces' could not have been I can now explain them by saying
more to the point. that I was collecting them to put in
In the first place, there are some a book.
old games that I have wanted to I have a self-interested motive
analyse closely for a long time. too. I am afraid that some tourna-
With the passing of time the an- ment organizers (not without acer-
noyance at missing the wins has tain justification, alas!) may have
subsided, and perhaps I can now formed an unfavourable impres-
look at these games more objec- sion of my chess style. They may
tively and critically. consider my play too solid and
Secondly, I actually happen to boring (unfortunately there is more
be a leading expert in this field (I truth in the latter). I want to turn
mean an expert in spoiling master- this image into a positive one: "He
pieces, rather than in critically ex- plays badly but interestingly."
amining my own games). In the And finally, my woeful experi-
course of my career I have suc- ences may prove useful to others,
ceeded in creating some games although I have to admit that I per-
which I can be proud of even now sonally have learned little even
- they have stood up to the strict from my own mistakes.
test of time and analysis. Yet for
every such game there are ten oth- Yusupov - Rebel 8
ers which were played beautifully 'Action Chess' (30 min. per game)
up to a point. then hopelessly ruined. match (13), Ischia 1997
Thirdly, I can imagine what Queen's Pawn Game
'pleasure' it will give my coach and
collaborator (my collaborator in 1 ltif3 ltif6 2 d4 e6 3 e3 c5 4 Jld3
writing this book, that is - not in b6 5 b3 .i.e7 6 .i.b2 0-0 7 0-0 dS 8
152 Missed Brilliancy Prizes

lDe5 ll:lfd7 9 f4 t2)xes 10 dxe5 This natural move cost me ten

.i.a6 11 c4 ll)c6 12 a3 dxc4 13 precious minutes and turned out to
bxc4 f5 14 exf6 .i.xf6 15 ll)cJ be an outright error. The immedi-
lDaS 16 'ii'c2 (D) ate l 8 :tf3 ! was better. I was afraid
of 18 ... 'ifd2 - for no good reason,
since the simple 19 'ii'xd2 ll)xd2
20 J:th3 is adequate. If instead
18 ...lDxb2 19 l:th3 .i.h4 (alterna-
tively 19 ... lDc4 20 'ife2) 20 Jle4
Jld3 21 'ifxb2 Jlxe4 22 ll)xe4
~g8 23 'ife5, White has a fierce at-
tack.
18 ... 'fie7? (D)
The correct reply was 18...ll)xe3 !.
If 19 .:.Xd8,then 19 ...l:.axd8 is not
bad: 20 'ifg6 .i.d3 21 'ir'xd3 l:.xd3
22 .i.xd3 Jld4!?, while 19... QJxc2!
20 :C.xa8:Xa8 21 Jlxc2 Jld4+ 22
My most recent example of a l:.f2l:.d8 ! is an even stronger line. I
misplayed attack occurred in a dem- had of course overlooked the last
onstration match against a com- move of this variation.
puter program. In this position
Black took a decision typical of a
computer:
16 ... lDxc4
After the 'human' continuation
16 ... h6 17 :..adl 'iie7 18 lDb5
Jlxb2 19 'Wxb2, White would gain
a clear plus, since Black's castled
position would be weakened and
his knight would be out of play on
a5. Now, however, White has to cal-
culate some variations that would
prove a difficult task for me even w
with a normal time-limit.
17 bh7+ liti>h8 19 lU3 t2)xb2
18 .l:adl 20 l:.h3 ll:lxdl
 Missed Brilliancy Prizes 153

If a computer sees the chance to intermediate check 24 .i.h7+!, fol-

devour something without detect- lowed after 24 ... <;t;>hs(24 .. .'itif7 25
ing a concrete reason not to, it will fr'h5+) by 25 fxg5:
do so without fail. A human player a) 25 ... ~xgS 26 .i.g6+ 'itig8
knows that gluttony leads to no (26 ... 'ifh6 27 l:.xh6+ gxh6 28
good, but the concrete proof of this 'ii'al+!) 27 lt'le7+! 'iiixe7 28 l:th8+
principle lies beyond the com- ..t>xh8 29 'iih5+ @g8 30 'iJih7#.
puter's horizon. However, being in b) 25 ...l:tfl+ 26 W°xfl .i.xfl 27
time-trouble, White failed to find ~f5+ 'it?g8 28 ~xe6+ <;t;>fs29
the win and settled for a repetition l:th8#.
of moves. The correct line was dis- I don't know which of these
covered the following day by some mates is prettier, but it's a pity I
chess enthusiasts who were inter- didn't manage to end the game fit-
ested in analysing White's attack. tingly with one or the other. In-
21 Ji.g6+ @g8 (D) stead, there followed:
22 1'.h7 + @h8
23 Ji.g6+ @g8
½-½

Yusupov - Xie Jun

Linares 1997
RuyLopez

1 e4 eS 2 ttlf3 lbc6 3 ~bS a6 4

1'.xc6 dxc6 5 0-0 'ii'd6 6 lt'la3 bS 7
c3 c5 8 ttlc2 f6?! (8 ...lbe7; 8 ...Ji.b7)
9 a4 Ji.b7 10 axbS ~xe4 11 d4
w cxd4 (D)
Black is behind in development
I was intending to bring the queen and White has attacking chances.
into the attack with 22 'ifxdl !, but Such trivialities as the loss of a
was somewhat discouraged when I pawn should not, of course, worry
noticed the defence 22 ...Ji.gS !. I him.
did realize that White's attacking 12 l:.el! ~b7
resources were not exhausted, and 13 cxd4 axbS
that I had 23 ttld5!. On 23 ... 'iWd8, 14 l:ua8+ Ji.xa8
however, what escaped me was the 15 'i¥e2 e4
154 Missed Brilliancy Prizes

not be underestimated! If she gets

the chance to complete her devel-
opment, it is White who will be
badly off. I felt that the critical mo-
ment in the struggle had arrived,
and immersed myself in the search
for the best solution.

w
16 'iVxbS+ .i.c6
17 'iiaS
White has already won his pawn
back and continues to develop his
initiative. 17 .:Xe4+?! is weaker in
view of 17.....tn.
17 t:f1e7
18 lt:ld2 rs
19 b3! I now made an amusing mistake
A good move, setting up the in analysis. I was very keen to dis-
threat of .i.a3 and preparing to patch my knight to e5, and quite
transfer a knight via c4 to the cen- quickly hit on the right method: 22
tral square e5. lt:lxf5 !! 'iVxf523 ltJe5 - only to dis-
19 'iVd7 card it, with much regret and just
20 ~c4! ~g6 about as quickly, on account of the
21 lt)2e3' .i.bS (D) simple 23 ...lDxe5 24 'ilxb5+ loo+.
In the event of the optimistic The fact that in this tempting varia-
21...f4, White had prepared the tion Black's king is jn check and
powerful rejoinder 22 ltJe5. In- her last move is therefore strictly
stead Black plays the natural and prohibited by the rules of chess,
necessary move. After this her was of course the point I over-
backward development assumes looked. The reason for such a curi-
dangerous proportions, yet on the ous error must lie in my indistinct
other hand her pawn centre should image of the chessboard while I
 Missed Brilliancy Prizes 155

was analysing~ a vital detail of the 22 fxg4

position was simply excluded from 23 'ifa8+ <3X7
my internal field of vision. 24 'ii'xe4 .i.b4! (D)
I believe the knight sacrifice
would have won. On 23 ...c6, White
has a pretty mate with 24 'fla8+
</;e7 25 .itg5+! 'ifxg5 26 'ii'b7+
@e6 27 'ii'd7+. A stiffer defence is
23 ....itd?, but White still has a de-
cisive attack after 24 'ii'a8+ .itc8
25 lhe4 tj)e? 26 .ita3.
It was also worth considering 22
h4!?, which virtually forces the re-
ply 22 ... h5. Then White has the
same sacrifice: 23 lbxf5! 'ifxf5 24
lbe5 .itd7 25 'it'a8+, although in
this case :EVackmay get the chance
to bring her rook into play via h6. Black makes use of the breath-
Thomas Wedberg suggested an ing space to finish her develop-
alternative knight move: 22 tj)e5 !. ment at top speed. White wins a
According to his analysis, after pawn, but his initiative evaporates
22 ... lQxe5 23 dxe5 c6 (23 ...~d3? completely.
24 tj)d5 !; 23 ... .itc5 24 'ii'a8+ 'i!i'd8 25 l:tdl .:t.e8
25 'ii'b7 with advantage) 24 J:.dl 26 'fixg4 'ii'xg4+
.itd3 25 .ila3 Jlxa3 (25 ...f 4 is bad Assessing the position correctly.
because of 26 tj)c4 'ii'g4 27 l:.al In the endgame Black's king will
~xc4 28 .ilxf8 .:.Xf8 29 'fie? ~a6 be safe, and her strong bishops will
30 e6!) 26 'ifa8+ 'ii'd8 27 'ii'xa3 fully compensate for the slight loss
J:.f8 28 'f/c5, White maintains an of material.
advantage. 27 tj)xg4 l:te4
Straying from the correct line of 28 h3 hS
attack, I was tempted to undermine 29 tj)geS+ tj)xeS
the c~ntre in a way that appears 30 tj)xeS+ <ate6
strategically attractive. We can now take stock of the
22 g4? situation. Black has obtained suffi-
This move looks stronger than it cient compensation for the pawn.
actually is. In view of his scattered position,
 156 Missed Brilliancy Prizes

White has no real chance to do The right way! By sacrificing a

more than share the point - which pawn I increase my lead in devel-
indeed is what occurred a little opment still more. The slow 14 g3
later. would allow Black to obtain a fully
acceptable position after 14...il.d7
Yusupov - Ivanchuk 15 f4 .i.xc3 16 lhc3 .li.c6, e.g.: 17
Tal Memorial Tournament, e5 4Jd5 18 .i.xd5 exd5 19 .i.e3 d4
Riga 1995 20 .i.xd4 W/d7 with counterplay.
Queen's Gambit Accepted 14 .li.xf4
15 .i.xf4 °iVxf4
1 d4 d5 2 l2Jf3e6 3 c4 dxc4 4 e3 a6 16 eS! 4Jd7 (D)
5 .i.xc4 cS 6 ilb3 lbf6 7 0-0 lbc6 8 The point of the pawn sacrifice
'ii'e2 cxd4 9 l:.dl d3 (9 ... .i.e7) 10 is that the active 16... lLJg4? is met
lhd3 'flic7 11 lbc3 ild6?? by the simple 17 g3, and if then
1l.. ..i.c5 may be better. 17...'it'f5 White has either 18 .li.c2
12 e4 lbes or 18 :t.fl, when his attack pro-
13 ltJxeS he5 (D) ceeds unopposed.

w w
In his attempt to gain control of 11 mu
the b8-h2 diagonal, my opponent Again White adopts the most
has rather fallen behind in devel- energetic solution. Of course, he
opment. White must of course take could have maintained the attack
the initiative at once. without additional sacrifices: 17
14 f4? l:te3!? 0-0 18 :r1 1i'd4 19 'iti>hl.In
 Missed Brilliancy Prizes 157

that case, however, the active black didn't have enough thinking time.
queen in the middle of the board I ought to have had more faith in
would have been a serious hin- my own powers, but just try main-
drance to White's offensive. After taining your confidence and sang-
the move played, White drives the froid when one of the world's
queen out of the centre; the loss of strongest players is sitting oppo-
the e5-pawn is compensated for by site you, quickly answering your
the gain of time and the opening of moves with an imperturbable air!
lines. 21 ... tl)e5
Instead 17 l:tad 1 would be a car- In answer to 21. ..l:td8, Ljubomir
dinal error, allowing Black to bail Ftacnik suggested 22 t2Jf6+ ! gxf6
out with a small sacrifice and beat (22 ...c;1;>h8
23 ~h5 h6 24 t2Jg4) 23
off the attack: 17 ... 0-0 ! 18 l:txd7 'ii'g4+ c;1;>f8(23 ...Wh8 24 l:tg3) 24
it.xd7 19 l:txd7 l:tad8. 'ii'b4+, and if 24 ...'ii'cS (but not
17 ... 'fi'xeS 24 ...c;1;>e8
25 l:tg3), the simple 25
18 l:te3 'ii'd4 'ii'xc5+ t2Jxc5 26 l:txd8+ <;1;,e727
Natural}.y, 18 .. .'iVc5 is weaker l:th8 gives White a large plus in the
because of 19 t2Je4. Pinning the ending.
rook is Black's best chance. He 22 l:th3 t2Jg6 (D)
would lose at once with 18 ...'iVd6? 22 ...h6 would be weaker in view
(18 ...'fic7 is answered by the same of 23 l:E.xh6gxh6 24 t2Jf6+, shatter-
stroke) 19 l:txf7 cj;xf7 20 J:E.xe6 ing Black's castled position. Black
'ft'd4+ 21 cj;hl ltk5 221ff3+. brings the knight closer to his king,
19 l:tdl 'i/ia7 but White has already concen-
Not 19 ... 1fb6 on account of 20 trated nearly all his forces for the
.i.a4. attack.
20 tl)e4 0-0 23 'iihs
21 cj;bl White seeks a rational continua-
White is doing everything right tion suited to his limited time for
but expending too much effort and calculation. He could already have
time on it. The only reason why I launched the decisive onslaught
couldn't conduct the game to its with 23 l::txh7!?.According to Ser-
logical conclusion was that I didn't gei Dolmatov's analysis, Black
trust my judgement and was trying loses if he accepts the sacrifice at
to calculate the_variations more or once:
les,s\ exhaustively. The result was a) 23 ...c;1;>xh724 'ii'h5+ Wg8 25
that at the critical stage I simply tlJg5 lle8 26 :n and now:
 158 Missed Brilliancy Prizes

23 ... h6
At this moment I only had five
minutes left on the clock. I saw that
my planned combination guaran-
teed perpetual check, and hoped
for the best.
24 tl'lf6+?
Directly after the game, much to
my annoyance, I easily found a de-
cisive strengthening of the attack. I
should have brought my bishop
into the fight with 24 .itc2!, after
which it turns out that Black is
al) 26 ...b5 27 l:txt7 'ifxt7 28 helpless against the threat of 25
lbxn ~ 291i'f3+ l2Jr430 Wxa8. 'l:lf6+. He only has the unpleasant
a2) 26 ...lte7 27 'ii'h7+ ~f8 28 choice between:
'ii'xg6 b5 291i'h7 and wins (then a) 24 ...eS 25 lof6+ 'if.>h8(or
29 .. .'if.>eSis answered by 30 l:ldl). 25 ... gxf6 26 l:tg3 fS 27 l:txg6+
a3) 26 ...llle5 27 lht7 lllxf7 28 fxg6 28 'ii'xg6+ 'if->h829 'iVxh6+
'iWxf7+'iti>b829 'iVxeS#. ~g8 30 .itb3+) 26 l:lg3 and now:
What confused me during the al) 26 ....ite6 27 :Xg6 fxg6 28
game was that Black also has an:. 'ifxg6 .i.g8 29 lbxg8 (29 .l:r.d7!)
other defensive possibility, but in 29 ...'iti>xg830 'ii'h7+ rJilt731 Jlb3+
that case too White can find a win: ~f632l:lfl+.
b) 23 .. ,q)f4 24 1i'g4 ~xh7 25 a2) 26 ... e4 27 .i.xe4 (27 :txg6
lbg5+, and now: fxg6 28 'iVxg6 .i.f5!) 27 ... lllf4 28
bl) 25 .. .'ith6 26 1i'xf4 f6 (or 'ikxf7 .i.g4 29 'ife7.
26 ... rJilg627 1i'g3) 27 'ifh4+ rJilg6 a3) 26 ... lllf4 27 'iixe5 log6 28
281i'h7+ rJilxg529 Wxg7+ 'if.>h530 Jlxg6 fxg6 29 lhg6 b5 30 lhh6+
:td5+ and mates. gxh6 31 :td7.
b2) 25 ... 'iti>gS26 1i'h4 lte8 27 b) 24 ... b6 25 lllf6+.gxf6 (or
11i'h7+ ""f8 28 1i'h8+ rJile7 29 25 .. .c.th8 26 'ii'g5) 26 'ii'xh6 :te8
iixg7 :tf8 30 tbh7 .i.d7 31 ti'f6+ 27 .itxg6.
~e8 32 wt'xf4with a won position. c) 24 ... lllf4 25 lllf6+ 'ith8 26
Howev~r. the natural move I j//xt7 ! Ji.d7 27 l:txh6+.
played is an equally good alterna- In all these variations the black
tive. king is unable to escape a quick
 Missed Brilliancy Prizes 159

mate. My premature combination, then after 27 l%fl ! 'ir'd4 28 l:txg6+

by contrast, allows Ivanchuk to fxg6 29 'iixg6+ ~f8 30 1i'h6+
maintain the balance. White can force perpetual check at
24 ... gxf6 will, as 30 ...~e7 fails to 31 'ir'g7+
24 ...<iii>hS?
would lose to 25 'ir'g5! ~d8 32 l:tdl.
e5 26 l:txh6+ gxh6 27 'ir'xh6#. Now that Black has countered
25 'ir'xh6 l:te8 the threat of h2-h4, White has
26 l:g3 (D) nothing better than to settle for a
Here Black can answer 26 .i.c2 draw.
with 26 ... f5, leading to perpetual 27 l:txg6+ fxg6
check after 27 'fi'h7+ ~f8 28 28 'ifxg6+ ~8
'ir'h6+. 29 'ii'h6+ <M7
30 'ii'h7+ ~f8
31 'it'h8+ 1/2- 1/2

Yusupov - Hilbner
Tilburg 1987
Slav Defence

1 d4 d5 2 ltJf3 ltJf6 3 c4 dxc4 4

ltJc3 c6 5 a4 .i.f5 6 e3 e6 7 .i.xc4
.i.b4 8 0-0 ltJbd7 9 ltJh4 .i.g6 10
ltJxg6 hxg6 11 f4
11 h3.
1I-.ltJd5
ll...'it'a5.
26 ... •a! 12 .i.d2
By an irony of fate it is Black's 12 ltJe4!?.
queen that intervenes to save him. 12 ..Ji'aS 13 'WetltJSb6 14 .i.b3
(We recall White's exertions to shut c5 15 l:tdl cxd4 16 ltJb5 .i.xd2 17
this key piece out of the game.) l:txd2 ltJc518 .i.c2 l:td8 19 exd4?!
Black might try playing for a win, (D)
banking~ on his opponent's time- A more solid line is 19 l:txd4
trouble, but the attempt would 1i'xel 20 l:txel, and if 20 ...l:txd4 21
most likely boomerang. Thus, exd4 locxa4, then 22 d5.
26 ...b6 27 J:td4! 'ilc7 28 l:tdg4 is no White has rather overestimated
good' at all for Black. If 26 ...b5, his chances, and now the cool
 160 Missed Brilliancy Prizes

21 ... <j;;xf7
The consequences of 21 ...'it'xb5
are unclear: 22 'fi'xe6+ ~f8 23
.i.g6 'Wd7 24 l:te2 fi'xe6 25 :xe6
~d5 26 l::lfel ~f6. I think that af-
ter 27 g3 !? or 27 rl,..e7White would
have enough compensation for the
piece sacrifice.
22 fS eS
The following defences are in-
adequate: 22 ...Ithe8 23 fxe6+ <ili>g8
24 e7. and 22 ...'it'xb5 23 'it'xe6+
@f8 24 f6 'ikd7 (or 24 .. JWxfl+ 25
19 ...0-0! might h~ve set him some @xfl l:te8 26 Wd6+ (j;f7 27 l:[f2),
serious problems. Instead, Hilbner which fails to 25 fxg7+ ~g7 26
falls in with my intentions. 'ikf6+ c;tig827 'Mi'g6+.
19 ... ~cxa4 23 f6
20 .i.xg6 The immediate 23 'it'xe5 was a
Virtually a forced move. Sharp serious alternative. The following
play suddenly flares up. variations do not, of course, ex-
20 ... @e7 haust all the possibilities in the po-
Accepting the sacrifice would sition, but they do show how strong
have lost: 20 ...fxg6 21 'ii'xe6+ ~f8 White's attack is:
22 f5. A more circumspect line, a) 23 ....It.he8 24 ~d6+.
however, was 20 ...0-0!? 21 .i.d3 b) 23 ... fi'xd2 24 f6! (24 'ii'e6+
a6, as the inviting 22 f5?! is met by ~f8 25 f6 l%.d7!)and now:
the unpleasant 22 ...l2JX.b2!(stronger bl) 24 ... g5 25 ~d6+! ltxd6
than 22 ...axb5 23 f6 or 22 ...exf5 23 (25 ...<it>g626 'ii'e4+ -.th6 27 ~f5+
.:Xf5). @g6 28 f7) 26 'it'e7+ <ili>g627
21 !Lxf7 'ii'g7+ @h5 28 'iixh8+.
It is too late for White to stop b2) 24 .. .'ili'g8 2~ fxg7 l:th6 (or
half-way (21 Jtd3 is no good in 25 ...ltxh2 26 'ike6+ 'itth7 27 gS'if+
view of 21...a6 22 f5 axb5 23 fxe6 :txg8 28 :tf7+ l:g7 29 'iff5+ ~h8
f6 24 'it'g3 'iltxd2). Both players 30 :tf8+ l:tg8 31 'ii'f6+ -.th7 32
are caught up in the surging cur- l:Ul+) 26 'it'e7+ and White wins.
rent and have no inkling where c) 23 ...~d524'We6+~f825f6
they will manage to surface. g6 26 .It.e2and now:
 Missed Brilliancy Prizes 161

cl) 26 ... 'ii'xb5 27 f7 'ii'xe2 (or

27 ...'illd7 28 'ti'e5 ! l:[b7 29 'i'i'e8+)
28 'ii'xe2 loab6 29 'fi'e5 with ad-
vantage.
c2) 26 ...'ii'M 27 f7 'ife7 28 'ifg4
'6'h4 (28 ... loe3 is answered by 29
'Wxg6 or 29 'fi'f4) 29 .:e8+! J:.xe8
30 fxe8'6'+ @xe8 31 ~d6+ 'std8
32 '6'c8+ @e7 33 l:[f7+ @xd6 34
'ild7#.
c3) 26 ...'ifb6 27 Wg4 <3;f7 28
'ii'g5! intending 29 .:r.e7+.
The continuation I chose is
probably just as good; in many 27 l:.e7+ loxe7
lines it merely transposes. 27 ...J:.xe7 28 fxe7+ lbf6 is no
23 ... g6 better, in view of 29 'ii'e5 with the
On 23 ...g5, White has either 24 decisive threat of 30 lbd6+.
dxe5 or 24 1fxe5 with a powerful 28 fxe7+ c;t>es
attack. 29 'ite5
24 1i'xe5 29 l:[f6 was probably an even
An interesting alternative is 24 stronger move.
dxe5!? 'ir'xd2 25 e6+ @f8! 26 e7+ 29 ... lhe7
~f7 27 exd8lo+ 'if xd8 28 'ii'b4! 30 'itxh8+ (D)
with threats of 29 lbd6+ or 29 b3.
24 ... ~d5
25 l:.e2 'fr'b6
If 25 ...'6'xb5, then 26 'ii'e6+
'it>f827 f7 'ii'xe2 28 'ifxe2 with ad-
vantage to White.
26 'iig5! (D)
Short of time, White nonethe-
less finds new ways to augment his
attack. The threat is 27 J:.e7+, for
example: 26 ...'fi'xb5 27 1:[e7+ 'it>f8
28 'itxg6 'ir'xfl + 29 @xfl ~xe7 30
'illS,7+and mates.
26 ... J:td7
162 Missed Brilliancy Prizes

Playing this move in fairly bad

time-trouble, I faint-heartedly of-
fered a draw, which my opponent
prudently accepted ( 1h.-¼). Imag-
ine my astonishment in the post-
mortem, when I realized I had an
extra pawn in the final position! I
had been material down for so
many moves, and was so glad to
have won it back, that I didn't even
notice I had picked up a bonus! Of
course White's sound extra pawn w
dictates the verdict on the position,
and after the natural 30 ... ~d7 31 14 tl)gS lbf8
'ii'h3+ 'iVe6321Wxe6+ l:txe6 33 b3, 15 dxcS
even my technique should be ade- A standard exchange, opening
quate to win. the long diagonal for the bishop.
15 ... bxcS
Yusupov - Anand 16 :Sdl!
Linares 1991 Bringing his rook into the game
Queen's Pawn Opening and offering a piece sacrifice, White
goes all out for sharp play. It was
1 d4 ~ 2 tl)f3 e6 3 e3 b6 4 Ji.d3 impossible to calculate all the varia-
.i.b7 5 0-0 dS 6 ltle5 ltlbd7 7 f4 g6 tions, but it seemed to me that a
8 b3 Ji.g7 9 tl)d2 cS 10 Ji.b20-011 couple of pawns plus the initiative
'iff3! tl)e8 12 'ifh3 tl)d6 13 ltldf3 should be adequate compensation.
l:te8(D) 16 ... f6 (D)
Both sides have more or less Anand accepts the challenge.
completed their mobilization, and The unassuming 16...:CS!? is less
White resolutely goes into action. in keeping with his temperament;
On the other hand Black is well he very rarely shirks complica-
prepared for his opponent's attack; tions.
he has constructed sturdy lines of 17 ~7! ~7
defence. Possibly I should have 18 lrntg6
preferred the restrained 14 J:ad 1 !?, 18 .i.xg6 is weaker on account
but already I was filled with a of the simple 18...tl)f8.
mood for enterprise. 18 ... 'ikc7
 Missed Brilliancy Prizes 163

organize counterplay on the open

d-file. •
21 l:tad8
22 .:Xd8 'iVxd8
23 'ii'g4
23 'ii'h5? would be a mistake
since White would have to answer'
23 ...'ii'd5 (23 ...'ii'd2!? 24 h3 W'xc2)
with 24 lbe5. Then after 24 ...l:te7
Black retains too many defensive
resources.
w 23 ... 'ii'dS (D)
Black had some other possibili-
18 ...J:.c8 19 J:.f3 c4 is risky in ties here:
view of 20 t2Je7+ 'ilxe7 21 'ii'xh7+ a) On 23 ...'ii'd2 White would
'itiif8 22 .i.g6 l:ted8 23 l:.g3 or 23 have played 24 h4, and if 24 ...W'xc2,
.i.a3 with Q strong attack. But per- then 25 lbf8·J:.e7 26 lbxh7 'ifxb2
haps Black should have gone in for 27 'ii'g6 J:.n 28 h5 .i.d5 29 h6 f5
18 ...f5!? 19 .i.xg7 rJ;xg7. Then 20 30 hxg7 with advantage; if instead
J:.f3 lbf6 21 l:.g3 leads to a repeti- 24 ...f5, there would follow 25 'ii'h5
tion of moves: 21...lbf'e4 (21...lbg4? 'fi'el+ 26 'it>h2 'ii'xg3+ 27 ~xg3
22 lbe5 ! and wins) 22 .i.xe4 lbxe4 .i.xb2 28 lbe5 J:.e7 29 'it'd 1 with
23 lbe5+ lbxg3 24 'ii'xg3+ ~f8 25 the better game for White.
t2Jg6+; while an unclear position b) 23 ...fS was worth consider-
results from 20 W'g3 'it'f6 21 lbeS+ ing. In my view, after the continua-
<:i>h822 lbd7 'ii'c3 23 lbe5. tion 24 'fi'h5 lbf6 25 t2Je7+ J:.xe7
19 J:.f3 lbe4 (stronger than 25 ...1Wxe726 .i.xf6)
19... c4 is risky because of 20 26 .i.xf6 W'd2 27 h3 (27 J:.xg7+
ltJe7+J:.xe7 21.i.xh7+ (better than J:.xg7 28 W'e8+ leads to perpetual
21 'ii'xh7+ ~f8 22 .i.g6) 21. ..rJ;fS check) 27 ...W'el+ 28 ~h2 '6'xg3+
22 J:.g3 c3 23 .i.a3 f5 24 J:.g6 with 29 'it>xg3.i.xf6 30 ~f2 White has
an attack. only slightly the better chances.
20 .i.xe4 dxe4 24 h4
21 J:.g3 The most natural way to go
The rook joins in the assault on about things. White gives his king
Black's weakened king position. some luft and brings his pawn into
Naturally enough, Anand tries to the attack. However, he had a less
164 Missed Brilliancy Prizes

the game was over. White should

pursue the black knight with 25
ltih7!. This time, as the following
variations show, it is Black who
has difficult problems to solve.
b 1) 25 ...l:td8 26 lbxf6+ .i.xf6
27 .i.xf6 °ildl+ 28 'ii'xdl .:txdl+
29 ~f2. White regains his piece
and should win.
b2) 25 ...f5 26 lbf6+! (not 26
'ii'e2 .i.xb2 27 l:txgS+ '3;f7 28 c4
w 'ii'd3 29 'iixb2 '6'xe3+) 26 ....i.xf6
27 'ii'h5 and now:
obvious way of conducting the as- b21) 27 ...l:td8 28 ir'g6+ (28
sault. I rejected 24 ltif8 !? on ac- .i.xf6 'ii'dl+ 29 'ii'xdl lbdl+ 30
count of 24 ... ltig5 (24 ...:Z.e7 is <li'f2l:.d2+ 31 ~el :Xc2 32fxg5 is
weaker in view of 25 loxh7 <ili'xh7 not bad either} 28 ...~f8 29 'ii'xf6+
26 .i.xf6): <tie8 30 ltxg5 'i!fdl + 31 <tif2'i!i'd2+
a) At this point it is true that 25 32 <tig3 'ir'xe3+ 33 <;i;ih4'iixf4+ 34
fxg5 fS 26 'ith5 promises nothing, <;t.ih3.The king is now sheltered
since Black would reply not with from checks, and White wins.
26 ....:.xf8 (when 27 g6 ~h6 28 b22) 27 ...iLxb2 28 'ii'xe8+ @h7
.i.d4! cxd4 29 'ifxh6 'ild7 30 g7 29 h4!? with a plus (29 'il'h5+ ~g8
wins for White) but with either 30 l:txg5+ <li'f8 31 h3 promises
26 ...~ or 26 ....i.xb2. The former less).
would lead to perpetual check: b3) _25...'ii'd2 26 h3 ii'el+ 27
26 ...<li'xf8 27 .i.xg7+ <li'xg7 28 ci>h2 'ifxg3+ 28 'ifxg3 lt::lxh7 29
'ir'h6+ ~-n 29 'ir'f6+.The more in- 'iVel and again the advantage is on
teresting line is 26 ....i.xb2 27 lbh7 White's side.
'l;g7 28 'ir'h6+(or 28 c3 .i.xc3 29 24 ... 'Wes
ltif6 l:th8 30 lbxd5 l:.xh5 31 lt::lxc3 If 24 ...f5, then 25 'fkh5.
~c6) 28 ...~f7. and it is not simple 25 'ir'dl _,d5
for White to demonstrate the cor- 26 'ir'g4 1i'f5
rectness of his attack. You might expect the game to
b )- However, there is at move end in repetition, especially since,
25another, stronger, continuation true to my usual habit, I was al-
which only entered my head when ready running short of time.
 Missed Brilliancy Prizes 165

27 'ir'dl 'ii'dS Viswanathan Anand defends

28 'We2! himself with great ingenuity. The
White plucks up courage and counter-sacrifice of a piece is his
continues the fight. It isn't simple best practical chance.
for Black to find a useful move 31 fxgS?!
here. Thus, 28 ... 'ifd6 is met by 29 Unfortunately, time shortage and
'tlig4l:id8 30 ~h2 l:id7 31 ltJe5 (31 fatigue had already begun to affect
h5 f5) 31...l:t.e7 (not 31...fxe5 32 me - instead of searching for the
..i.xeS 'i!ie7 33 ..i.xg7) 32 h5 'if a6 strongest move, I was content with
33 h6 with a ferocious attack; 'a bird in the hand'. The correct
while if 28 ... ..i.c6, White has the line is 31 ltJe5+! fxe5 32 fxg5. The
decisive 29 h5 @f7 30 'ii'g4 l!g8 strong connected passed pawns
31 h6. The line Black chooses should quickly decide the game,
makes hardly any essential differ- for example: 32 ...l:id8 33 h6 'fi'd 1+
ence to the situation. 34 'ir'xdl l:t.xdl+ 35 ct,h2.
28 ... ..i.c8 31 fS
29 hS 32 ,_e2 ..i.xb2
This pawn takes on the role of a 33 c4 'i!Vd6
battering ram. It breaks up the 34 'ifxb2 eS!
seemingly harmonious co-ordina- I had underestimated this move.
tion of Black's pieces. Of course Black would lose at
29 'M7 once with 34 ... 'ii'xg3? 35 'ii'f6+
30 'Wg4 ltJgS!(DJ ~g8 36 'iVh8+ @f7 37 'iih7#.
Now, however, my pieces lose
their co-ordination, which not only
makes it hard to exploit the extra
material but hands the initiative to
my opponent.
35 l:ih3?!
An ineffective manoeuvre. A
better move was 35 @h2 (35
'ii'cl !?), so as to answer 35 ... f4
with 36 exf4 exf4 37 'iff2 e3 38
'ii'xf4+ °fi'xf4 39 ltJxf4 e2 40 ltJxe2
D.xe2 41 g6+ @g7 421lg5, when
w White has winning chances.
35 ... f4
166 Missed Brilliancy Prizes

36 :Ji4 situations brings success more of-

The consequences of the ex- ten than catastrophe; in any event it
change sacrifice 36 exf4 are un- brings more creative satisfaction
clear: 36 ....i.xh3 37 ll'lxe5+ ~g8 than prematurely terminating the
38 gxh3 'ii'dl+ 39 ~t'2 J:.d8 40 fight.
~g3.
36 ... fxe3 Yusupov - Gulko
37 ~h2? Novgorod 1995
In time-trouble White loses the Nimzo-Indian Defence
thread altogether and blunders.
The right move was 37 'i!fe2!, 1 d4 ll'lf6 2 c4 e6 3 ll'lc3 .i.b4 4 e3
when at least there would be no 0-0 5 .i.d3 dS 6 tbf3 c5 7 0-0 b6 8
risk of losing - as the following cxd5 exdS 9 ll'le5 .C.e8(D)
variations demonstrate:
a) 37 ...1!i'd238 'iffl +.
b) 37 ...'ifd4 can be met by 38
~h2!? or 38 h6!? .i.f5 39 h7 .i.xg6
40 h8'if ltxh8 41 lhh8.
c) 37 ...'ii'd3! 38 'ii'fl+! ~g8 39
'it'f6! (39 h6? 'it'xfl+ 40 ~xfl
~h7) 39 ...e2 40 'ifh8+.
37 .i.f5! -
38 'ii'e2 1i'd2
39 'iffl ~e6
40 lbe4 .i.xe4
0-1 w
Although White gained no re- This game was played in the last
ward for his boldness in this game, round of a tournament in which I
I was not unduly upset about the had done very badly; I was right
half-point I threw away through down among the also-rans. I was
declining the draw. I would have therefore keen to score my first
been far more annoyed if I had set- win, albeit at the very finish. I can-
tled for repetition and afterwards not say I was prepared for the
disc_overed the win. From my own variation my opponent played, but
experience I can assure the reader the shape of the game was more or
that playing for the win in such less familiar to me - after all, the
 Missed Brilliancy Prizes 167

Nimzo-Indian has a firm place in that Black cannot maintain the

my opening repertoire. Black's central tension; he must either ex-
somewhat mysterious ninth move change on d4, thereby liberating
is quite simple to explain: he wants the white queen's bishop, or re-
to preserve the option of playing lease the central pressure by ad-
....i.a6 without loss of tempo. (If vancing his c-pawn.
9 ....i.a6 at once, White has the un- 10 ... c4
pleasant 10 lDc6!.) However, with- In the case of 10...cxd4 11 exd4
out undue effort, I now succeeded .i.a6 12 .i.xa6 l2Jxa6, Gulko didn't
in devising an idea which, though like 13 .i.g5.
not original, is perfectly reason- 11 .i.c2 .i.d6?! (D)
able; in this position, as I later A loss of tempo which increases
found out, it constituted a novelty. Black's difficulties. After the natu-
10 l2Je2! ral 1 l.. ..i.b7 Gulko was afraid of
After I O .i.d2 Black can carry 12 b3 !?, but this would have been
out his plan with 10... .i.a6; the re- the lesser evil; Black could defend
ply 11 ~? is no longer any good, with 12...cxb3 13 .i.xb3 .i.d6.
on account of 11...~xc6 12 _j_xa6
cxd413 .i.b7 dxc314bxc3 lDa5!15
.i.xa8 .i.c5. This and several other
variations are taken from Boris
Gulko's notes in lnfonnator 63.
On the other hand, 11 ha6 lDxa6
121i'a41i'c8 13 l:tacl 'iib7 14 'ii'c6
l:tab8 leads to equality, Portisch-
Spassky, Candidates match (14),
Geneva 1977.
The point of the move White
played is clear: the knight is head-
ing for g3, where it will not only
control e4 - a particularly impor-
tant square in this variation - but 12 f4!
also be ready to join in an attack on A standard idea. This set-up is
the kirig via f5. White's slight loss good if White succeeds in control-
of time is fully compensated for by ling e4 - as he can in the present
the precarious position of Black's case. It would be illogical to ini-
dark-squared bishop, which means tiate play on the queenside: 12 b3
 168 Missed Brilliancy Prizes

b5 13 a4 cxb3 14 ~xb3 b4, with something more than the standard

chances for both sides. development of the attack by 16
12 ... bS li:Jf5 g6 17 'ii'h3 gxf5 18 .i.xf5. By
13 lbg3 lbbd7?! weakening his opponent's queen-
Black gives his opponent addi- side he brings additional factors
tional chances; 13....i.b 7 was more into play.
accurate. 16 ... b4!?
14 'ii'f3 The thematic continuation. Af-
I decided not to deviate from my ter l 6 ...a6 17 /J:Jf5g6 18 'ii'h3 gxf5
basic plan. The alternative 14 e4!? 19 .i.xf5 li:Jb6 20 .:tf3 White has a
would have obscure consequences powerful attack.
after 14...lbb6 15 lbc6!? (15 li:Jxf7 17 as
'itiixf7 16 e5 .i.g4 17 'ii'd2 ~g8 is Not, of course, 17 li:Jxc4? l:tc8
unconvincing for White) 15.. .'ilc7 18 b3 ii.a8, when White loses a
16 e51'.g4 (Gulko gives 16...lbfd7 piece.
17 exd61i'xc6 18 tbf5 Citi>h8 19 'it'g4 17 ... • lic8
g6 20 lbh6) 17 'ii'd2 Wxc6 18 exf6. Perhaps he should have tried
14 ... 1'.b7 17 ... c3. Black probably feared 18
15 iLd2 .i.f8?! (D) a6, not liking 18 ....i.xa6 19 l:r.xa6
15...lLJfS!?is more logical. cxd2 20 'ii'e2 'ii'c8 21 Ji.f5 'fi'b7
22 :lfal, though 18...cxb2 19 lta2
.i.c8 was worth considering.
18 a6 .i.a8
19 q)f5 c3!?
19 ...lbe4? would be a mistake
due to 20 ii.xe4 dxe4 21 'i!ig4, when
21...g6 fails to 22 tl)xd7. On the
other hand if 19...g6, White would
have a new resource thanks to the
queenside advance he has pro-
voked: 20 .i.xb4!? gxf~ (20 ....i.xb4
21 lLJh6+ ri;g7 22 li:Jhxf7 'ile7 23
w 'ii'h3) 21 'ii'g3+ /Lg7 22 .i.xfS.
20 bxc3 g6
16 a4!? 21 'ii'b3! b3!
An .interesting and somewhat un- Accepting the knight sacrifice is
expected de<;jsion.White is seeking hazardous. After 21...gxf5 22 .i.xf5
 Missed Brilliancy Prizes 169

lbb6 23 .l:tf3 ii.g7 24 ltg3 c;ii>f8

25
cxb4 White already has three pawns
for the piece as well as a dangerous
initiative. Gulko endeavours to get
at least some counterplay.
22 ii.xb3 lbe4
At the cost of two pawns Black
has succeeded in establishing his
knight in the centre. Although ob-
jectively his counterplay is insuffi-
cient, he has managed to set his
opponent some serious practical w
problems by suddenly transform-
ing the situation on the board. Af- and I experience the desire to 'cre-
ter 22 ... gxf5 23 ii'xf5 it would ate a work of art'. Excessive emo-
have been much easier for White tions during play have, alas, proved
to conduct his attack. harmful to me more than once.
23 ~7! 25 'ii'g8+?!
This of course is the right con- In our post-mortem analysis we
tinuation, as the black king will established that White should have
now have to set off on a walk. 23 played 25 'ii'xg6+ 'iif6 26 lbg7+ !
.i.el was out of the question, since (26 'ii'g8+ 'fkf7 27 'ii'xf7+ <;l;xf728
Black can defend by 23 ...lbxe5 24 .i.el lbxc3 is unclear). If now
fxe5 gxf5 25 'ii'xf5 l:tc7. 26 ... .i.xg7, then 27 'ii'xe4+ c;t,d6
23 ... <j;xf7 (27 ... c;t,f7 28 .i.xd5+) 28 'ife5+!
24 1'xh7+ ~e6 (D) and wins. There is more point in
The critical moment of the game. 26 ... <j;e7 27 ii'xf6+ lt)dxf6 28
White saw the correct continua- lbxe8 lbxd2 29 lbxf6 @xf6 30
tion, but failed to reach the right ii.a4 lbxfl 31 <itxfl l:hc3, but af-
verdict on the ending to which the ter 32 ~e2 (32 l::tbl !? ii.c6 33
main variation leads. In addition I .i.xc6 :.xc6 34 Wt2 is also strong)
was in a strongly combinative 32 ...l:tc4 (or 32 ....i.a3 33 l:tbl .i.cl
mood, and the temptation to draw 34 l:tb3) there is no doubt about
the black king further forward was White's advantage.
too great. Another of my weak- Now, however, he simply has in-
nesses took its toll: I only have to sufficient reserves to achieve more
play a-.couple of attractive moves, than perpetual check.
 170 Missed Brilliancy Prizes

25 ... ~5
26 g4+
White gains nothing from 26
jff7+ in view of the reply 26 ...'iff6
27 'ii'xd7+ 'ife6.
26 ... 'M6
27 fS
White had set great store by this
move. On the other hand 27 .i.el is
bad on account of 27, ..q;e7 28
.i.h4+ ©df6 29 'ifxg6 :Xc3.
27 ... .i.g7! w
A sober assessment of the posi-
tion; Black decides to force a draw. confirm that the win, sadly, has
Instead, 27 ...gxf5?! 28 :Xf5+ ,J;e7 slipped away.
was too risky in view of 29 .i.xd5 ! 29 'iff7+ q;b7
(29 '6f7+ q;d6 30 l:txd5+ q;c7) 30 'ifrs+ q;gS
29 ....i.xd5 30 1Vxd5. White now 31 'iff7+ rJ;b8
obtains a promising position after 32 "ii'hS+ q;gS
either 30 ...©ef6 31 1i'f3 rJ;d6 32 33 'iff7+ q;b8
e4! ©xe4 33 .i.f4+ q;c6 34 l:tel 34 'ihis+ rJ;gS
(but not 34 .i.e5 ©xe5 35 dxe5; be- 35 .i.xdS+ .i.xd5
cause of 35 ....i.cS+ 36 citfl 'fld5), 36 'iixdS+ q;h8
or 30 ...©d6 31 l:tafl! 'irb6 (Black 37 "ii'b5+ q;gS
is mated after 31...©xf5 32 l:txfS 38 'iif7+ q;hS
'file? 33 e4 q;dS 34 l:txf8 l:lxf8 35 There is no sense in g1vmg
.i.g5+) 32 l:lf?+ q;ds 33 c4 Wxa6 White chances with 38 ...«~h7?! 39
34 .i.a5+ l:tc7 35 l:tlf6!. l:lf5 ©clf6 40 :aa5 'flxa5 41 Jlxa5
28 fxg6+ ©xd242g5.
28 'iVh7 l:lg8 291fxg6+ q;e7 30 39 'ii'hs+
'ii'e6+ ffl 31.i.cl fails to 31...©g5 40 'ii'f7+
32 .i.a3+ ©c5 33 .g6 .i.xd4.
28 ... chg6(D)
White now repeats moves a few To conclude, I will give another
times, __so that he can reach the example of a botched attack, al-
time-control and check the varia- though this time things ended hap-
tions accurately. Alas, this will pily for the author.
 Missed Brilliancy Prizes 171

Yusupov - Adams Adams simply didn't know about

Dortmund 1994 that game, which was played a few
Pirc Defence months before Dortmund.
16 lbg4
This too was a last-round game A stock idea in this variation.
which I very much wanted to win, From g4 the knight gives quite good
to do something to rectify my tour- support to the attack, and threatens
nament standing. The encounter to go to h6 at a suitable moment.
meant even more to my opponent, 16 .i.xb2
who would have had a chance of 17 llbl ~g7
first place if he had won. 18 f5
1 d4 d6 2 e4 liJf6 3 liJc3 g6 4 f4 An essential link in the plan. Re-
.i.g7 5 ©f3 c5 6 .i.bS+ .i.d7 7 eS gaining the pawn can wait (18
©g4 8 .i.xd7+ i¥xd7 9 dS dxeS 10 llxb7 would be met by 18... f5). It
h3 e4 11 ltixe4 ©f6 12 ©xf6+ is far more important to develop
.i.xf6 13 0-0 0-0 14 .i.e3!? l2Ja615 White's kingside initiative and se-
lbeS 'ifd6 6P) cure h6 for the knight.
18 ... lbc7?!
In the game I mentioned, Hort
played the stronger 18... ©b4!? in
an effort to obtain counterplay.
Adams' s move is rather passive,
enabling White to dictate the fur-
ther course of events.
After the text-move (18 ... lbc7)
White seems to have only too wide
a choice:
a) 19 c4 b5.
b) 19 f6 exf6 20 .i.f41i'd8.
w c) 19 tbh6+!? .i.xh6 20 .i.xh6
:t°d8 21 c4 gxf5 22 .i.f4 e5.
White has played the opening d) 19 .i.f4!? .i.d4+! (as Adams
confidently, the chief explanation pointed out, 19...IVdS is dangerous
being that I had had this variation - for Black: 20 .i.xc7 'flxc7 21 d6
and the diagram position in par- 'fid7 22 dxe7 .i.d4+ 23 'itihl 'fixe7
ticular - in a Bundesliga game 24 f6, with an attack) 20 'flxd4
against Vlastimil Hort. Perhaps cxd4 21 .i.xd6 exd6 22 lhb7
 172 Missed Brilliancy Prizes

%:tac8!(22 ...lbxd5 23 f6, with lbh6+ 22 lbh6+

to follow - Black is weak on f7), But not 22 lbf2? in view of
and if 23 f6, then 23 ...h5. 22 . .-.'ii'g3 23 .i.xc5 ~f4.
During the game I was mainly 22 <ifi>h7
choosing between this last varia- 23 lbfS (D)
tion, which I didn't think was too
clear, and the move I actually
played.
' • 19 l'hb7!
The start of a forcing operation.
Black's reply is forced, as White
threatens 20 ~f4.
19 ... ~xd5 (D)

White has not succeeded in

winning a piece, but with this
move he totally wrecks his oppo-
nent's pawn structure.
23 ... gxfS
Although Black is nominally a
pawn up after 23 ...'ii'c6 24 •xd5
'ii'xd5 25 cxd5 gxf5 26 .i.xc5, an
20 f6! ending with that kind of pawn
The point of White's idea is to structure can scarcely give him any
•utilize the position of the queens pleasure.
opposing one another. 24 cxd5 . ct>g8
20 ... exf6 25 llxf5
20 ....i.xf6 is weaker on account An interesting alternative is 25
of 21 ltxf6 ~f6 22 •xd6 exd6 23 .i.f4!? 'ii'a6 26 'ii'hl, when White
~_xf6+ <l;g7 24 lbg4! with a won ignores the f-pawns altogether and
position. concentrates his efforts on forcing
21 /c4 h5 his passed pawn forward.
 Missed Brilliancy Prizes 173

25 ... 'ii'a6 should, of course, have put a little

26 'Wbl llie8 (D) effort into checking that after the
I expected 26 ... 'ife2. when the straightforward 27 .i.xc5 Black's
simplest way to maintain the ad- minimal activity gives no great
vantage is 27 .i.f2. White could cause for concern: 27 ... 'i!Va5is an-
also play 27 .i.xc5 l:tfc8 and then swered by 28 .i.f2, and 27 ... .:tac8
28 .l:.b2 -·but not 28 d6? .l:.xc5 29 by 28 d6. The sharpest continua-
:xc5 'ir'e3+ 30 ~h 1 'i!Vxc5 31 tion, 27 ...l::te2,leads to an easy win
:b8+ l:.xb8 32 'i!i'xb8+ ~h7 33 d7, for White after 28 .l:.b8+ (28 lhh5?
on account of 33 ...'ir'cl + (inverting l:.xg2+) 28 ...:xb8 29 'ii'xb8+ '.t>h7
the move-order doesn't work; on 30 .:txh5+ 'it>g631 l:.h4!.
33 ....i.h6? White wins by 34 'ifbl+) 27 .i.f2?
34 <ifi,h2.i.h6 35 d8'ii' .i.f4+. I reckoned that after the virtu-
The move played similarly par- ally forced exchange of rooks my
ries the obvious threat of 27 l:.xh5. passed pawn would decide the
game, but I was overlooking a
strong defensive manoeuvre for
my opponent.
27 l:.e5
28 .l:.b8+ :Xb8
29 Vxb8+ <ifi,b7
30 'ifbl
A useful ploy - White repeats
moves to avoid time-trouble.
30 WgS
31 'ifbS+ ~h7
32 'ifbl WgS
w 33 :Xe5 fxe5
34 'ifb8+ Wh7
I was in no doubt about my as- 35 Vc7
sessment of the position, but on I had been pinning my hopes on
analysing the variations I started to this move. There is no danger for
get confused. Everywhere I saw Black in 35 d6 'it'd3 (35 ... ~f6) 36
visions of some kind of counter- "filc7 Wg6 (36 ... c4 37 'ii'xf7 'ii'xd6
play for Black. Hence I decided to 38 'ii'xh5+ 'ir'h6 is also playable)
pla.y something fairly simple, by 37 d7 .i.f6. But now, seeing that
analogy with 26 ... 'ife2 27 .i.f2. I White is threatening to take on n
174 Missed Brilliancy Prizes

as well as to push his passed pawn, offered him no chance of first prize
I was feeling optimistic - until I in the tournament. Unable to con-
spotted Black's defence. My oppo- centrate fully on fighting for the
nent spotted it too, of course - Ad- draw, he played this natural but los-
atns doesn't miss a chance like this! ing move almost instantaneously.
35 ... 'ii'xa2! Yet Black could have saved him-
36 "iixf7 'i!lbl+ self by 39 ...'ii'f5 ! !. After 40 .i.g3
This is the point! The queen \i'e4 41 d6 (41 "ikf7 also draws)
crosses to the kingside with tempo. 41...h4 or 40 ~e3 'ii'e4 41 ~g5
37 @h2 'ii'g6! "fi'xd5 (41. .. @g6 42 d6 ..i.f6 43
38 "iixa7 ~xf6 'iVf4+ 44 'it>g1 'ii'e3+ 45 ~fl
Mindful of his bad play in the "ifd3+ is also possible) 42 .i.f6 'ii'g8
technical stage of the game, White 43 .i.xe5 'it'h8 44 .i.xg7+ "fi'xg7 45
talces a sensible practical decision 'ii'xc4 "fi'e5+Black should be able
- to minimize the possibility of to draw.
losing. The bolder 38 Wc7 1i'f5 40 d6 c3 ,
would leave the black passed a- 40 ...~g6 41 d7 .i.f6 was rather
pawn alive. more tenacious, though after 42
38 ... c4 .i.b6 Black's position is difficult.
39 'iic7 (D) My opponent had obviously over-
looked White's 42nd move.
41 d7 cl(D)

-39 ... 'iVd3?

Adams was disappointed by the
way the game had gone - it had 42 .i.e3!
 Missed Brilliancy Prizes 175

Black's downfall is due to the 44 W'c7!

fact that his king is on the same Simplest. There is no point in
rank as the white queen, so that if working out the more complicated
the black bishop moves, White has 44 d8'ii' .i.e5+ (though that too is
a decisive discovered check. adequate to win) when a simple so-
42 W'xe3 lution is available.
43 W'xc2+ e4 1-0
 10 A Controversial Position
Mark Dvoretsky

To risk or not to risk? To sacrifice choose between the two - to de-

or not to sacrifice? We quite fre- cide whose interpretation of the
quently have to solve this kind of position was nearer the truth.
problem. Clearly there is not, and
cannot be, any general recipe here. Shamkovich - Simagin
The best advice 1 can give my Leningrad 1951
readers is to consult books and ar- Griinfeld Defence
ticles in which such situations are
discussed. Try them out on your- 1 d4 lbf6
self; study the position deeply and 2 c4 g6
try to decide how you would han- 3 lbc3 dS
dle it, then compare your ideas 4 lt)f3 !ILg7
with the conclusions drawn by the 5 'ii'b3 dxc4
annotator. In this way you wilf not 6 1i'xc4 0-0
only develop your technique of 7 e4 !ILg4
analysis, you will also learn to de- 8 !ILe3 lbl'd1 (D)
cide intuitively whatdegree of risk
is acceptable.
I want to draw your attention to
the analysis of a sharp position
which occurred in·a game by Vla-
dimir Simagin (a master at the time;
he gained the grandmaster title
much later). Eleven Jears after this
game, and without prior knowl-
edge of it, Bobby Fischer reached
th~ same position. F1scher's opin-
ion of it differed from Simagin's.
You will _have the opportunity to w
 A Controversial Position 177

9 'S'b3 .i.xf3 11 ... eS

Black wants to develop his 12 dxe5
knight on c6, but the immediate 12 d5?! 4::'id4is bad for White.
9 ...tlk6 comes up against 10 Wxb7 12 ... ttlcxeS
4::'ia511 'ii'a6, as in Polugaevsky- 13 .i.b3 (D)
Simagin, USSR Championship, Leonid Shamkovich goes onto
Leningrad 1960. (According to the the offensive, hoping to exploit the
Encyclopaedia of Chess Open- pinned d7-knight. Black can an-
ings, though, the position is un- swer 13 .i.e2 with the excellent
clear after l l...c5 12 dxc5 l:.b8.) move 13...'ii'h4! (and if 14 f4, then
The preliminary exchange on f3, 14 ...ttlg4).
removing one of the defenders of
d4, gives White no time to take the
b-pawn. The move has its draw-
backs too, however; hence 9 ...ttlb6
has become the chief theoretical
continuatiQn.
10 gxf3 4:'ic6(D)

13 ... lbxf3+!
14 @e2
On 14 <if->fl,Simagin gives the
continuation 14 ... ttlfe5 ! 15 .i.xd7
ttlxd7 16 'ii'b5 c6 17 'ii'xb7 .i.xc3
18 bxc3 'ii'h4 ! (I may add that
Black also has 18 ...ttlc5!). Now 19
11 l:tdl? °fi'xd7? fails to 19 ...l:.ad8, while if
This· justifies Black's play. As 19 l:t.xd7, then either 19 ...l:tab8 or
subsequent practice has shown, 19 ...i¥xe4 is strong.
W,hite can gain an advantage with 14 lDfeS
11'0-0-01. 15 .i.xd7
178 A Controversial Position

Or 15 l:txd7 'ii'h4 ! (better than In the game Evans-Fischer, USA

15 ...tbxd? 16 l:tdl). Another bad Championship, New York 1962/3
line is 15 f4 1Wh416 il.xd7 tbxd7 Black didn't risk the complications
17 lhd71r'g4+. and settled for the simple line
15 ~xd7 19...'ii'xd7 20 l:txd7 .i.xc3, when a
16 1!t'bs c6 draw became a virtual certainty:
17 1Vxb7 l:r.b8 21 lha7 l:te8 22 l:ta4 .i.b4 23 il.d4
18 1Vxd7 ::tc2 24 l:txb4 c5 25 hcS l:txc5 26
• Not, of course, 18 'ii'xc6? :Xb2+ ~g2 l:tc2 27 a4 l:td8 28 ~g3 l:ta2
19 ,;j;fl 1ih4!. 29 l:tcl l:tdd2 30 lU1 l:td3+ 31 f3
18 ... l:txb2+ ::tda3 32 l:tdl lha4 33 l:td8+ 1;g7
• 19. ffi (D) ¼-¼.
Simagin played differently. He
declined the draw that was offered
at this moment, and sacrificed a
piece.
19 ... 'ii'f6!?
Fischer considered the sacrifice
completely unsound; Simagin tried
to demonstrate that the combina-
tion wins for Black. I suggest that
the truth lies somewhere in be-
tween. Let us look at the variations.
a) 20 tba4?. This is what Sham- ·
kovich played. After 20 .. .l:txa221
tbcS 1i'.f3 22 ~gl (22 l:tgl 'ii'e2+
Here is the key position, where I 23 g.>g2'ifxe3) 22 ... .i.h6! Black's
suggest you try to decide how attack was irresistible: 23 1'd4
Black should continue. 'iVxdl + 24 <it>g2 'iVd225 'iVd61'e3!
He has the choice between win- 26 tbd7 'ir'xf2+ 27 ~h3 'iig2+ 0-1.
ning the knight back- which leads b) 20 l£ie2? 'ii'f3. 21 l£ig3 .i.h6
to an approximately level ending - 22 .i.xa7 (22 .i.c5 doesn't alter any-
and trying to attack while remain- thing: 22 ...l:tc2 23 .i.xa7 l:ta8 ! 24
ing a piece down, by means of l:tel l:txa2) 22 ...J:.aS! (threatening
19 .. Ji'f6 (aiming at the weak f3- 23 ...l:r.xa7) 23 l:r.el (23 .i.c5 .i.f8
square). Which would you have 24 .i.e3 l:.axa2; 23 lhl l:txf2+! 24
preferred? .i.xf2 .i.e3) 23 ...J:.xa2 24 .i.c5
 A Controversial Position 179

.i.d2! (24 ...l:tc2 is also strong) 25 drawn result, or whether he is jus-

.:C.bll:tal 26 'ir'b7 .i.c3 ! with the tified in continuing the attack with
decisive threat of 27 ....l:.xbl+ 28 21...c5. Simagin thinks this is jus-
ifxbl l:tal (analysis by Simagin). tified. He gives the variation 22
c) 201'd4!? and now: l:tg31ifhl+ 23 l:tgl 'fi'xh2 24 l:lg2
cl) 20 ... 1i'h4 is inferior owing 'iihl+ 25 l:.gl 'ii'h4!, with an at-
to the excellent 21 ltld5 !, as indi- tack. He is mistaken, though;
cated by Fischer. After 21...11fxe4 White can play more strongly with
(21. ..cxd5 22 .i.xb2; 21. ..1'xd4 22 22 .i.xc5 ! .i.xc3 23 'ii'd3 ! 'ii'f6 24
©e7+) 22 ltle7+ @h8 23 .i.xg7+ l:tg3. While permitting his oppo-
~xg7 24 ti'd4+ 1i'xd4 25 l1xd4, nent to restore the material bal-
White should win with his extra ance, White has activated his forces
piece. and seized the initiative. Black
c2) 20 ... 'iVf321 l:tgl (D) (better loses at once with 24 ... 1'b4?? 25
than 21 <i>gl c5). .i.d4 (this occurred in a correspon-
dence game Maclellan-Kokorin,
1968; Black resigned). In the event
of 24 ....i.eS 25 l:tf3 'ii'h4, White
achieves no more than a draw with
26 .i.xf8 'ii'xh2 27 .:td2 .:tbl+ 28
~e2 'iWgl 29 l:.dl l:tb2+ 30 l:td2
l:.bl, but the simple 26 h3 !! places
Black in insuperable difficulties.
For instance: 26 ...l:lfb8 (26 ...l:.c8?
27 W'd7 l:txc5 28 •xf7+ and
mates) 27 l:.xf7 ! (stronger than 27
'iVd5 .i.f 4 ! 28 .t.d6? l:txf2+ ! 29
B l:txf2 'iixh3+) 27 .. .'~xf7 28 'iid7+
~g8 29 'iie6+, and White wins
It was because of this continua- (analysis by Larry Evans).
tion that Fischer rejected the piece So with correct play, 20 .t.d4
sacrifice, but he was wrong. After leads to a draw; in other words it
all, if Black wants, he can force a does not refute Simagin's combi-
draw here with 21. ..l:tc2! 22 l:tg3 nation.
(necessary) 22 ...'ii'hl + 23 l:.gl 'it't3. d) There is just one more possi-
Th~ only question is whether bility for us to examine; in my
Black should be satisfied with a view it is the strongest:
180 A Controversial.Position

20 l:td3! 'iv'f3 22 lb:c3 l:tbb8

21 ct,gl! (D) An unexpected retreat; Black
21 ltgl? would be a mistake: sets up the threat of 23 ...l:tfdS. In-
21.. ..i.xc3 22 l:.xc3 l:.fb8 (threat- stead, 22 ...l:.e2 fails to 23 'fi'dl !.
ening 23 ...l:txf2+!) 23 'ifxc6 l:td8 23 l:tcl l:.fd8 (DJ
24 :Cl 'ii'e2+ 25 <:i'g2'ifxe3. Simagin now continues with 24
'ii'xc6 l:.dr+ 25 .:txdl W'xdl+ 26
~g2 'ii'g4+ 27 'iii>fll:.d8! and wins.
However, the defence can be im-
proved.

21 ... Ji.xc3
Simagin recommended this-way
of playing the attack in his notes to w
the game. I have also looked at
21. ..l:tc2. Then 22 Ji.d2? l:[xd2 ! 24 'ii'c7!
and 22 Ji.c5?! 1i'f4 23 'itd6 'ifxd6! By b_ringing his queen back to
24 Ji.xd6 J:td8 25 ~dl .i.e5 are g3, White repulses the attack.
both mistaken replies, leading to a Black still has some initiative, but
pleasant endgame for Black. 22 it obviously does not compensate
iLxa7 is playable, though after for the sacrificed piece.
22 ...l:tcl+ 23 ~dl 1i'xe4 24 .i.e3 Let us sum up. Objectively
l:tal Black retains distinct counter- Fischer was right; Simagin's bold
chances. Best of all is the immedi- attack could have been refuted.
ate 22 ltldl !; White isn't afraid of However, from the practical stand-
22 ...l:.e2 because of 23 Ji.d2. Black point, Black's risk seems to me to
ha~ insufficient compensation for be justified. In the majority of
his sacrificed piece. variations he creates lethal threats.
 A Controversial Position 181

Even after 20 J:td3 his position still ponent will analyse the position
looks menacing, and he has reason infallibly and discover the strong-
to hope that he will still find ways est moves over the board is fairly
of reaching at least a draw. In ad- slim. Such chances are worth tak-
dition, the likelihood that his op- ing!
 11 Attacking after Castling on
Opposite Wings
Mark Dvoretsky

The subject of this chapter is one 7 ... JlxcS (D)

with which most of you are sure to 7 ...tbxc5 is not bad either.
be familiar. It is treated in numer-
ous books on the middlegame. I can
hardly say anything new about it,
nor am I attempting to. We are sim-
ply going to examine a few games.
In the course of analysing them,
we shall recall the chief principles
applicable to positions where the
kings have castled on different
sides and we shall practise putting
those principles into operation.

T. Georgadze - Dvoretsky w
USSR Spanalciad, Moscow 1967
French Defence 8 .i.d3 f6
8 ...0-0? is a mistake in view of
1 e4 e6 the standard bishop sacrifice 9
2 d4 dS Jlxh7+ ! ~xh7 10 ll)gS+. The
3tbc3 ll:\r6 move played is perfe.ctly logical - I
4 eS ll)fd7 prepare to castle, and at the same
s f4 cS time I exchange the strong central
6 t2:)f3 ll)c6 pawn on eS. Other moves have also
..- 7 dxcS been played here: 8...tbb4, or 8...a6
-This move promises White no followed by 9 .. Ji'c7. The knight
advantage. 7 Jle3 ! is thematic. sortie isn't to my liking - White
 Attacking after Castling on Opposite Wings 183

may simply withdraw his bishop to Of course, taking the g-pawn .

e2, but he can also permit an ex- was unthinkable.
change on d3 and recapture with
the c-pawn. Is it worth losing two
tempi in the opening, even for the
sake of exchanging the opponent's
powerful bishop?
9 exf6 ~6
10 'ii'e2 a6
I could simply have castled, but
was attracted by the idea of a posi-
tional trap. The enticing 11 f5? !
would allow an advantageous pawn
sacrifice: 11.. .e5 ! 12 ll:ixe5 loxe5
13 1Wxe5+ ~n. and the white w
king, stuck in the centre, will be
exposed to..a dangerous attack. Now, a question: what would
11 .i.d2 0-0 you play if you had White here?
12 0-0-0 'ii'c7 With the kings on opposite
After castling in different di- flanks you must play as energeti-
rections the players usually cally as you can, and try to seize
carry out pawn-storms on oppo- the initiative at all costs. In such
site wings; each tries to be the situations the slightest delay
first to weaken the enemy king's tends to be fatal.
cover. From this point of view, the The principle itself is perfectly
moves 12 ...b5 13 g4 b4 would ap- clear, but its implementation is
pear logical. Then, however, White sometimes anything but simple.
replies 14 ll:ia4! .i.d6 15 g5, and Here, for example, the attempt to
15...ll:ih5? is bad on account of 16 undermine the opponent's centre
loe5. Hence, before beginning the with 14 f5?! exf5 15 g5 fails
pawn attack, Black must make a against 15...ll:ie4 16 ll:ixd5 Win.
preparatory move to fortify his po- To me, the sharp continuation
sition in the centre. Control of the 14 g5 ! ll:ih5 15 ll:ie5! seems cor-
centre must not be forgotten even rect. Acceptance of the pawn sacri-
when conducting sharp flank at- fice is then extremely dangerous:
tacks. 15...ll:ixf4 16 .i.xf4 lhJ4 17 ll:ixc6
\ 13 g4 bS (D) 'ii'xc6 18 fi'h5 (18 l:thfl!?) 18...g6
 184 Attacking after Castling on Opposite Wings

19 ~xg6 hxg6 20 'it'xg6+ ~h8 21 I disagree with Kotov here. It is

lbxd5. true that in such cases the result of
Tamaz Georgadze must have the battle may hang by a hair - it
been worried about the reply may depend on every tempo. Cal-
15 ... g6!, leaving the f-pawn pat- culating variations plays an impor-
ently weak. In sharp situations tant role - but not the central one.
with the kings castled on oppo- It helps in solving particular prob-
site wings, you often have to lems, but usually (as in the present
make positional or material con- game) it does not enable you to
cessions for the sake of the initia- foresee the exact outcome of the
tive. Don't be afraid of this. attack you are undertaking. For
White continues 16 foxc6 'ifxc6 that reason it doesn't pay to be-
17 ~e4! it.d4, and then either come obsessed with calculating,
bums his boats with 18 h4, with a and you cannot of course confine
view to 18...~xf4?! 19 ~xf4llxf4 yourself to it entirely. It is impor:;
20 ~f6+ _j_xf6 21 gxf6 l:.xf6 22 tant to have a feel for the position -
h5, or else plays the more re- to form an intuitive assessment of
strained 18 l:tdfl !? , intending 19 this or that course which events
l2Jg3. may take - and to gauge the pros-
14 J:ldfl?! pects offered to either side.
This might seem a sensible 14 b4
move - White fortifies his f-pawn 15 fodl ~d6
in good time, and frees d 1 as a 16 ~eS b3! (D)
retreat-square for his knight. How-
ever, even such a slight hesitation
on White's part is enough to let
Black launch his onslaught first.
I will take this opportunity to
quote what Alexander Kotov said
about mutual attacks against the
kings on opposite flanks. He con-
sidered this maxim very important:
"Bear in mind that a pawn-storm is
in the nature of a forced variation,
and th_at when launching it you
have to calculate it just as precisely w
as you wquld a combination."
 Attacking after Castling on Opposite Wings 185

Much stronger than crudely t2Je4 23 .i.xe4 dxe4 24 .:thfl .i.b7

capturing on e5, which would lead (D) was most likely to arise.
to unclear play. Black accom-
plishes his main strategic task -
weakening the pawn protection in
front of the enemy king.
17 axb3 t2Jd4
18 'ir'el (D)

w
Black has an extra pawn but it is
doubled, and the white knight is
ready to occupy the excellent block-
ading square e3. I wasn't con-
vinced that I had the advantage
here, and couldn't find anything
Here I made what was probably better. So I took the draw.
my one serious mistake in the game Do you see where the mistake
- after a long think I accepted my was? It was in my assessment of
opponent's offer of a draw ( 1'2-1/2). the end position. I didn't take the
Mistakes like this need analysing; weakness of f3 into account (if
you have to clarify why they occur. White had a pawn on g2 the posi-
The reasons may be purely techni- tion would indeed be unclear). As
cal, or psychological, or sometimes soon as the knight goes to e3,
a mixture of both. Black's rook will occupy the f3-
First, a reason that can be ex- point. From there it will exert un-
plained on the chessboard. From pleasant pressure on the opponent's
working·out the variations, I con- position, since an exchange on f3
cluded that the position after would be quite hopeless for White.
18... t2Jxb3+ 19 <J.ibl t2Jxd2+ 20 Of course it all seems obvious
'ii'xd2 .i.xe5 21 fxe5 'fi'xe5 22 .:e1 when you look at the diagram, but
 186 Attacking after Castling on Opposite Wings

during the game it is quite possible had examined in it. After asking
to miss something at the end of a • how much time I had left, the ex-
long variation. Still, is it worth World Champion said I was right
making excuses? What is more im- to agree a draw.
portant is to think how to avoid It is clear all the same that a
such mistakes in future. cool-headed, self-assured player
When analysing, you may would never have accepted the
sometimes be unable to visualize peace offer in such a position. It is
the positions very clearly. (Some essential to work at developing
sectors of the board may get ex- your resilience and toughness,
cluded from your field of vision; your ability to keep your bear-
you forget the exact locations of ings in all circumstances - even
particular pieces, etc.) In that the most complex ones - and not
case, special training methods are to be cowed by strong opponents.
called for. Get into the habit of Without all this, you cannot ex-
analysing positions that interest pect major successes at chess.
you without moving the pieces; An effort to play with redou-
go over games in books or maga- bled accuracy and solidity at cru-
zines without using a board; play cial moments is sometimes detri-
blindfold chess with your friends. mental. You are being untrue to
Now for the psychological as- your usual programme of thought
pect of my error. At that time I was and behaviour, and this brings
a young and inexperienced player. considerable danger. Why? Let
I was playing (on the junior board) me give you a simple analogy. None
in the same team as such illustrious of you would have much difficulty
figures as Mikhail Botvinnik and walking along a tree-trunk that was
Vasily Smyslov. Understandably, I lying on the ground. But if the
was very nervous and afraid of let- same tree-trunk were placed across
ting the team down. Consequently a chasm, you would be likely to
I spent more time than usual on ex- fall off if you weren't prepared. On
amining and re-examining the the ground, when ~ere is no dan-
variations, so that at the moment of ger, our movements are to a large
concluding peace I had less than extent automatic; they are suffi-
half an hour for the remaining 23 ciently well guided by our sub-
moves. Incidentally, after the game conscious. Above a chasm, the fear
I showed Mikhail Botvinnik the fi- of falling makes us watch every
nal positic~,nand the possibilities I step, we exert ourselves to avoid
 Attacking after Castling on Opposite Wings 187

putting a single foot wrong. In con- about the black knight coming to
sequence. the natural harmony be- b4. However, this is achieved at
tween conscious and subconscious too great a cost.
is destroyed. It is always more dif- 8 0-0
ficult to act according to an unac- 9 .1'.d3 f6
customed programme. 10 exf6 ~xf6
11 'ii'e2 a6
Bronstein - Dvoretsky 12 i..d2 i..d7 (D)
USSR Championship I didn't hurry to advance my b-
(First League), Odessa 1974 pawn, as I didn't want White to be
French Defence 'scared off'. In answer to 12 ... bS
he would surely have played 13
1 e4 e6 ~dl.
2 d4 d5
3 ~c3 ~f6
4 e5 ~fd7
5
6
7
"
l2)f3
dxcS
cS
~6
hc5
8 a3?!
In my view this move is anti-
positional. Let me explain why:
I) In the opening the pieces
must be developed quickly, with-
out wasting time on relatively in-
significant pawn moves. w
2) Kingside castling has been
made difficult for White, but if he 13 0-0-0?! b5
castles long, the advanced position 14 g4 b4
of his a-pawn will help his oppo- 15 axb4 ~xb4(D)
nent to open lines for the attack. It The shortcomings of a2-a3 are
is highly dangerous to castle on now obvious. Black has opened
opposiJe wings if there are any lines on the queenside and gained
defects ·in the pawn cover in the advantage.
front of your king. 16 g5?!
White wants his bishop to feel Pawn advances often create
co~ortable on d3, with no worries weaknesses in your own camp. It
188 Attacking after Castling on Opposite Wings

Of course Black is not content

to simplify into an ending with
23 ...llxd4 24 'ifxe6+ 'ii'xe6 25
l:txe6 .itf4+ 26 'it;bl. His advan-
tage is much more significant in
the middlegame.
24 lbxdS llxd4 (DJ

is worth it if you obtain a strong

attack in return - but if the at-
tack fails, the weaknesses will
make themselves felt. In this case
it is clear that Black will be doing
all the attacking, so why weaken
the pawn on f4?
16 ... fud3+
The immediate 16... lt}h5 is also 25 lbc3 .i.f4+
playable, since the bishop sacrifice 26 'it;bl .i.d2
on h7 is not dangerous. 26 .. J:td2 27 We4 l:.b8 looks
17 cxd3 tempting, but after 28 'ifxf4 I
If 17 'ifxd3 lbh5, then 18 lbxd5 couldn't see a way to deliver mate.
(or 18 lbe2) fails to 18... .i.bS. On 27 l:tdl .itxc3
18 ~5. Black has 18 ... ~xf4 19 28 bxc3 l:.b8+
Jlxf 4 ltxf 4, and this time 20 lbxd5 29 <ticl (DJ
fails to 20 ...J:td4. Quite a simple little poser: what
17 tbhs is the most precise way to continue
18 d4 .i.d6 the attack?
19 lbe5 lbxr4 I didn't want to exchange a pair
20 .i.xf4 lbf4 ofrooks. To avoid it, there is quite
~21 h4 'ilc7 an easy tactical ploy.
22 lbxd7 'ifxd7 29 'ifc6!
23 J:.del l:.e8? 30 'ifc2 lh4
 Attacking after Castling on Opposite Wings 189

35 'ii'g3
35 .l:tg3 'fihl+ 36 'iffl 'ti'e4+.
35 ... 'ife4+

Even great chess players have

bad days when they play below
their usual strength. Evidently, for
David Bronstein this was one of
those days - which made it easier
for me to win.

B Let us.look at one more 'light-

weight' game.
31 .:thgl lh3
32 c;pd2 'iVc4 Mariasin - Dvoretsky
33 'iVd3 ::tb2+ Kiev 1970
34 ~el (D) Pirc Defence
Or 34 <t>e3l:txc3. Now an ex-
change on d3 followed by a check 1 e4 g6
on the back rank would pick up a 2 d4 .i..g7
rook, but the move I played leads 3 ~c3 c6
to a forced mate. 4 ~c4 d6
s 'iff3 e6
6 J..e3
6 .i..f4 and 6 t2;ige2 have also
been played here.
6 lbf6
7 0-0-0 (D)
7 ... bS?!
8 .i..xb5! 0-0
8 ... cxb5? 9 e5 is hopeless for
Black.
I freely confess that I didn't sac-
rifice the pawn - I just blundered it
away. Hence my seventh move de-
serves its question mark. The ex-
clamation mark is added because
190 Attacking after Castling on Opposite Wings

Black does nonetheless acquire Of course Black cannot allow

some positional compensation (al- h4-h5 - the h-pawn has to be sol-
though certainly not enough) for idly blockaded. When preparing
the lost material. With the kings an offensive on one Dank, don't
castled on opposite sides, it is forget the essential prophylactic
,-
very important to start your at- measures on the other.
tack first. To speed up the open~ 15 'ii'g2 lt.)h5
ing of lines near the enemy king, 16 li.)ge2 l:[b8
it may be worth sacrificing a 17 li.)g3 ~7f6 (D)
pawn.
In the next phase of the game
my opponent's play was noticea-
bly unsure, and the advantage
gradually passed to me.
9 .i.84 'fkc7
10 h4
It was worth considering 10 g4,
intending 11 g5 ~h5 12 li.)ge2 fol-
lowed by 13 ~g3.
10 ... hS! (DJ
11 ~g5?! ~,
. li.)d7
12- ~e3
13 g4 hxg4
14 _1i'xg4 ~6 18 ~5?
 Attacking after Castling on Opposite Wings 191

White is hoping to break the 'itie2 lhbl 29 lhbl :Xbl 30 ll'ixg7

blockade by bringing his other would tum out in Black's favour
knight to g3, but there is no time but this is too complicated. Any-'
for this. A stronger move was 18 way, why close any lines on the
,J.g5, aiming for 19 Ji.xf6 tbxf6 20 queenside, when Black has an im-
h5. mediate chance to do the very op-
18 tbxhS posite?
19 tbe2 cS 22 ii.xe2!
20 c3?! 'ilfaS 23 :Xe2 cxd4
21 Ji.b3 Ji.a6 24 cxd4 lhb3! (DJ
22 l:td2 (DJ
How should Black continue the
attack?

Such sacrifices don't need to be

calculated. It is clear that all the
winning chances are now with
Going for the a-pawn is dubi- Black.
ous: 22 ... c4?! 23 .J.c2 'ifxa2 24 25 axb3 'ilfal+
.i.bl (this is why White played 22 26 ~d2 'i!fxb2+
l:td2 - the b2-pawn is now de- 27 ~el 'ii'al+
fended) 24 ... Wal 25 ll'ig3. White 28 ~d2 'ii'h2+
threatens to go over to the attack 29 ~el 'iVxb3
himself, ·e.g. after 25 ... tbxg3 26 30 J:td2 l:tc8
fxg3!, with h5 to follow. It is just 31 ~e2
possible that the line 25 .. .l:[b3!? 26 White wants to consolidate with
ll'ixh5 llfb8 27 ~d 1 !? l:txb2 28 ~f3, 'ilffl and ~g2. However, in
192 Attacking after Castling on Opposite Wings

the meantime I succeeded in break- ~f5 49 l:tf8 e5 50 .:xd5 .:C3+ 51

ing up his centre. ~d2 lhn+ 52 'iti>e3lU3+ 53 <iiiid2
31 ... 'ii'c4+! ~e4 54 l:td7 g4 55 l:gS f5 56 :es
32 r.;tin l:ta3 57 .1%de7~f3+ 0-1
If 32 l:r.d3, then 32 ... .i.xd4!. If
White plays instead 32 ~e 1, Black Pawns attacking the opponent's
has 32 ...'ifb5! (simply threatening king position cannot give mate on
to push the a-pawn) 33 fi'fl 1r'hl+ their own. The purpose of a pawn-
and 34 ...'iixe4. storm is to open lines for the
32 d5 pieces - for the queen and rooks
33 eS he5! first and foremost, though some-
34 'ii'n .i.f6 times it is the minor pieces that
35 'ii'xc4 lhc4 play the key role in an attack.
36 ~e2? (D) The hero of the next game is my
He had to play 36 l:al with the dark-squared bishop. After occu-
idea of doubling rooks on the sev- pying the long diagonal as early as
enth as soon as possible. Now move two, it never makes any
Black wins without any trouble. more moves. Yet its influence on
events is enormous.

Dvoretsky - Khramtsov
Moscow 1970
Nimzowitsch/Larsen/Simagin
Attack

1 b3 eS
2 .i.b2 l2Jc6
3 e3 dS
4 .i.b5 .i.d6
5 f4 (D)
Of course Black.must not be al-
lowed to maintain his powerful
36 ...~g7 37 .l:.al ltJfS 38 lha7 pawn centre unopposed. The break
.b:d4! 39 :SS+ '3;g7 40 .i.f 4 .i.c3 on the other wing with c2-c4 looks
41 J:td3lhf4 42 l:txc3 lt:Jxh4 43 like a quieter line; the riskier move
l:tc7 gS 44 l:taa7 '3;g645 %ld7ltif5 I played had, however, brought me
46 %lab7f6 47 l:tb8 l2Jd4+48 ~e3 success in some previous games.
 Attacking after Castling on Opposite Wings 193

unavoidable. Success will go to the

side that plays more construc-
tively.
The game Semeniuk-Dvoretsky,
Sverdlovsk 1987 continued 12
'ii'h3 l£ie7!? 13 .J.d3?! (13 .llf4! is
better) 13... tZle4! 14 l£ie2 f5, and
Black seized the initiative.
Alanakian-Dvoretsky, Moscow
1971 went 12 Jl.xc6!? bxc6 13 a4?
(13 'ii'f4) 13... a5 14 'ilff5 l:fe8
B (14 ... 'ii'd6) 15 l:ael J:ad8 16 l:e2
c5 17 'ii'xe5 l:lxe5 18 .:r.ef2 c6 19
5 ... 'ile7 llf5 llde8 20 h3 l18e7 21 llxe5
On 5 ...f6, I intended 6 'it'h5+ !? :Xe5 22 lli4 ~f8 23 g4 ~e7, with
(inducing Black to weaken the a good endgame for Black.
long diag~al) 6 ...g6 7 'ii'h4. 7 ... ~xf3
6 lt)f3 .J.g4 8 'ilxf3 ltJf6 (D)
After 6 ... f6!?, the win of a pawn In Dvoretsky-Makarov, Mos-
is hazardous: 7 fxe5?! fxe5 8 cow 1971, Black went wrong with
.i.xc6+ (8 lbxe5? .i.xe5 9 .i.xc6+ 8 ... e4? 9 'i!Vg3f6 10 lbc3 'ilif7 11
~d8 ! is thoroughly bad for White) l£ixd5 !. There followed: 11...0-0-0
8 ...bxc6 9 l£ixe5 'ii'h4+ (9 ...he5? (11...ii°xd5 is answered by 12
10 'ilh5+) 10 g3 'ilh3 (better than ~c4, and then 12 ...'ii'h5 13 'fixg7
10... 'ii'e4 11 0-0!) 11 1i'e2 ltJf6, or 12... 'ii'g5 13 'ii'xg5 fxg5 14
and Black has a dangerous attack. Jl.xg7) 12Jl.c4 'ii°d713 0-0-0 a6 14
However, after the correct 7 0-0, 'ii°g4 lDa5? 15 lbb6+! 1-0.
the black centre remains vulner- How original would you say the
able. position in the next diagram was? I
7 h3 was astonished to discover that the
I have also happened to play this well-known theoretician Vsevolod
position with Black. My oppo- Rauzer had discussed it several
nents _always opted for 7 fxe5 decades earlier, when annotating
he5 s·Ji.xe5 Ji.xf3 9 'ifxf3 'ifxe5 his game with Black against Via-
10 l£ic3 lbf6 11 0-0 0-0. Objec- cheslav Ragozin in the 1936 Young
tively the chances are about equal, Masters' Tournament in Lenin-
but' this doesn't mean a draw is grad. Rauzer indicated that after 9
194 Attacking after Castling on Opposite Wings

11 ©c3
I also thought about 11 g3 h4 12
. gxh4, but decided it was too corn-
promising.
11 h4
12 0-0-0 ~g3?!
13 wn 0-0
14 tj.)bt (D)
Now what would you play with
Black? To answer this question, it
helps to think about the intention
of my last move.

0-0 exf4 10 exf4 0-0 11 ~xc6 bxc6

12 lcic3 l:.fe8 the chances are on
Black's side; he recommended 9
g3 with approximate equality. I
think 9 ~c3 !? is also worth look-
ing at.
9 fS?! ,
A highly committal continua-
tion. I sensed how risky it was,·but
wanted to force my opponent into
as complex and unconventional a
struggle as possible. At the end of
the day my thinking paid off.
9 ... e4 I came close to exchanging a
10 'irf2 hS few pieces: 14 .t.xc6 bxc6 15 lt)e2
White would like to castle ~e5 16 .t.xe51Vxe5 171ff4. After
queenside. Therefore the most un- 17 ...'ili'xf4 18 lt)xf 4 White has the
pleasant move for him to meet is better ending. Unfo~nately this
10 ...lcih5!, forcing him to castle idea doesn't work because of the
short~ Admittedly, after 11 0-0 the mate on al, but after 14 ~bi Black
attempt to clamp down on the king- has to reckon with it.
side irnmediately with 11...lcig3 is After the text (14 ~bi) I recom-
refuted by 12 f6!, but instead mend 14... a5!, an aggressive and
11...0-0-0, for example, is not bad. prophylactic move which disrupts
 Anacking after Castling on Opposite Wings 195

White's plan. Then 15 .i.xc6?! a) 18 loxg3 hxg3 19-'iff4 (or 19

bxc6 16 loe2 .i.e5 17 .i.xe5 'ii'xe5 'ii'f5).
18 'li'f4 is met by 18... 11fe7! 19 a4 b) 18 lod4 (aiming for 19 ~f5).
(19 'ii'xh4 a4) 19...l:ttb8, and the c) The crude 18 .i.xf6 'ii'xf6 19
white king is in trouble. 'li'xf6 gxf6 20 fug3 hxg3 21 :Iifl.
14 ... 'ii'cS? If Black had taken the pawn I
A major positional error. Black would have had to choose between
must always be ready to meet ~2 the· above lines. but there was no
with a bishop exchange on e5, but sense in wasting time on the deci-
now he lacks that possibility. In sion in advance.
consequence, the b2-bishop ac- 18 ~xg3 hxg3 (D)
quires fonnidable power. Not, of course, 18 ... loxg3 19
15 .i.xc6 'ii'xc6 'ii'f4. •
16 loe2 lobs (D)

19 'li'e2!
17 f6! White has the advantage, but
Stronger than 17 lbxg3 lbxg3 breaking down his opponent's de-
18 '6'f4 f6! (but not 18 ...loxhl? 19 fence is not so simple. On the king-
f6 with lethal mate threats). side Black has built something like
17. ... g6 a fortress. If White plays l:thfl
I didn't even consider Black's with the idea of l:lf5 !, Black replies
acceptance of the pawn sacrifice. .. .'ii'e6, and the rook has no entry
On 17 ... ~xf6 White would have squares on the f-file. White can
the pleasant choice between: transfer his queen via g4 to g5,
196 Attacking after Castling on Opposite Wings

threatening to penetrate on h6, but 27 'iig7+ ~e6 28 h4!. Black then

the threat will be parried by ...Wh7. has nothing with which to oppose
White would have to open up a 'se- the advance of the h-pawn; the en-
cond front' by undermining his op- tire board is under fire from the
ponent's centre with d2-d3 at a bishop.
suitable moment. 25 ~a3+ ~d7
19 ... ~6? Again 25 ...<itte6was more tena-
Khramtsov makes my task much cious.
easier. In such situations greed is 26 'ii'h7+ c;i;>e6
entirely out of place. 27 'iJ/e7+ <MS(D)
20 l:tdfi il)h5
21 lll5!
21 'ili'g4 would be a mistake on
account of 21...f5 22 'Wg5 l:.t7.
After the text-move the threat is
22 l:.xh5 gxh5 23 'if xh5 f6 24
'Wg6+ <it>h825 ltfl.
21 ... f6
White has two strong replies to
21 ...liJg7: 22 l:tg5 and 22 ltf6 'itid7
23 lthfl.
22 'iJ/g4 c;t>r7
23 :.XhS w
This obvious exchange sacrifice
was hard to resist. However, 23 28 'iJ/g7!
.:lhfl might have decided the game More precise than 281:tfl + ~g6
even more simply. 29 %tf4f5 (or 29 ....:tfe&).
23 ... gxhS 28 :Ce8
24 'iJ/xhS+ 9iJe7 29 .:rt+ 1-0
24 ...<itg7 25 l:[fl is wholly bad
for Black. A more stubborn de- To conclude, I suggest you try to
fence is 24 ...~e6, which I in- find the best continuation in the
tended to answer with 25 'if g4+ following positions with the kings
(25 l:.fl !?) 25 ... f5 26 'ii'g6+ 'itd7 on opposite sides.
 Attacking after Castling on Opposite Wings 197

Exercises

3. White to move

2. Black to move
198 Attacking after Castling on Opposite Wings

5. White to move 7. White to move

6. Black to move

Solutions are given on pages 276-81.

 12 Training Session: Making
Difficult Choices
Igor Belov

A large number of the moves we

play require no deep investigation
of the position. It is enough to ap-
ply some standard principles of
evaluation and add just a few
variations. Nevertheless, in nearly
every game a moment of crisis is
bound to occur (sometimes sev-
eral such moments arise); at
that moment the right decision
cannot possibly be found on the
surface - and yet it determines w
the whole future course of the
struggle, perhaps its very out- Belov - Vaulin
come. Katowice 1991
It is some situations of this kind,
arising from my own games, that I Exercise 1. In this position there is
want to bring to your attention. In an unusual balance of material.
conditions of limited thinking time, Who is playing for a win? How
try to solve the problems that con- should White continue?
fronted me. We will then compare
our conclusions. Rook, knight and pawn are worth
Let us._start with the relatively about the same as a queen, but here·
(only relatively!) simpler exam- Black is better off positionally.
ples, and end with those that are There are no weaknesses for White
high!~ complex, almost unfathom- to fasten onto, whereas Black threat-
able. ens to work up pressure against the
200 Training Session: Making Difficult Choices

weak d4-pawn. White can have no consider a pawn sacrifice after 1

high ambitions. Ld5 cxd5: 2 h4 JJ..j63 l:tc 1! Ld4
The most precise way to equal- 4 l:tc7. However, after 4 ... b6 or
ize was pointed out by Ilia Mak.ar- 4 ... b5 White still has problems, so I
iev. think the manoeuvre 11al-dl -d3-
1 .i.xdS! cxdS c3 is more convincing. - Dvoret-
If 1...exd5 then 2 l:tel or 2 'i!Vc3 sky]
a6 3 l:lel. All other plans are weak.er.
2 l:dl lUcS Maxim Boguslavsky, for example,
[Black can improve with 2 ... b6, suggested 1 '1Wc4threatening 2 a6.
with a view to 3 a6? ! b5 followed Black replies 1...a6, and what now?
by 4 ... 'l:.b6 or 4 ... b4. However, af- Does White exchange on d5?
ter 3 axb61lxb6 4 'iic3 in conjunc- There is something illogical about
tion with 5 'l:.al, White seems to be that, since Black recaptures with
out of danger. - Dvoretsky] tempo. Vasia Emelin took the analy-
3 l:d3! sis further: 2 JJ..xd5cxd5 (2 ...exd5
The rook heads for c3. In some 3 1fb4 and 4 'l:.el) 3 '1Wc7l:tfc8 4
lines White may even seize the ini- 'iid6 JJ..f6 5 J:.d 1. The queen is ac-
tiative. The side with the queen tive, of course, but White has had
should seek exchanges! The to postpone the exchange of rooks.
queen's power is easier to exploit An idea of Inna Gaponenko's is
when there are fewer pieces op- questionable: 1 a6 bxa6 2 '1Wc4.
posing it - the chances of break- The black rook comes into play at
ing into the opponent's camp are once on the b-file; the bishop will
increased. try to get at the white d-pawn and
Peter Svidler devised a different attack f2.
route for the rook (after 1 .i.xd5 In the actual game, I am afraid I
cxd5): 2 'ifb.4?! llfc8 (2 .. .llfe8 3 h4 missed my way in this position too.
!ll.e7 4 'ii'd2 and 5 l:tcl) 3 l:.a3. This I understood that I had to aim for
is too artificial. Try to implement exchanges, and considered 1 JJ..xd5.
your plans in the simplest and most However, I didn't s~ the rook ma-
reliable way, or you risk missing noeuvre to c3; I only looked at
something, which is just what Pe- l ...cxd5 2 h4 .i.f6 3 'iie3. Then
ter did: 3 ...l:c4 ! 4 'it'd6 :ds. 3 ...l:tbc8! (threatening 4 ...l:.c4) is
[Once you have clearly grasped strong; if 4 l:cl, Black has the re-
that it is imperative to bring the ply 4 ...JJ..xd4!.
rook to the c-file, you may even 1 :S4? .i.e7!
 Training Session: Making Difficult Choices 201

Black of course prevents 2

~xd5 and 3 l:lb4. Even so, it was
not too late to take on d5, but I de-
cided to occupy the c-file with my
rook first.
2 ltc4?! 4Jc7! (D)

Kamshonkov - Belov
Podolsk 1991

Exercise 2. The game had been

adjourned. This is the position af-
ter Black's sealed move ... ~a5-
c7!, which came as a surprise to
Alas, this simple move escaped White.
me completely. With the retreat of Recollect the positional princi-
the knight, White is suddenly in ples that operate in situations like
trouble. My opponent is planning a this, and with their help, select a
multiple attack against d4. He has plan of defence for White.
more pieces than I have, so the
pawn is essentially defenceless. I There are opposite-coloured
can only hope for some random bishops on the board. It would
tactical counterchances. help to exchange rooks and go
3 .i.e4 .:Cd8 4 1!fd3 h6 5 1i'e3 into a 'pure' bishop ending, but
iU66~c2a6! unfortunately that is not feasible at
Before placing his knight on b5, present.
Black does well to prevent the ad- When defending an endgame
vance a:S-a6. with a material disadvantage,
7 .i.a4 l£ib5 8 .i.xbS axbS 9 you should exchange pawns.
l:lb4 .:as 10 'iff4 l:td7 11 '.tg2 This is a useful rule. Are any other
l:tad8, and Black won. general considerations relevant?
202 Training Session: Making Difficult Choices

Many years ago I attended a lec- are increased. I am glad to say that
ture by Dvoretsky on bishops of half the students in the contest
opposite colours, and from that sized up the position correctly and
time on I have thoroughly absorbed made the same choice.
the chief principles for playing 1 iLh3! <it>d8
these endings. One of the rules he 1...l:.d5 2 1Lxe6+.
formulated says: "The stronger 2 f5 e5
side should keep its pawns on the 3 l:td2!?
squares of the same colour as The rook exchange leads to a
those on which the opponent's forced draw; 3 lLg2 is not bad ei-
bishop moves; the weaker side ther.
should place its pawns on the 3 ~e7
colour of its own bishop." In ac- 4 l:lxd6 iLxd6
cordance with this rule. though in s iLg2 <i>f6
time-trouble, I had not hesitated to 6 iLb7! (D)
arrange my pawns on light squares
(there was a choice between a7/b6
and a6/b5). My opponent's king-
side pawns, by contrast, are on the
'wrong' squares and therefore vul-
nerable. All the same. opposite-
coloured bishop endings have a
strong drawish tendency, and in
my adjournment analysis I found a
forced draw for my opponent. I
had hopes that he wouldn't find it,
though. since he didn't know my
sealed move and the saving line B
was not entirely obvious.
Diana Darchia suggests 1 iLh3, On the queenside too, it is im-
to prepare f4-f5. Correct! We ei- portant to drive Bla~k's pawns
ther exchange pawns or else force onto the same colour squares as his
...e5, when the white f-pawn will bishop. You see how easy it is to
be on the 'right' square and the play good moves, once you know
black e-pawn on the 'wrong• col- the principle.
our.- that of its own bishop. The 6 aS
chances of a successful blockade 7 1Lc6! b4
 Training Session: Making Difficult Choices 203

8 q.,f'J! (DJ you are at a disadvantage, precision

Simplest. I analysed a sharper and accuracy are called for, and not
continuation: 8 .i.e4 <it.>g5 9 ~f3 all methods of defence are of equal
c;tih4 10 f6 h6 11 f7 c;t,h3 121Lf5+ worth. For instance, 1 ~e3 looks
~xh2 13 .i.g4, and no matter what tempting, depriving Black's rook of
I tried, I could find no win here ei- d4 and aiming to meet 1....i.b6+?!
ther. 2 ~e2 l:.d4 with 3 l:.d2 or 3 l:.c6.
However, Black has the unpleasant
reply 1...l:.dl (and if 2 .i.e4, then
2 ...l:.el + winning another pawn).
Sasha Chemosvitov and Inna
Gaponenko recommend 1 l:[c5.
Why? Is the idea to attack the h7-
pawn? I would surrender it with
pleasure if I could get at the b3-
pawn. My answer would be l...l:td4
2 l:th5 lhf4 3 l:txh.7+ q.>d6.Or if 2
f5, then 2 ... exf5 3 l:.xf5 l:tb4 4
l:U7+ <itd8. After 5 l:.xh7 .:.Xb3
B White is hard put to survive, but
otherwise he comes out two pawns
8 ... c;tigS down (5 l:.f3 .i.xh2; 5 .i.d5 l:th4).
If 8 ...c.ii>xf5,
then 9 .i.e4+ and 10 [A good defensive idea was found
.i.xh7. by Grandmaster Evgeny Bareev: 1
9 c.ii>e4 .i.c7 'JL:3/? (threatening 2 l:.h3) l ... l:td4
10 .i.d7 2 .i.b7!. For example, 2 ... iLx/4 3
By now the game is obviously a La6 b4 4 l:td3 with a draw. The
draw. same idea can be carried out a lit-
In the plan of defence we have tle differently: 1 l:c5 l:td4 2 .i.b7!.
examined, an idea characteristic of In this connection it is worth re-
opposite-colour bishop positions is calling another principle for de-
consistently implemented, namely fending opposite-bishop endings:
the correct arrangement of the attack your opponent's pawns with
pawns. Of course, a practical game your bishop. - Dvoretsky]
is not a study, and there may be Ilakha Kadymova suggested 1
v~ous possible ways of handling .i.e4. This is just what my oppo-
the position. All the same, when nent played, and no good came of
204 Training Session: Making Difficult Choices

it; the black pieces immediately

sprang to life.
1 ~e4? :d4 2 <if.ie3(2 Slab?!?)
2.. .ltb4 3 J:tc3(D)

Lempert - Belov
Katowice 1990

Exercise 3. Who is better here?

How should Black continue?
3...hS
Better than 3 ...h6- the stronger Unfortunately, two students were
side should not place pawns on the unable to come to any conclusion
same colour squares as its bishop. at all. Vladimir Baklan suggested a
The position is now difficult for move which never came into my
White, as his pieces are tied down head: l...e3. His idea looks highly
to the defence of his weak pawns. dubious, for instance 2 fxe3 tbe4 3
This is a suitable moment to recall d6 :xcl 4 l:.xcl .i.xd6 5 l:.dl l:te6
another important principle. 6 Jlc4 J:tf6 7 .i.e5.
If there are other pieces on the 1 ... .i.h6!
board apart from the opposite- If Black manages to exchange
coloured bishops, you must on the strong f4-bishop without a seri-
no account remain passive; you ous mishap, his position will be
have to seek counterplay and preferable.
fight for the initiative at all costs. However, if you only looked at 2
4 h3 .i.b6+ 5 <i1if3.t.d4 6 :d3 :Xc8 lbxc8 3 'ilc7, your move was
(j;e7 7 f5 e5 8 .i.d5 <itir69 Sl.e6 as superficial. The real test ofBlack's
10 :d2 e4+! 11 ~e2 a4 12 bxa4 idea is 2 'iVxd6! 'iixd6 3 l'bc8. If
bxa4, and Black won. then 3 ... 'ii'f8?! 4 :c7+ (j;h8 (not
 Training Session: Making Difficult Choices 205

4 ... J:.e7 5 d6 l::txc7 6 dxc7 ii.xf4 7 one recommendation, 6 ... 'ii'xb3!?,

l:.d8) 5 ii.xh6 'i!i'xh6 6 d6, the ad- is very interesting. Where is the
vantage is with White. Black has to rook to go? Black's chances here
play 3 ....i.xf4! 4 :Xe8 (D). are not at all worse.
I admit I only examined 6 ...f4!?
7 l::txe4f3 8 l::td2fxg2. Then 9 d6 is
answered by 9 ... 'ii'h3 10 l::txg2
'ii°d3 11 l:.e7+ -.t>h612 l::te6 'ii'd4+,
with a draw. White might try 9
l:.ee2 !? , so as to keep a rook on the
d-file, but even this is unclear.
We can now give an objective
assessment of the starting position.
Black has some problems to solve,
but after l ....i.h6! he is justified in
expecting to save the game.
' B Let us now look at how the ac-
tual game went. Hardly giving it a
Peter Svidler reached this posi- thought, my opponent exchanged
tion in his analysis, and judged it to rooks.
be in White's favour. Without the 2 .:.Xc8?! lbxc8
bishops on the board, this would 3 'W/c7 l:td8
certainly be the right verdict - the 3 ...l::te7 is not bad either.
two rooks combined with the dan- 4 l:tcl 1'.xf4
gerous passed d-pawn are stronger 5 'Wixf4?!(D)
than the queen. A bad line is The line 5 'ifxd7+ l:[xd7 6 lhc8
4 ... 'ii'd7? 5 gxf4! 'ifxe8 6 d61i'd7 7 is sounder; White would certainly
.i.c4 followed by 8 h6, and White not lose the resulting opposite-
wins. coloured bishop ending.
It was Emelin who calculated 5 ... lbd6
furthest -he gave 4 ... 1'.xg3!. Af- 6 'Wies
ter 5 l::te6, Black has 5 ... 'ii'c5 or Here my opponent offered a
5 ... i&.xf2+. We need to look at 5 draw, but I refused. Black's pawns
fxg3 'iVxg3+ 61'.g2. What happens are excellently placed, confining
now? the enemy bishop. The 'Nimzo-
6~..e3? is anti-positional; after 7 witsch knight', blockading the
llfl Black's pawns are blocked. But passed d-pawn, is very strong.
206 Training Session: Making Difficult Choices

Exercise 4. My opponent's last.

move, 1 ~d3-b5, has set me no
easy problem - that of rescuing the
pinned a4-knight. Your task (like
mine during the game) is:
a) to evaluate the position;
b) to find various ways of play-
ing the black side, and supply the
essential variations;
c) to select the most promising
line.

As to the evaluation, opinions

6 ,,l:e8 varied. "White's better'' - ''White's
7 'ii'd4 l:e7 worse" - "It's equal". The students
8 'ti'b(; • :r, nearly all observed that l ...l:txf3?
9 .i.113 'ii'e7 fails to 2 l:txa4!. They suggested
10 'iVcS - 'i!feS 1...d3!?, some of them without giv-
Black has an undoubted plus. ing any analysis. Three moves were
Unfortunately I later blundered analysed:
twice, and let the win slip. a) Many of them gave the varia-
tion 2 l:xa4 :Xa4 3 ~xa4 dxe2 4
'ifxe2 1i'a8. Black wins his pawn
back and gains the advantage.
b) After 2 ~f4:
bl) Svidler analysed 2 ...~c3 ! 3
l:xc3 'ifxa.5,when after 4 :Xa3
'Wxb5, the e4-pawn is under at-
tack.
b2) Instead Baklan suggested a
line giving up two pieces for a
rook: 2 ...~b2 3 l:txa4lha4 4 .i.xa4
Juel 5 jfxcl 'ifa5. This is an in-
genious idea, but we already
know that there is no need for it; 2
- - Yachmennik - Belov lbf4 is more strongly answered by
Smolensk 1989 2 ...~c3!.
 Training Session: Making Difficult Choices 207

b3) [In addition, White has to c21) 7 J:.e3 l:la8 [In fact after
reckon with 2 ... Le4!? 3 l:xa4 7... :.dB! it is not at all easy for
'D.xa44 La4 e5. - Dolmatov] White to exploit his material plus;
c) Only Makariev examined 2 his opponent's pieces are just too
lng3 !, but he stopped after 2 ....i.c3 ! active. - Yusupov] 8 a3 J:la49 tbct2
(the idea suggested by Baklan with is a possibility.
the knight on f4, 2 ....i.b2 3 l:.xa4 c22) 7 :td4 l:ta8 8 l:.c4 l:ta5 9
:Xa4 4 .i.xa4 .i.xcl 5 'ii'xcl 'iia5, tbd4. Analysing this far, I saw that
is bad for Black here; White plays I was a pawn down with a difficult
6 .i.b3, with the terrible threat of position. [Black can regain the
1Wh6and log5) 3 l:.xc3 'ii'xa5 [the pawn with 9 ... fue4. After 10
pretty counter-stroke 3 ... l:xa2 liJxe4 ( 10 li)b3 l:e5) 10 ... Le4 11
would lead to equality in the case tbb3 l:te5 12 /4 :.e6 13 l:.cB+ <3;g7
of 4 'ikxa2? lt:l:cc35 'iial d2; un- 14 li)c5 Ac6 15 .:Xc6 hc6 we
fortunately it is refuted by the pro- reach a drawn ending. The conclu-
saic 4 'ZL:a4! 'D.xd2 5 liJxd2 - sion is that l ... d3!? gives excellent
Dolmatov1' Actually the variation saving chances. - Yusupov]
needs to be continued: 4 l:.xa3 c3) [As Grandmaster Bareev
'ii'xb5 (D) and now: has shown, instead of taking the
d3-pawn White has the strong in-
termediate move 5 'IJ..b3!.After
5 ... 'ffa6, he continues not with 6
'ikh6?! f6 but with 6 :Z.Xd3tbc5 7
:dB, maintaining a substantial
plus, for example: 7.../6 ( capturing
on e4 would lose a piece) 8 :xJB+
~x/8 9 'ii'dB+ 9;j7 JO 'ikh8. -
Dvoretsky]
Are there any other candidate
moves in the diagram position? Va-
dim Zviagintsev mentioned 1...lbc:3
w (though without giving any analy-
sis). After 2 :Xa3 ti::JxbS3 l:.b3
cl) 5 :Xd3 is poor due to Black has no compensation for the
5 ... loc5. exchange.
~2) 5 1i'xd3 'ii'xd3 6 l:.xd3 locs Weighing up the variations, I
and here: came to the conclusion that in the
208 Training Session: Making Difficult Choices

normal course of events I was most penetrate with his rook to White's
likely to lose. I didn't want to rec- second rank.
oncile myself to such a dreary fate,
and kept on looking. Finally I suc-
ceeded in finding an amazing
chance.
In principle, Black's position
does have some good points - for
instance the bishop-pair and a
compact pawn-chain. The idea of
constructing a fortress suddenly
occurred to me ...
[For my own part I would like to
suggest one other idea:,1. .. Le4!?
2 :Xa4 llxa4 3 La4 'ila8 4 .i.b3
( or 4 ~dl) 4 ... L/3 5 gxf3 °fkxj3.
Without a doubt, the resulting po- 6 ~c6?! l:.d2
sition is in White's favour; his 7 ~ 1Lxc6
passed a-pawn may become ex- 8 .i.xc6 .i.d4
tremely dangerous. But for the mo- 9 ltle4 .l:.b2
ment he has to worry about his 10 l:tdl .i.b6
shattered kingside and deal with- The threat is 11...fS.
Black 's threat to advance the cen- 11 .:.d2 lhd2
tre pawns. Frankly, this continua- 12 ~d2(D)
tion seems to me more promising
than what occurred in the game. -
Dolmatov]
1 lba2!?
2 'ii'xa2 'ii'xaS
3 'ii'xa4 'ii'xa4
4 ha4 he4
5 t°Mxd4 l:td8 (D)
Black has just one pawn for the
sacrificed piece. Yet realizing
White's advantage is not so simple.
If 6 ~f3 or 6 itlb3, Black would
continue with'6 ...l:ta8, aiming to B
 Training Session: Making Difficult Choices 209

I had seen this position a long my assessment of the position was

way ahead and judged it to be right. On resumption, there was
drawn. White's only object of at- later on an interesting episode:
tack is the IT-pawn, and I couldn't
see how both his pieces could at-
tack it at once - since the dark-
square approaches would be de-
fended by my king and bishop. Of
course I was not totally confident
of a favourable result, but I think
that from the practical standpoint
the decision I took was justified. I
had a clear idea of how White was
going to win in the lines where he
had an extra pawn, whereas I could
not imagine how he could succeed
here.
12 .i.d4 My opponent tried 1 f5!?. After
13 ffl hS the end of the game he pointed out
14 'it-e2 '3;g7 that even I.. .exf5 !? 2 !i)xf7 <tif6
15 f3 e6(D) would leave him with no chance of
success. As usual, the drawing po-
tential of opposite-coloured bish-
ops is immense!
The game continuation also led
to a draw: 1... gxfS 2 .i.bS bgS! 3
hxgS <atig64 <tif4 eS+! 5 <atixeS
<t;xgS 6 .i.d3 f6+ 7 c.te6 r4 8 .i.e2
· 'it>h4.
What main conclusion follows
from the example we have just ex-
amined?
When defending a difficult
position, it pays to consider the
most implausible resources, to
In,\ this position the game was have faith in yourself and boldly
adjourned. Analysis confirmed that transform the character of the
210 Training Session: Making Difficult Choices

fight. Your opponent is most Exercise S. White threatens f2-f4;

likely to cope successfully in a how does Black resist? At first I
standard position that is techni- thought I was in a very bad way,
cal in nature (with an extra but then ...
pawn, for instance). It is much
harder for him in a position with Sergei Movsesian decided on
an unconventional line-up of 1...~f8. This is no good! Nearly
forces - here the probability of everyone gave the variation 2 ~e4
error greatly increases. tbc4 3 l:tb3 with advantage to
Incidentally, having found this White, and Svidler continued
difficult and unexpected defensive 3 ...l:tb2 4 l:.xb2 .i.xb2 5 .i.f6!. I
idea, I not only saved half a point don't know if he saw the counter-
but experienced a huge emotional stroke 5 ... tbd2+, but it doesn't
boost, thanks to which I began change the assessment of the posi-
winning game after game. tion. I thought that in place of 2
tbe4, a more natural move was 2
lbb5 (why should White block the
e-file?). For example, 2 ...:Xa2 3
tbxd6 l:.a6 4 l:.b3 tbd7 5 tbc4 with
a clear plus. However, after 2 ...tbc4
3 .i.e7+ ~e8 (3 ... ~g8 4 l:.el) 4
loxd6+ €lxd6 5 .i.xd6+ ~d7 6
.i.f8 .i.f6 (or 6 ....i.d4 7 l:.e7+ <;pd8)
the outcome still remains unclear.
Evidently the knight's move to e4
is stronger after all.
Emelin found nothing better
B than 1...f6 2 f4 €ld7, but he rightly
assessed the position as difficult
Mitiaev - Belov for Black.
Moscow 1989 An ingenious idea, (in the spirit
of the preceding game) was de-
I had had a winning position, vised by Zviagintsev: 1...f5 2 f4
but with the onset of the endgame I <;pf7!?3 fxe5 .i.xe5. However, af-
weakened and completely forgot ter 4 .i.e 1 I doubt if Black has real
about the possibility of getting compensation for the sacrificed
mated. piece. In any case, White doesn't
 Training Session: Making Difficult Choices 211

have to take the knight - 3 ~b5 ! is outside passed a-pawn and at the
very strong. same time avoid losing his d6-
There is one more possibility to pawn?
consider. Let us look more closely at
1 ... g5! White's main threat of f2-f4. This
2 hg5 move may win the game, yet in it-
But what now? Chemosvitov self it is anti-positional. The sec-
rightly observed that the exchang- ond rank is suddenly weakened,
ing combination 2 ...l:tx.f2+ 3 'ii;ix.f2 and the activity of White's bishop
lDg4+ 4 'ii;if3 ~xe3 5 .i.xe3 .i.xc3 is reduced. Can't Black somehow
leads to a hopeless ending for make use of all this, and organize a
Black after 6 Jlf4. counter-attack against the enemy
Most of the participants re- king?
stricted themselves to the modest 2 ... f5!?
2 ...f6 3 .i.h4 'ii;if7(3 ...~g6 4 ~b5! 3 f4 (D)
lDxh4 5 :t.e8+ 'ii;if76 ~xd6+ 'ii;ig6 [As Bareev pointed out, 3
7 l:.xh8 iB scarcely any better). lDb5!? was also strong, while the
How do we rate the ensuing posi- attempt to exclude this possibility
tion? According to Svidler and by inverting the move-order - I ...f5
Makariev it is unclear, perhaps a 2 /4 g5 - doesn't work in view of 3
shade more comfortable for White. fxe5! gxh4 4 exd6. - Dvoretsky]
Well, compared with me they are
great optimists. I thought the posi-
tion was hopeless. White has an
extra pawn, and the h8-bishop has
no moves. Let us look at a specific
line: 4 loe4 lt)c4 5 l:Lb3?f5 6 l:tb7+
"'g6 - here, without a doubt, ex-
cellent counter-chances arise. White
cannot play 7 ~xd6 ~xd6 8 l:Lb6
~h5. But why should he go into
the complications at all? I think 5
l:.c3 !,_as suggested by Dvoretsky,
throws told water on Black's hopes. B
After 5 ...:Xc3 (5 ...lDd2+ 6 "'e2) 6
~xc3 he probably wins back the 3 ~g4
h6:.pawn, but how can he stop the 4 l:te8+ ~7
212 Training Session: Making Difficult Choices

s :W.8 lbe3+ very unclear consequences, since

5 ...l:txc3 is hopeless in view of 6 the king jumps in at h3. Unfortu-
1:lxh7+ 'itg8 7 l:tg7+ <;t,hs 8 l:td7 nately, after 11 ..i.d8 ! (Dolmatov),
l:k2 9'itel !?, withl:txd6to follow. White wins after all. Maybe Black
6 ..t>gt(D) should try 7 ... lbg4 8 g3 lbh2 (but
not 8...%le1+ 9 <;t,g2.l:c2+ 10 @fl
lt'ie3+ 11 @gl) 9 Ae8 ! (preparing
.l:e3, so as to destroy the well-
known drawing mechanism ...l:r.d2
and ...lt'if3+) 9 ...@h5! followed by
10 ...~g4. White would need to
tread carefully, though with cor-
rect play his position would still
probably be won.
7 ... ~hS (D)

If Black now takes the knight,

the advanced white h-pawn de-
cides the game: 6 ...lhc3? 7 lhh7+
~g6 8 l:tg7+ @h5 9 h7. There is no
imptovementin6 ...lbg4?7 lhh7+
<ili>g6 8 l:[g7+ Wh5 9 lbdl ! :kl 10
'itfl.
6 ... ~g6!
By now, the main strategic basis
of Black's pawn sacrifice with
... g5 ! is comprehensible. It was 8 lt'ib5?
important to give the king some air My opponent thought for about
and a path to freedom. fifteen minutes here, but still failed
7 l:tg8+ to find the right solution. He was
In answer to 7 lt'ib5!? I was clearly very keen to hang on to all
planning 7 ...ltxg2+ 8 'ifi'hl lha2 9 his extra material. However, if he
lbxd6 lt'ixd5 10 Ag8+ ~h5, and if wanted to go after the d6-pawn, it
11 lbxf5?, then l l...~g4! with was more logical to do it a move
 Training Session: Making Difficult Choices 213

earlier, without driving the black against a rook - either in the previ-
king forward. ous line (with 10 lhg4 instead of
The move I feared most was 8 10 'iii>e2), or by 8 .ilf6!? l:tcl+
ltg7, as I couldn't see what to do (8 ...lDg4 9 ltg5+ and then 9 ...'iii>xh6
about the powerful h-pawn. For 10 .ilg7# or 9 ...'iii>h4 10 :C.xf5+
example: 8 ...:Xg2+ 9 'iii>h 1 .l:tc2 10 'iii>g3?11 lDe4#) 9 'iii>f2~g4+ 10
.:Xh71:.xc3 1 l l1e7 'it>g412 h7 'itg3 l:txg4 'iii>xg411 a4. White should
(12 ... 'iii>f313 l:.xe3+!) 13 .ilh4+!. be able to realize his advantage,
Dvoretsky found a defence, how- though there are still some techni-
ever: 8 ...~g4! 9 g3 (9 'iii>fllDe3+) cal difficulties to overcome.
9 ...ltxc3 10 l:.xh7 :.et+ 11 'iii>g2 s ... lDg4
ltc2+ 12 'iii>fllDe3+ 13 'it>gl (Black 9 g3 lDh2 (D)
has the same answer to 13 'iii>el) Black wants to construct the
13 .. .'it>g4! 14 l:.e7 'iii>f3(threaten- drawing mechanism I mentioned
ing 15 ...licl+ 16 'iii>h2lDg4+) 15 before: ...l:.d2 and ...lDf3+. An idea
:Xe3+ 'itxe3 16 h7 ltcl+ 17 'itg2 of Dvoretsky's also deserves to be
ltc2+ 18 'iii>h3:tc 1. studied: 9 ...lic 1+ 10 'it>g2lk2+ 11
The black king isn't much more @fl ltJe3+ 12 c;t,gl 'it>g4, but as-
comfortable on the h-file than on sessing the consequences with in-
the back rank. The attempt to play adequate thinking time would be
for mate with 8 .ile7!? looks invit- difficult.
ing. If 8 ...l':txc3 (reckoning on 9
l':tg5+ 'iii>h4! with unclear conse-
quences), then the quiet move 9
g3!!, found by Dolmatov, is deci-
sive - the king cannot escape from
the mating net. It follows that
Black must defend with 8 ...l':tcl + 9
'iiiiif2(9 <it.>h2lDg4+ 10 ~h3 .:Xc3+
11 g3 ~f2+ 12 'ittg2 ltJe4) 9 ...lDg4+
10 'ite2 .:Xc3. There is, of course,
, not much joy here for Black, but he
can still fight on.
Without doubt White's simplest w
method, and the most unpleasant
one fo(_his opponent, consisted in At this point White wasted most
reaching a position with two pieces of his last few minutes looking for
214 Training Session: Making Difficult Choices

a forced win that is not there. He 15 ffl c4!?

should definitely have settled for 16 ll:lxd6 <itf3
10 ~f6 lllf3+ 11 <itfl l:td2 12 17 <itel c3(D)
l:tg5+ ll:lxg5 13 fxg5 l:txd5, when
Black :would still have had to de-
fend accurately.
10 l:te8?! <itg4
11 l:te3 lLJf3+
12 <itfi?! (DJ
12 l:txf3 °it>xf3leads to a draw.

w
18 lfuaS??
18 <itdl was essential, with a
probable draw.
18 ... l:tcl#
(0-1)
From the practical point of
view, the decision I took ( l...g5
My opponent's errors in the fi- and 2... f5) was, I think, correct,
nal stage of the game are easily ex- everi though analysis has revealed
plained. He considered the position more than one refutation. As in the
won for White. He viewed my previous example, I wanted to de-
pawn sacrifice (l...g5) as despera- flect my opponent from a purely
tion. The move 6 ...<it>g6!,confus- technical course - I endeavoured
ing the issue, came as a surprise to to 'randomize' the position and
him. In severe time-trouble (which create active counterplay at any
was mutual) he loses the thread, cost. For my opponent to find his
and even the game. bearings in the new situation, with
12 mutual attacks and an unconven-
13 lhf3+ tional distribution of material,
14 "'el proved far from simple.
 Training Session: Making Difficult Choices 215

In seeking a solution, you students performed. The best score

should not confine yourself purely was made by Svidler. Zviagintsev
to examining variations. Gen- was second, and Emelin third. But
eral positional considerations of- then, they all made quite a good
ten come to your aid too. In this showing. I hope the experience
case, for one thing, I was looking gained from solving and discuss-
for a way to utilize the defects of ing these exercises will be of some
f2-f 4, and for another thing I was help to them at the board, where
trying to let my king out of its cage. they are sure to keep encountering
Now let me announce how the such tough problems.
 13 Virtuoso Defence
Mark Dvoretsky

While I was studying Grigory Sanakoev - Maeder

Sanakoev's collection of games 10th World Co"espondence
(which we discussed in Chapter 8, Championship 1979-84
Thoughts about a Book), the fol- Sicilian Defence
lowing ex~eptionally tense and
fascinating duel attracted my at- 1 e4 cS
tention. White worked up an ex- 2 'iJf3 d6
tremely dangerous attack against 3 d4 cxd4
the king and seemed certain of vic- 4 ~d4 ~6
tory. Yet by cool defensive play his 5 liJc3 a6
opponent not only managed to parry 6 .i.gS e6
the immediate threats but even 7 f4 .i.e7
seized the initiative and eventually 8 1i'f3 1i'c7
won. In such cases the cause of 9 0-0-0 liJbd7
failure usually lies in errors in the 10 .i.d3 bS(D)
conduct of the a.ttack. Here, on the
other hand, as shown by Sanako-
ev's analysis - which in place~ has
been significantly refined and sup-
plemented by the aµthor of these
lines and by Grandmaster Zviag-
intsev - it is practically impossible
to fault White's play; the result
was determined by Black's in-
spired defence.
Once again it is interesting to
consider the contrast between cor-
respondence and over-the-board w
play.
 Virtuoso Defence 217

11 .:bel ~b7
12 'Wg3 b4
13 lbdS exdS (DJ
13... lbxd5 is a mistake in view
of 14 exd5 ~xd5 15 ~xe6 .i.xe6
16 :xe6. 13...~xd5 is also bad: 14
exd5 ~xd5 15 .i.xe7 ~xe7 16
"i!lxg7.llf8 17 ltxe6! fxe618 ltJxe6,
Chudinovskikh-Semionov, USSR
1974.

turns out clearly in White's favour

after 17 'i!Vxg5'iif4+ 18 'ii'xf4 ~xf4
19 exd7+ ~xd7 20 .i.f5+ ~d6 21
g3 ~g6 22 ~h3 l2Je7 23 lte3, or
17... ~df6 18 exf7 + rbxf7 19 ltJe6
'ii'd6 20 .i.e2 g6 21 .i.xh5 .:ae8 22
~g4 ~c8 23 lUl lhe6 24 .i.xe6+
'ili'xe6 25 lbd5 (Lepeshkin). Black
has to accept the queen sacrifice
with 16...~xg3 17 exf7+ ~xf7 18
:xe7+ ~g8 19 hxg3 'iixg3
14 eS!? (19 ... 'ili'e5 20 ~f5) 20 tbe6 'fi'e5.
The main theoretical continua- After 21 ltfl, Black should play
tion is 14 exd5 ~dB, with unclear 21...~f8!? 22 .i.f5 .i.c8 23 :es
consequences. ~f7 24 J:[e7+ ~g8 25 :e8, when
14 ... dxeS the result is a draw (Lepeshkin). In
15 fxeS ltJhS (DJ the game Baluyev-Vadikan, corre-
Not 15...l2Je4? 16 ~xe4 ~xg5+ spondence 1976, Black instead
17 'ibg5 dxe4 18 ~f5 'ii'xe5 19 played 21...~5? and lost: 22 ~f5
~d6+ <;Pf8 20 'ili'xe5 l2Jxe5 21 l2Je4? (22 ...l2Jxe6 23 .i.xe6+ 'ili'xe6
~xb7. • 24 ltxe6 h6 with advantage to
16 't!Vh4 White) 23 .i.xe4 dxe4 24 l:[f6 !
The sharp try 16 e6?! is interest- 'ibg5+ 25 tbxg5 1'.d5 26 ltxa6
ing.\The cautious 16 ...1'.xg5+?! 1-0.
218 Virtuoso Defence

16 ... hg5+ 19 exd7 :Sd8

17 'ifxgS g6 20 g4 lllg7 (DJ
18 e6 (DJ Black has nothing better; aftet
Weigel-Hauemherin, correspon- 20 ...'iff4+ 21 'ifxf4 lllxf4 22 l%e7,
dence 1977 saw instead 18 g4 lbg7 20 ... lllf4 21 .i.f5 (followed by
19 e6 illc5 20 exf7++ ~xf7. .l:te7)or 20 ...f6 21 'iib4 lllg7 22 'IL:,7
9f 4+ (22 ...:Xd7 23 llle6) 23 l%d2
'ii'xd4 24 .i.xg6 'ilxd2+ 25 q;xd2
hxg6 26 g5, his affairs are in a very
bad state.

Now 18...llldf6 is hazardous, as

after 19 exf7++ q;xf7 20 :n the
threat of 21 g4 is hard to meet.
It certainly makes sense for Black
to exchange queens by 18...'iVf4+!? 21 .i.fS!?
19 't!Vxf4q)xf4 20 exd7++ ~xd7, An attractive-looking move.
as in the game Shakarov-Zhurav- Defending the d7-pawn temporar-
liov, correspondence 1976. After ily, White intends to double major
21 :n q)xd3+ 22 l%xd3 f5 23 g4 pieces on the h-file. The attempt to
White retains the better ·chances carry out this attacking manoeuvre
but Black's position is still defensi- at once is ineffective: :Z1 'fi'h6 lh.d7
ble. (21...1Vxd7?! 22 l%e3 'iVxg4 23
In two games in the 10th World .i.e2! and 24 l%h3)22 l%e3~6 23
Correspondence Championship, .i.f5 (23 l%h3'iff4+) 23 ...it:lxd4 24
a
Maeder chose different, riskier :Xd4 l%e7 (24 ... gxf5? 25 '#g5+
continuation: ~h8 26 1i'f6+ ~g8 27 gxf5 and
18 ..: 0-0?! wins) 25 l%h3f6.
 Virtuoso Defence 219

There was, however, another

plan of attack which was evidently
stronger; it involved pushing the
h-pawn. The game Estrin-Maeder
in the same tournament went 21
h4 ! l:.xd7 22 h5 1i'd6 23 .. h6 .i.c8
24 l:thl loe8 25 l:tdfl l:tc7 26 g5
'ile7 27 ltlc6 ! ! l:.xc6 28 hxg6 f6 29
gxf6 :t"xf6 30 l:tfgl .i.e6 31 gxh7+
'.Ph8 32 'iVg5! 1-0.
As the subsequent analysis will
show (if it is correct, of course),
White no longer has any advantage
after the move played. However, to the move he played was a mistake
reach this conclusion in advance is after all, and that he should have
quite impossible even in a postal played 22 l:te3. From analysing the
game, let al~ne over the board. position together, Zviagintsev and
21 ... .i.c6 (D) I came to the opposite conclusion.
The bishop is taboo; Black loses I would point out once again that
with 21...gxf5? 22 ll)xf5, or with the truth here can only be estab-
21 ... f6? 221i'h6 gxf5 23 l:te7 l:tf7 lished by analytical means; in prac-
24 l:tdel l:tdf8 (24 ...lhd7 25 tical play this is an insuperable task.
l:te8+) 25 l:txf7 '3;xf7 (25 ...lhf7 Let us look at the position aris-
26 l:te8+ l:tf8 27 l:.xf8+ '3;xf8 28 ing after 22 l:te3 .i.xd7 (D ).
~6+) 26 lDxf5 lDxf5 27 gxf5.
22 l:td3!?
Which rook should go to h3?
White's choice looks illogical,
since after 22 ....i.xd7 23 1i'h6 (or
23 l:.h3) Black will play 23 ...l:.feS
with tempo: White will have to lose
time removing his rook from e 1.
Sanakoev took this circumstance
into account, of course, but after a
deep study of the specific varia-
tions he nonetheless preferred the
text-move. Later he concluded that w
220 Wrtuoso Defence

a) Now 23 'ifh6 seems indi- a42) Sanakoev wanted to play

cated. Sanakoev gives these first 24 Afl. with a view to the winning
three variations: variations 24 ...toxd4 25 l:th3 l1fe8
al) 23 ...gxf5 24 gxf5 f6 25 l:tgl 26 .i.xg6 and 24 ...l:tfe8 25 tl\xe6
(25 l:te7 llf7 26 ltle6 is met by .i.xe6 26 llh3. Alas, there is a refu-
26 ...ltlxf5!) 25 ...:0 (25 ....i.c8? 26 tation: 24 ...'ii'c4!, and the fl-rook
l:.h3) 26 ltle6! .i.xe6 27 fxe6 lle7 is en prise.
28 l':lh3 is clearly better for White b) The attempt to reverse the
as Black must return the piece by move-order with 23 l:.h3 (D) (in-
28 ...l:txe6 29 'iixh7+ ~f8 30 'ifh8+ stead of 23 'ifh6) is interesting:
@e7.
a2) 23 .....i.xf5 24 l:th3 l2Jh5 25
lhh5 gxh5 26 gxfS l!ixh2 27 4)f3.
a3) 23 ...J:tfe8 24 lth3 gxf5 25
'ifxh7+ '.tf8 26 'iih8+ <3;e7 27
'i/ixg7 fxg4 28 :.n.
a4) Black's best defence is to
play 23 ...4)e6! (with the idea of24
l:th3? 1Wf4+;24 ~xe6? fxe6).
a41) In the course of the game,
Sanakoev thought that after 24
lbxe6 ..i.xe6 25 iLxg6 fxg6 26
ltxe6 'it'f 4+ 27 'ifxf 4 lhf 4 the end- B
ing was level. Afterwards it oc-
curred to him that the simple 25 bl) If now 23 ... toe6?, White
J:h3 guarantees White a very strong plays 24 lbxe6, answering 24 ...fxe6
attack, for instance: 25 ... .:.fe8 26 not with 25 ~xg6 'Wf4+ 26 'ifxf4
'it'xh7+ ~f8 27 'ii'h6+ (J;e7 28 ltxf4, but with 25 l:.xh7!. We have
°it'gS+. etc. This whole variation is already seen that the variation
of little importance, however, be- 24 ....i.xe6 25 'i!Vh6has dire conse-
cause it rests on the false assump- quences for Black. .
tion that after 24 ... fxe6 (instead of b2) White also keeps a plus in
24 .....i.xe6?) 25 ..i.xg6, White wins. the event of 23 ... hS?! 24 .i.xd7
That is far from the truth; the obvi- 'ii'xd7 25 .J:r.h4.
ous is ...iLc8 26 l:th3 :t.d7 stems b3) Yet with the rook on h3,
White's onslaught. Thus the ex- Black is justified in accepting the
change one6 gives White nothing. piece sacrifice: 23 ...f6! (or 23 ...gxf5
 Virtuoso Defence 221

24 °it'h6 f6!) 24 'ii'h6 gxf5. The

Point is that Black meets 25 gxf5
by the simple 25 ...ji,xf5, while af-
ter 25 'Wxh7+ ~f7 the black queen
obtains the important f4-square:
26 gxf5 (the threat was 26 ... l:th8)
26 ..."iff4+ 27 ~bl .:.lxf5. Even af-
ter the relatively best 25 l:.fl
'ffe5!?, White's compensation for
the sacrificed piece is plainly in-
adequate.
22 ... .i.xd7 B
23 'ihi6 rues
Black develops his rook with point in 24 ...'ifc4 because of 25 b3.
tempo, though once again 23 ...lbe6 Taking on f5 gives White a mating
deserved attention. White gets no- attack. In reply to 24 ...'iVeS, Sana-
where with,24 l:th3 'iff4+ or 24 koev had prepared quite a compli-
.:t'l 'ii'e5 ! ( this time, of course, not cated combination: 25 l:th3! 'ii'xd4
24 ...'ii'c4 ?). Winning back the pawn 26 ~xg6! iLe6! 27 iLxh7+ ~f8 28
after multiple captures on e6 leads rJ...d3'ife5 29 l:te3 'if d4 30 .i.f5 l:td6
to equality. The thematic line is 24 31 l:.dl li'c5 32 l:t.eel followed by
~xe6 fxe6 (24 ... .i.xe6? 25 l:th3) 33 'ifb8+. White recovers the piece
25 l:.h3 ! .i.c8 26 .i.xg6. An analo- and obtains a decisive plus.
gous position occurred in our 24 ... b3!!
analysis of 22 l:te3, except that the A magnificent riposte! Any cap-
white rook was then on dl. It is ture on b3 has its drawbacks. If 25
considerably better placed on e 1, ltlxb3, then at worst Black has
as becomes clear especially in the 25 ... .i.b5 26 l:l.c3 "ifxc3 27 bxc3
variation 26 ....:C.d7?!27 lue6 (when .i.xfl. If 25 axb3, there follows
27 ... rJ...g7fails to 28 .i.xh7+ l:txh7 25 ... 'ifa5 (threatening not only
29 l:tg6+). Black would have to re- 26 ...'ii'al+, but also 26 ...l:tel+) 26
ply 26 ...'iff 4+ 27 'ifxf4 l:txf4 28 @bl .i.b5. In either case the initia-
.i.xh7+-~g7, but here too the ad- tive passes to Black. The right con-
vantage is with White. tinuation was definitely 25 J:t.xb3
24 :rt! (D) "ifa5 26 c3, but even then, having
What is Black to do now? White repelled the immediate threats to
is threatening 25 l:[h3. There is no his king (the rook can no longer
222 Virtuoso Defence

reach h3), Black obtains counter-

chances.
25 l:.h3!? (D)
A brave decision! White breaks
through on the h-file after all, but
at a high price - the black pawn is
on the verge of queening, though it
may be possible to stop it with J:r.a3
orlbb3.

a) After 29 ...al'ii'+ 30@d2 Sa-

nakoev gives a beautiful variation
leading to a won ending for White;
30 ...1i'aa5+ 31 c3 lbh5 32 .:Xh5
gxh5 331i'h6+ ~e7 34 J:r.el+ .i.e6
35 .i.xe6 1i'xh2+ 36 lZ.e2 1i'xe2+
37 ~xe2 lZ.d6 38 lbf5+ ~d8 39
lbxd6 J:r.xe6+ 40 'it'xe6 f xe6 41
lbb7+. As a refutation of the at-
tack, he gives 30 ...1Wca5+ 31 c3
25 bxa2 1i'xb2+ 32 .i.c2 "ifbxc3+ 33 J:r.xc3
26 1i':xh7+ ~ 'f:..e7,but actually White wins in this
27 'ifb8+ @e7 line too: 34 'it'h4! lbh5 (34 ...@g8
28 'ifb4+ (D) 35 'flxe7) 35 gxh5 .:CS 36 J:r.f3.
This is the position Sanakoev b) Perhaps Black should try
was aiming for. He intended to an- 29 ...1i'f4+!? 30 J:r.xf4 al'ii+ 31
swer 28 ...~d6 with the simple 29 ciid2 'it'el+ 32 ~d3, and now not
J:r.33,stopping the pawn and main- 32 ...'ii'dl+ 33 <Ji>c3 .:tc8+ 34 ~b3,
taining a strong attack. He was not as indicated by Sanakoev (the king
afraid of 28 ...~8, if only because on the third rank can go to a2 to es-
White can settle for a draw if he cape the checks), but 32 ....i.xfS+!.
wants (29 'iih8+ ~e7 30 'iih4+). ff 33 gxf5, White has to reckon
He would also be justified in think- with 33 ...4Jh5, therefore he should
ing about 29 'it'f6!; play 33 lhf5 'ii'dl + 34 ~c3 'ii'el +
 Virtuoso Defence 223

35 ~b3. Black can choose be- 30 ... 'ii'c4!

tween 35 ...l:lbS+ 36 <at>a2 'ffa5+ 37 31 .i.d3 'ii'xb3
:a3 'ffxa3+ 38 ~xa3 (38 bxa3) 32 'ii'xf6+ ~d6
38 ...gxf5, and 35 ...lte3+ 36 l:r.xe3 33 1fe5+ <at>c6 _
'ii'xe3+ 37 ~a2 - and then either 34 cxb3 alW'+
37 ...l:td7 or 37 ... gxf5. However, all 35 ~c2· Was
these variations obviously favour 36 .i.xg6 .:rs
White. 37 .:C7 (D)
Unfortunately, Black has a much
tougher defence at his disposal.
28 ... f6
29 l:te3+ itle6!! (D)
29 ... .ile6 is weaker, for if (as in
the game) there follows 30 ~b3
'frc4 31 .i.d31i'xb3 32 'ii'xf6+, the
g7-knight is en pn·se.
Black wovld lose with 29 ...~f7?
30 .i.xg6+ or 29 ...'~f8? 30 .i.xg6 !
al 'ii'+ 31 ~d211f ca5+ 32 c3 'ifxb2+
33 lt)c2.
B

At this point White was still

feeling optimistic. His pieces are
very active, and his opponent's
material plus is not too • great.
Sanakoev analysed the first three
of the following variations:
a) 37 ...:Xf7 38 .i.xf7 ~c5
(38 ... ~c7 39 'iif6+ ~b7 40 'ifxd8
'ifc5+41 l:.c3 and wins) 39 .i.xd5+
~b6 40 'ifd6+ ~a7 41 b4 Wa4+ 42
.i.b3 itlxb3 43 'ii'c7+ ~as 44
'ifxd8+.
30 itlb3 b) 37 ...l:.de8 38 'iVf6d4 39 l:.e5 !
'I'.here are no combinations~ so 'iVd8 40 :.Xe6+ ! .i.xe6 (40 ...lbe6
White must retreat. 41 .i.e4+ ~b6 42 'ii'xd4+ ~b5 43
224 Virtuoso Defence

jj'c4+ 'i!i>b644 'Wb4+) 41 Jle4+ check after 38 ...l:1fe8 39 l:r.f6+

'i!i>c5(41...'iPb5 42 l1b7+ 'if->c543 '3;b7 40 Jlxe8 l:txe8 41 'ikxd7
b4+ ~c4 44 b3#) 42 b4+ <if.,xb443 l:txe3 42 1i'c6+ q.>a743 l1f7 l:.e2+
l:r.b7+ 'ita5 (43 ...'iti>c544 We5+ 44 'i!i>bl Wei+ 45 '3;a2 'ii'a5+
Jld5 45 b4+ 'iitc6 46 .i..xd5+ Wxd5 (45 ...l:r.e7??46 'i¥c5+).
47 Yikc7#)44 b4+ 'ii?a4 45 'ii?b2! 39 b4 11ra4+(D)
and 46 .i..c2#. 39 ...1Yh5? loses to 40 Jlf5.
c) 37 ....:lde8 38 'Wf6 Yikd839
lhe6+ l:1xe6 40 'if c3+ and the
black rook on f8 will be en prise
after 41 'Wb4+.
d) An interesting defence was
suggested by Vadim Zviagintsev:
37 ...<if.,b6!?38 'ii'd6+ .i..c6 39 Wxe6
l:r.de8! 40 'Wd6! (40 l:C.xf8?l:r.xe641
J:.xe6 'Wc5+; 40 lte7? l:r.f2+ 41
'ii?bl l:tfl + 42 'iitc2 'ifcS+) 40 ...l:txf7
41 l:txe8 .:lf2+ 42 ~bl l:tfl + 43
q.>c2JU2+ with a draw.
However, Maeder found an even w
stronger continuation.
37 ... lbc7!! 40 <itid2 ~b7
The f8-rook is defended indi- 41 Jtrs
rectly: 38 1i'f6+ <if.,b739 l:1xf8 From White's viewpoint, a bad
"ifc5+.White has to forget his am- symptom: he has to exchange
bitious plans once and for all and pieces, as otherwise it is his own
start fighting for the draw. Nor is it king that comes under attack. For
clear how he can succeed. example after 41 J:C.f6(threatening
38 "ilie7 l:r.g8! 42 l1a3), there is 4 l...~xg4 !. Nor
Again the most precise move. is 41 l:.c3 l:tge8 ! any good for
Black loses with 38 ... d4? 39 J:.e5 White, for example: 42_"ilid6l:[e2+ !
lbd5 40 l1xd5!, or with 38 ...l1h8? 43 ~xe2 Jlxg4+ 44 ~f2 l1xd6 45
39 b4 lhh2+ 40 q.>b3. A roughly l:tcxc7+ <if.,b8(but not 45 ... ~b6??
equal position would result from 46 l:tb7+ ~c6 47 l:r.fc7#)46 l:r.b7+
38 ... 1:lxf7 391:lc3+ ~b7 40 .i.xf7 @c8 47 1:la7'tfc6, or 42 'ii'c5 Wc6
lbb5 41 Ld5+ '3;a7 42 b4 '&b6 43 43 'ii'f2, when Black must avoid
.:lc5. The game ends in perpetual 43 ... 'Wb6? 44 'ii'xb6+ ~xb6 45
 Virtuoso Defence 225

.:.xd7, while a draw results from occurs, after which Black still has
43 ... 'i!Vxg6? 44 l:txc7+ r:i;xc7 45 enough advantage to win.
'ii'a7+; the winning move is 44 nc3+ ~d6
43 ...'ifd6!. 45 lU6+ ~eS! (D)
41 ... l:tge8
42 '1i'd6(D)

46 lb:c7 ~xf6
47 lhd7
42 ... 'iic6! After 47 JTLxd7 nh8, Black
42 .. Jbe3? is a mistake on ac- quickly works up a decisive attack
count of 43 lhd7 !. with the combined forces of his
43 'ii'xc6+ two rooks and the king.
43 'ii'xd7? lb.d7 44 .i.xd7 would 47 ... lhd7
fail against44 ... 'ii'h6! 45 .i.xe8 d4. 48 .i.xd7 l:th8
43 ... ~c6! 49 h3 ~e5!
In the case of 43 ... .i.xc6 44 h4, The king should stay in the cen-
the position would remain unclear tre while the rook copes with the
-the white passed pawns are very kingside pawns. On 49 ... r:J;gs?,
dangerous. To deal with them, it is play might continue 50 b5 axb5 51
important above all to exchange ~d3 l:txh3+ 52 ~d4 .:.b3 53 '.txd5
White's strong bishop; this explains l;.xb2 54 ~c5 b4 55 jLa4 with a
why Black takes on c6 with his draw.
king. Admittedly this allows his . 50 g5 l:tg8
opponent to regain some of the lost 51 h4 nhs (DJ
material, but a major simplification 52 g6
226 Virtuoso Defence

Or 55 .i.e8 l:th2+ 56 'iii>c 1 (56

<it;eI <j;e3 57 <iPfl c.1tf3!)56 ....:tg2
(56 ...<j;e3?! 57 i.c6) 57 i.f7 '.li>e4
58 b4 d4 59 b5 d3 and Black wins. •
SS ... l:tg4
56 .i.e8 q.,e3
57 b4 d4
57 ...:.Xb4? 58 .i.f7.
58 g7 Jhg7
59 'ita>b3 l:.gS
A good move, though not
obligatory. The immediate 59 ...d3
was also playable, for example 60
By some means or: other White .i.h5 l:tg5 6 I .i.d I J:.gI 62 .i.h5
has to swap pawns on the queen- l:.hl 63 i.g4 ~4.
side. The immediate 52 b5 is re- 60 <j;c4 d3
futed by 52 ...<Jtd6! 53 bxa6 (what 61 .i.a4 d2
else?) 53 ...~xd7 54 a7 l:ta8! (but 62 .i.b3 l:tgl
not 54 ... q.,c7? 55 h5 <iPb7 56 h6 63 bS l:tcl+! (D)
<i;;xa757 <i;;d3~b6 58 'iii>d4<it;c6 Not, of course, 63 ... dl'ii'?? 64
59 b4 ~d6 60 b5 ~e6 61 b6 and .i.xd 1 lhd 1 65 b6 and draws.
White saves himself).
52 ... l:.xh4
53 b5
There is hardly any point in 53
.i.c8, as Black could then choose
between taking both b-pawns and
winning the g6-pawn.
53 ... axbS
54 .i.xbS ~d4!
The most precise continuation;
the black king supports the ad-
vance of the d-pawn. The conse-
quences of 54 ...l:tb4 55 .i.d3 (55
.i.e8) 55 ... J:.xb2+ 56 'ite3 .l:tg2 57
~f3 are less clear-cut. 64 'ita>dS .:tbl
55 ~ 0-1
 Vinuoso Defence 227

"I have played about 300 games the 5th issue of the Danish maga-
by correspondence and won most zine Skakbladet for 1982.) I call
of them, but few of those victories the reader's attention to the follow-
have given me so much creative ing extract.
satisfaction as this unsuccessful at-
tack. The excitement of the imagi-
native contest took such a hold of
me that at a certain point the bare
result ceased to be all that impor-
tant; creativity occupied the fore-
ground ...
"In this game, heaven knows, I
did everything that was then in my
power. My opponent played better
- all praise to the winner! How-
ever, I conducted the attack with-
out heed fo~ the circumstances, B
and in the final analysis human be-
ings are responsible for their ac- Rivas - Mestel
tions, not for the result ... Of course Marbella Znnal 1982
it was madness to let the black (notes by Larsen)
pawn get to a2, but 'he who has
never done anything reckless is When annotating this game for Ek-
less wise than he thinks' (La strabladet [a Danish tabloid news-
Rochefoucauld). Surely creative paper], I was influenced by Rivas' s
pleasure counts for no less than analysis and· also by the editor
miserable half-points or even a full Dinesen, who was doing his best
point? And did not Caissa repay to hurry me. And so ... I confirmed
me a hundred-fold for those glori- that in this position Black was
ous deeds of recklessness which I doomed: 24 ... gxh5? 25 Wt'g5;
permitted myself not only in this 24 ....h6? 25 'it'g5 .i.xd5 26 .i.xg6+
game but in others that did not end fxg6 27 h5, and White wins (ac-
so sadly?"_ (Sanakoev). cording to Rivas, the sequel could
be 27 ...1i'xf6 28 hxg6+ <iPg8 29
Examining this game, I cannot 1i'xd5+ ~g7 30 l:.h7+ c.f.>xg631
help recalling a vivid article by 'ii'h5#- beautiful, without a doubt).
Bent Larsen. (It was published in I would point out one instructive
228 Virtuoso Defence

feature: the queen on g5 blocks the .:i.e2+ and 30 .:i.b5. We shall pres-
g6-pawn (this would normally be a ently see why I mention this second
pawn's job). possibility. The dl-a4 diagonal
Maybe Mestel overlooked some- might have been blocked!
thing here, seeing that he lost in Apart from this line, there is
another four moves. As he thought 26 ...'it'f8 27 .:i.xg6+ 'ith8 (27 ...fxg6
for a full hour over move 26 in a 28 h5) 28 i&.f5 ilxf5 29 exf5 ~h7
vain effort to find a defence, it is 30 tbg8 ! !. Perhaps this is just what
precisely here, at move 24, that a Mestel overlooked? And yet 30
diligent reader should be looking lt.Jg8 isn't hard to find if you ask
for Black's last hope. We will come yourself how White is going to
back to the diagram position, but prevent the defence 30 ...'fi'h6.
first let us see what happened in 27 idl ~h8 (D)
the game. Or 27 ..Jhe7 28 h5.
24 ... l:e8??
25 1i'g5!! :es
What else could he do? White
was threatening .ilxg6+ and h5.
26 lbe7! (D)

28 'ii'h6+ 1-0
Very pretty. But let us return to
move 24 - when Mestel clearly
B had plenty of thinking time - and
see what moves and ideas were
26 ... 'ii'g8 available aside from 24 ... gxh5,
The main variation is 26 ...:txg5 24 ....i.e6 and 24 ...lle8.
27 hxg5 'flixe7 28 fxe7 .:i.e6 and One possibility, for instance, is
then 29 i~t+ <:Jilg730 ~a4 or 29 24 .. :ii'e8. The queen supports the
 Virtuoso Defence 229

g6-pawn, and also eyes the square The idea of ...b3 (which was
e5 (as the rook did after 24 ...l:te8) there all along) in conjunction with
and the e4-pawn. There might fol- ...'ii'e8 (which has just become
low 25 !be7 c4 26 'ii'g5 l:tb5, but possible) is something that comes
we already know the refutation: 27 to light when you survey the posi-
Ji.xg6+! fxg6 28 h5. tion in desperation - when you see
So 24 ...'ii'e8 fails to save Black, that all the natural replies are un-
but it was worth looking at it all the satisfactory. At the fateful moment
same; eccentric moves sometimes you play 24 ... b3 ! ! without even
lead us to the right ideas. [Let us looking at 25 ~bl or 25 axb3.
recall what was said about Reshev- After 25 'iir'bl (the king is on a
sky - that he deliberately slipped light square!), there is of course no
into time-trouble after first analys- point in taking the a2-pawn. A good
ing all ( !) the tactical refinements move is 25 .. .'~e8, but 25 ... Ji.b7 is
of the position, and then played with also playable, so as to take on e4
complete assurance 'with his flag with check ·(25 ~bl? Ji.b7!? 26
dangling'. This is an obvious fabri- 'Wg5 Ji.xd5 27 Ji.xg6+? fxg6 28 h5
cation. He couldn't have grasped Ji.xe4+ 29 <it>al'ifxf6).
all the tactical refinements - new After 25 axb3, Black can oc-
ones would have cropped up.] But cupy the open a-file. The defensive
there is something else for us to idea ...l:txb3-h3 also suggests it-
think about: at every move since self, but unfortunately it does not
the seventeenth, the possibility of work. The simple 25 ...l:ta8 forces
... b3 has been in the position. the reply 26 <it>c2,and again the
There you are! At the moment king is on a light square. Black can
Black is not threatened with mate play 26 ...Ji.b7 or even 26 ... c4, but
in two, and on 24 ... b3 White has why give White another pawn for
cause for alarm about ...bxa2. If25 the exchange?
a3, we come back to the ... 'ii'e8 There is scope here for lengthy
idea: 25 ... 'ifeS 26 0.e7 'ii'a4!, and analysis. However, in practical
Black has unexpectedly come to play, the most plausible course of
life. The queen attacks c4, d4 and events is 24 ... b3!! 25 a3 'ife8!, and
e4, for example 27 !bxg6? 'ii'c4+!, now the optimist who is playing
or 27 Ji.f3 'i!i'c4+ 28 ~bl 'ii'c2+. WJ:ritewill have a long think-if he
White has a pawn for the exchange has the time. Gradually his ears
and s~me positional trumps, so the will go red, his breathing will be-
chances are about equal. come heavy, his knees will start
230 Virtuoso Defence

shaking slightly, and ... so will the won the game. Optimists on the at-
whole board. tack are very bad at-readjusting to
Level-headed defence saves changed circumstances.
many a point. I have seen Jonathan The trouble for annotators is
Mestel wriggle out of tighter cor- that games ending in a beautiful,
ners than this. I am convinced that overwhelming victory can be hard
after 24 ...b3 he would even have to analyse objectively.
 14 Errors and What Lies Behind
Them
Mark Dvoretsky

'To err is human.' It is a truism that (which he constantly replenishes)

no chess player has the gift of play- and, latterly, some powerful com-
ing faultlessly. Sometimes blun- puters have been working tirelessly
ders stem from the sheer difficulty to perfect his opening arsenal. As a
of solving the problem in hand. or result he gets into -difficult posi-
from some chance conjunction of tions comparatively rarely; in this
circumstances. Very often, however. department his experience is quite
mistakes conform to a distinct pat- modest. It should come as no sur-
tern; they result from some of your prise, therefore, that defence is one
failings as a player or as a person. of his relatively vulnerable points.
This applies not only to ordinary When he is compelled to defend,
chess players but also to the lead- Kasparov always does so as ac-
ing grandmasters and even the tively as he can; he immediately
World Champion. tries to undertake something, to al-
The most difficult thing for any ter the complexion of the fight by
of us is to handle unfamiliar types sharp play. This characteristic of
of situation in which we lack ade- his was discussed, in particular, by
quate experience. To some extent Grandmaster Evgeny Bareev in a
this deficiency can be cured by lecture he gave in 1992 at a session
purposeful training. but unfortu- of our chess school.
nately there are few players who
engage in this. In the diagram on the following
It is well known that one of the page, Black has the choice between
main strengths of Garry Kasparov exchanging queens and sharply sac-
lies in his deep, large-scale open- rificing a piece on e4 ( 17... 'iic7? is
ing preparation. For many years he
\
anti-positional; White continues
himself, the members of his team 18 ~ followed by l:.cl and tofl).
232 Errors and What Lies Behind Them

inferior ending. so he decided on

rather a dubious piece sacrifice ..."
17 lofxe4?!
18 fxe4 ltlxe4
19 1fxa5 .:XaS
20 ltJc3! (D)

Bareev - Kasparov
Linares 1992

In Bareev's opinion the right move

is 17.. Ji'xd2+!. Then 18 <ifiixd2?
lDfxe4+ is too risky for White, and B
so is 18 jlxd2?! (with the idea of
developing the knight on a3) ••simple and strong. If 20 ...ltJxc3
18...Jl.xg4 ! (better than l 8 ...ltlfxe4 21 bxc3 .:r.xd5, then 22 ltJf2 and
19 fxe4 ltJxe4 20 ltJc3) 19 fxg4 White will set up_ a light-square
tlxxe4 (intending 20 ...ltlxd5), and blockade. Without counterplay,
if 20 Jl.f3, then 20 ...ltlxd2 and Black's three pawns are not
21 ... e4. White would have to play enough for the piece. He needs to
18 ltlxd2 b5 19 ltlf2 l:lic8 with ap- have some dynamic weapon, such
proximate equality. as two connected passed pawns"
"We all have our own styles, our (Bareev).
own idiosyncrasies as players," There followed: 20 ... log3 21
Bareev remarked in his lecture. "In :gl loxe2 22 <itxe2 e4 23 .:act!
Black's place I would have resigned f5 24 gxfS lhf5?! (24 ...Jl.e8) 25
myself to the queen exchange. How- iof2 Jl.e8 26 l:thl il.bS+ 27 loxbS
ever. Kasparov doesn't like posi- .:Xb5 28 l:.c8+ @h7 and now, in-
tions where he hasn't any counter- stead of 29 .:r.ct1?.:r.xb2+ 30 .:.d2
play.·He didn't want to go into a a3!. with an unclear position - as
quiet and (as- he thought) slightly occurred in the game - White
 Errors and What Lies Behind Them 233

could have gained a decisive ad- was wholly out of place (though in
vantage with the simple 29 l::tbl ! the second half of the match An-
l:tfxd5 30 b4! axb3 31 axb3. and was demoralized and unable to
punish him for it).
In many cases active defence is 27 l:[d5!
precisely what holds out the most Obviously, accepting the Indian
promise, but this is by no means al- grandmaster's positional exchange
ways so. Any kind of one-sidedness sacrifice is extremely dangerous;
is a bad thing. There are times the attacking force that it gives
when you need to parry your oppo- White in the centre and on the
nent's threats calmly and cope queenside is too strong. Black
with the problems patiently and should have parried the threat of
accurately. A lack of flexibility in 28 ~c7 with 27 .. Jlec8! (not fear-
his methods of fighting makes a ing 28 llxe5 'ifxc4, and envisag-
player vulnerable. ing ... 'ifc6 at a suitable moment).
Black's position would still be un-
pleasant but by no means lost.
27 ... ~xd5?
28 exd5 1i'g6
Kasparov took the rook all the
same. Why? As I see it, the expla-
nation is that he was hoping for ac-
tive play of his own. His queen has
targeted the white rook as well as
the squares c2 and d3. The active
move ...e5-e4, attacking the bishop,
is coming ... Alas, these are all de-
lusions. The strategic strengths of
White's position count for far more.
Anand - Kasparov 29 c5 •e4
World Championship match (9), 30 jle2 lleS
New York 1995 31 'ilid7! llg5? (DJ
In Chernin's view, 31...e3 would
Interestingly, in his match with have put up much stiffer· resis-
Anand (New York 1995), Kaspa- tance, e.g.: 32 llfl llg5 33 jld3 e2 !
rov ~ore than once opted for ac- 34 ~xe2 llxg2 35 ~d3 llg 1+ I 36
tive defence in positions where it llxgl 'ii'xd3, and the position is
234 Errors and What Lies Behind Them

still sharp. 32 l:.gl is better, but

rather than transpose to the game
with 32 ...l:g5, Black could play a
different, more useful, move.

Anand - Kasparov
World Championship match ( 11),
New York 1995

29 axb4 l:tc4 30 l2Jb6??. I should

32 J:tgl e3 add that Anand only took a few min-
33 d6 l:tg3 utes over this sequence and played
34 1Vxb7 ii'e6 the fateful knight move almost in-
35 ~h2! 1.0 stantaneously. Why? On the one
Black resigned as after 35 ...1We5 hand, evidently, the immense nerv-
36 'it'xa8 he has no good way to ous strain - intensified by the emo-
make use of the discovered check. tions of losing the previous (tenth)
game - was taking its toll. On the
In the next diagram, the e7-pawn other hand, Anand is lacking in the
is under attack. After the natural discipline of scrutinizing variations
27 ...~f8 Black would have had an intensively and painstakingly. He
acceptable though somewhat infe- possesses remarkable intuition -
rior position. Instead, Kasparov many of his decision~ (including
played 'actively'. quite difficult ones) are taken
27 ... ..i.e6? quickly, and yet speed of thought
In the actual game his stratagem is partly at odds with correctness
paid off, because his opponent was and precision of calculation.
tempted into an unsound continua- Now there followed 30 ...lb:b4+
tion based on a fork: 28 b4? axb4 31 'iPa3.White assumed he was
 Errors and What Lies Behind Them 235

winning the exchange for a pawn - White could have reached this
after, for instance, 31...l:tbc4 - but position by force. Of course we will
had missed the terrible rejoinder not stop here. Some specific but not
31 .. .:Xc2!. He could only resign at very complicated analysis is needed
once (0-1), seeing that after 32 to take the variation to its logical
.:txc2 l:tb3+ 33 ~a2 l:te3+ Black conclusion. It is amazing that nei-
comes out two pawns up. ther Anand (during the game) nor
In this case Kasparov was just Kasparov (in his commentary for
lucky. In actual fact the move he lnformator) was equal to the task.
played was bad! By capturing on 32 ... lbb4+
e7 White could have forced a dou- If 32 ...l:tec8? (suggested by Kas-
ble rook ending with an extra pawn parov after the game), then 33 c3!
and excellent winning chances: (stronger than 33 l::te2 l::txb4+ 34
28 tbxe7! l::te8 ~cl l:tc6 35 l::ted2l::ta6) 33 ...l::txc3
29 lbdS .i.xd5 (D) 34 l::te2with 35 lhb5 to follow.
33 ~c3 (D)
Not, of course, 33 ~cl? f5 with
equality.

30 b4! axb4
31 axb4 B
White could also have reached
this position via the alternative 33 ... l::tc4+
move-order 29 b4 axb4 30 axb4 34 ~b3 l::tec8
:C4 31 lbd5. 34 ...f5 is more tenacious, but af-
31 l:tc4 ter 35 lhb5 l::td4 36 ~c3 Black is
\ 32 l:.xdS left with a difficult position.
236 Errors and What Lies Behind Them

35 l:te2 'active' line in place of the 'nor-

Black loses one of his pawns, mal' one.
and his chances of salvation are 32 ... lkS?
highly problematic. 33 :Xa7 gS
I am sure that if Anand had cal- Black seems to have everything
culated this variation accurately he under control; White loses his g-
would have played it. White's ad- pawn after 34 hxg5 l:ug5, or his f-
vantage here is greater than in the pawn after 34 g3 gxh4 35 gxh4
position he was trying to reach l:tf5. However, Anand finds a splen-
with 28 b4?. (If anything that posi- did rook manoeuvre to refute his
tion is drawish.) opponent's idea.
34 l:.a8! gxh4
35 l:.e8+! 'itid7
Evidently 35 .. /~fS was even
worse: 36 l:te4 @g5 37 a4!?.
36 l:te4 c3
At this stage White is faced with
rather a difficult choice. The In-
dian grandmaster simply decided
to win a pawn:
37 .l:.xh4?! cxb2+
38 ~b2 :lgS
39 a4
B However, in the position that
has resulted White's pawns are
Anand - Kasparov disunited and his opponent has real
World Championship match ( 17), chances of saving himself (given
New York 1995 the drawing propensities of rook
and pawn endgames). Indeed Kas-
In this position Black should parov eventually secured the draw.
play 32 ... a6, removing his pawn With a little more self-confidence
from attack. An important point is (his failures in the preceding games
that on 33 '.td2 (with the posi- were weighing heavily on him),
tional threat of @c3-d4) Black has Anand would most likely have pre-
33 .. ,c3+ r 34 bxc3 l:tc5, after which ferred 37 b4! l:tg5 38 l:txh4 .l:txg2
the game should be drawn. How- 39 @bl. In this case White has yet
ever, Kasp~ov again chooses an to win a pawn, but he holds a major
 Errors and What Lies Behind Them 237

positional trump - two connected The moral of these examples

passed pawns on the queenside. If (more of them could be added to
he manages to transfer his king to the list) is obvious. For chess play-
b3 (as in the variation 39 ... l:tf2 40 ers of all levels it is extremely im-
l:txh5 !? ltxf3 41 @a2), the posi- portant to analyse your own games
tion is definitely won for him. thoroughly and unearth the hidden,
Black's only counterchance (but I underlying causes of your errors.
doubt if it is sufficient) lies in This is always the indispensable
pushing his d~pawn to d4 with the first step towards the goal of elimi-
white king on bl. nating them.
15 Analysis of a Game
Mark Dvoretsky

We are going to look at an interest- 3 e3 lill'6

ing game played by Sasha Cher- 4 Jlxc4 e6
nosvi tov. He has annotated it in 5 ll)f3 c5
great detail, giving a wealth of 6 0-0 a6
variations and explaining what he 7 a4 lbc6
was thinking about during the 8 i¥e2 'ilc7
game, what he was afraid of, what 9 ll)c3 Jld6
points he missed. His analysis is 10 dxc5
not fault-free. Errors are of course Other moves to have been played
inevitable in cases like this; when are 10 b3, 10 d5 and 10 l:tdl. In his
you are examining complicated annotations Sasha sets out his de-
variations, it is not surprising if tailed views on the theory of this
you go wrong somewhere. How- variation. However, as his opening
ever, your mistakes sometimes - ideas are to some extent a personal
serve to reveal characteristic de- matter, there is no need to discuss
fects in your manner of thought, in them here. The problems arising
your approach to chess. I selected later will be of more interest to us.
this game for discussion for the 10 ... .i.xc5
very reason that some errors com- 11 h3?!
mitted both in play and in the notes 11 e4 is more thematic, though
seemed to me instructive. White then has to reckon with
11...lbg4 (threatening 12...lbd4)
Denisov - Chemosvitov or with an immediate l l ...lbd4.
Moscow Junior White wants to advance his e-
Championship 1991 pawn in complete comfort, but in
Queen's Gambit Accepted the opening every tempo matters;
with such slow play you cannot
1 d4 dS count on an advantage.
2 c4 /, dxc4 11 ... 0-0
 Analysis of a Game 239

12 e4 lLJd4! queen has left the c-file? Perhaps

13 lLJxd4 bd4 exchange the dark-squared bish-
The opening phase of the game ops with 17 .i.e3, but most likely
has ended favourably for Black. he will prepare f2-f4 by moving
He controls the central squares, the his king to h 1. Can Black make it
c4-bishop is obstructed by the e6- hard for him to carry out these
pawn, while the white e-pawn is plans?
kept in restraint and in some varia- 16 ... 'ii'g3!
tions even comes under fire (for in- An excellent manoeuvre! It is
stance 14 ~e3? .i.xc3 15 bxc3 much stronger than 16 .. Ji'e5 17
lLJxe4). Black only has to develop '1.>hllLJh5 18 ~g4.
his light-squared bishop, and his 17 c;i;>hl 'ifh4
position will be preferable. The queen is splendidly placed
14 Jld2?! on h4. From here it pressurizes the
Chemosvitov recommends 14 white e-pawn and at the same time
.i.d3, preparing to bring the other eyes his king. If appropriate, the
bishop toe~ knight will jump to g4. If 18 f4, then
14 b6 18 ... t2Jh5 19 .i.el lLJg3+ 20 ilxg3
15 :act .i.b7 'Wxg3 with the better chances for
16 .i.d3 (D) Black. Perhaps this was the least of
the evils for White.
18 :eel :t"d8
The white pieces on the d-file
are starting to feel uneasy. Now on
19 Jlc2 (with the idea of ~h2 and
g3, or 'iff3-g3), Chemosvitov sug-
gests l 9 ...g5 ! 20 'ii'f3 .i.e5 21 l:te2
l:td4. If 19 t2Jd1 (preparing 20
.i.c3), then l 9 ...b5 20 axb5 axb5
21 .i.c3 e5 22 'it'f3 .i.xc3 23 bxc3
(23 tbxc3 b4) 23 ...l%a2, and Black
has a considerable plus. Sasha con-
B siders that the relatively best de-
fence is 19 'i!Vf3!?.i.xf2 (19 ....i.e5
The threat of 17 lbd5 (or 17 20 lLJdl) 20 l:txf2 l%xd3 21 'ii'xd3
lLJb5) is completely obvious. But 'ifxf2 22 'i!Ve3!? 'if xe3 23 .i.xe3
what' will White do after the black luxe4 (23 ...iud7 24 l:tdl) 24 iuxe4
240 Analysis of a Game

.ilxe4 25 .ilxb6 with chances of

salvation (for example, 25 ...f5?! is
ineffective in view of 26 l:k 1 ~n
27 .ild4).
19 <it>h2? (D)

first of all consider the more solid

23 ...ltJe5 !. White's reply 24 li'e3 is
forced. Let us continue the varia-
tion: 24 ...'ifxe3 25 ~xe3 ~c4 26
~cl (26 .ilf2 lbxb2 27 .ilxb6 .:r.cs
is also dismal, for example 28 .i.d4
Now the combination in the last l:.c4 29 .:r.bl .:xd4 30 lbb2 l:td7)
line gains dramatically in strength. 26 .. J:td8 27 b3 ~e5 28 .ilf4 (28
19 .ilxf'2! - .ile3 l:td3) 28 ...f6 and Black will
20 :XO ltxd3 exploit his extra pawn with no par-
21 \i'xd3 lbg4+! (D) ticular trouble.
22 'itgl li'xf'2+ Why didn't Sasha play this? He
23 '.thl wanted to carry on the attack - out
Chemosvitov has conducted the of inertia, and for fear that White
first half of the game very power- might save the endgame thanks to
fully. He has completely outplayed the opposite-coloured bishops.
his opponent and gained an advan- The main point, though, is that he
tage quite sufficient for victory. underestimated his .opponent's
However, from this moment on, threat, imagining that the queen
another player seems to talce over. sortie to d7 was unplayable be-
How should Black proceed now? cause of the loss of the e4-pawn. In
23 ...li'h4 looks inviting-, but then actual fact, after 23 ...'ii'h4 24 'it'd7
the counter-attack with 24 'ii'd7 is ltJf2+ 25 '.th2 Jlxe4, White can
unpleasant. Therefore Black should capture on e4 and then fork two
 Analysis of a Game 241

pieces with 'it'c6 or 'ifb7. To go in 24 ....:bs? is thoroughly bad: 25

for such adventures you need to be 'ill.c7'it'd8 26 ~f4. Therefore Black
armed with the most precise analy- must choose between 24 ...l:td8 and
sis. I would certainly have stopped 24 ... ltlf2+.
at this point and discarded the With his last move Sasha made
whole variation (ifl hadn't done so a typical psychological mistake:
already, the moment I spotted 24 realizing he had obtained a deci-
'i!Vd7). sive advantage, he relaxed and
When exploiting an advantage played carelessly. It now dawned
you must pay strict attention to all on him that matters were not as
the opponent's active tries and pre- simple as they had seemed - and
vent any unnecessary sharpening his second psychological error fol-
of the fight. It may happen that lowed. (You will recall Tarrasch's
your advantage hangs by a hair, famous dictum: "Errors never come
and that to maintain it you need to singly.") As often happens, the
exert yourself and calculate some sudden appearance of difficulties
long, complicated variations. How- dismayed him and prevented him
ever, the situation here is clearly from coolly examining the varia-
different - Black has a sound extra tions - which in the present situa-
pawn, and there is a safe way to tion are quite complicated. He
preserve all his assets. even failed to solved this problem
23 ... 'itb4? later, when analysing the game at
24 'iVd7! (D) home.
What should Black do now? Chemosvitov rejected 24 ... l:.d8
25 'iVxb7 lud2 on account of 26
l:.e2. However, this is too early to
break off the variation; Black can
play 26 ... hS! (or 26 ... h6!), giving
his king some luft. His rook is ta-
boo, and his threats look quite dan-
gerous.
Incidentally, in this kind of posi-
tion the pawn is usually pushed
two squares rather than one, as it
may join in the attack. However, in
the present case both moves are of
B roughly equal value.
242 Analysis of a Game

White's reply is obvious: 27 alter the verdict (if 32 'iifiig3,then

'ifxb6 (DJ (not 27 'ii'xa6? lhb2! or 32 ... 'Wel+ followed by 33 ... a5 is
27 .. .l:tc2!). The next problem is to good enough).
decide how Black should utilize If White plays instead 28 c;i;ib2,
his trumps? Black has a splendid knight sacri-
fice: 28 .. .lbxh3! 29 gxh3 l:r.d3!with
an irresistible attack, for example
30 'ii'b8+ ~h7 31 ~gl l:txh3 32
l:th2 'ii'g5+!, etc.
If White tries to evade the mate
he ends up in a lost endgame: 29
lhd2 l£if4+ 30 'itg 1 'ile 1+ 31 ~h2
'ifxd2, or 29 'ii'b8+ ~h7 30 'ilg3
'i!i'xg3+ 31 <i.>xg3l:td3+ 32 ~h2
~g5 (or 32 ...~f4) 33 .l:k2 a5.
In place of 26 l:.e2, we still have
to look at 26 J:trl! (DJ. Now how
should Black continue?

The simplest solution - 27 ...llxe2

28 l£ixe2 'ifel+ 29 l£igl l£if2+ 30
~h2 l£ixe4 - seems to me uncon-
vincing. After 31 l£if3 (or 31 'ifb8+
~h7 32 l£if3) the co-ordination of
Black's pieces is to some extent
spoilt, and White simply threatens
a rapid pawn advance on the
queenside. For example, 3L.1i'g3+
32 ~gl .d6 33 'ii'xd6 l£ixd6 34
b4, with l£Jd4 to follow.
Let us try 27 ... l£if2+!. On 28
<i>gl, Black plays 28 ... l£ixh3+! 29
gxh3 :Xe2 30 l£ixe2 'ii'e 1+ 31 <itg2 White's answer to 26 ... f6? is
_.xe2+, and the queen ending is a not 27 'Wxb6? g;f7 (threatening
win. In this line the pawn is clearly 28 ...'iig3 and 28 ...l£if2+), but 27
better placed on h5 than on h6, 'fie?!. So Black must choose be-
but having it on h6 would still not tween 26 ...h6 and 26 ...h5.
 Analysis of a Game 243

A line that looks good is 26 ...h6

27 'ii'xf7+ ~h7 28 'ii'xe6 ltJe3 29
:gl '1i'g3, and White is defenceless.
However, 29 l:tgl? is not obliga-
tory; 29 ltJd5 ! ~xfl 30 W'f5+ leads
to perpetual check.
Black's attack can be strength-
ened with 26 ... hS! (instead of
26 ...h6) 27 'ii'xf7+ ~h7. If 28 °ii'f3,
the simplest course is 28 ...~e5 29
'ii'e3 l:txb2, coming out a pawn up.
Sharper play (although still fa- B
vourable to Black) results from
28 ...:txb2 29 e5 !? t2Jxe5 30 'iie3 'ii'xb6, but there follows 31. .. li)g4!
'i!i'c4! (30 ...ltJg631 ~4),andif31 (31...l:tf3 32 'iVxa6 is unconvinc-
l:tf4, then 31...1i'c6 32 tt)e4 'ii'c2. ing) 32 hxg4 'iVd2 33 'ii'xf2 'iixf2
The main variation is 28 ,_xe6 34:tgl (34 gxh5 'iixb2) 34 ...'iVn+
ltJe3. This time, 29 ~d5 doesn't 35 c;t.>h2hxg4 36 eS @g8!?, and
work: 29 ...ltJxfl 30 'ilff5+ ~h6 31 Black retains substantial winning
'ii'e6+ g6 and the checks run out. chances.
Instead White plays 29 llgl, in- Grandmaster Dolmatov discov-
tending to counter 29 ...'ilfg3 with ered another unexpected and inter-
30 'iff7 (as you see, having the esting way to hal\dle the attack.
pawn on h5 has a minus side as After 24 ...:td8 25 'ii'xb7, he sug-
well as a plus side). Black must re- gested 2S... h6!? (D) instead of tak-
ply 29 ... 'ii'gS!, but then White has ing the bishop.
30 g3! (D).How should the attack Now 26 :te2 :Xd2 would trans-
be continued now? pose into lines we have seen al-
30 ... h4 31 gxh4 'ifh5 32 :tg3 ready. However, in reply to 26 :t'l,
li)fl 33 :tg2 is useless for Black, the capture on d2 is not compul-
while 30 ...:Xb2 31 ltJd5 leads sory. Black has the much stronger
only to a draw. The strongest move 26 ... tt)f2+! 27 :txf2 'ifxf2, when
is 30 ... :tl!, depriving the white the bishop cannot escape.
queen of the f5-square and threat- The best defence is probably 26
ening 31.. .ltJfl or 31.. .ltJg4 32 ltJdS! exd5 27 :n (27 exd5 is no
hxg,_4hxg4! (but not 32 ...'ifd2? 33 good at all in view of 27 ...lL}f2+
:tg2!). The best defence is 31 and 28 ...li)e4), but then Black has
 244 Analysis of a Game

made beforehand, of which I have

already spoken.
24 ...
25 ..th2
As Evgeny Bareev observed, it
was not too late to revert to lines
like those we have analysed, by
playing 25 ...1:.d8! 26 'it'xb7 lhd2.
Only this time, Black should meet
27 l::te2(27 'ii°b8+ l::td828 'ii'g3
l2Jg4+!) not with 27 ...h5 or 27 ... h6,
w in view of 28 ";i'xa6! (28 ... 4Jg4+
29 ~gl!), but with 27 ... gS!. For
27 ... lbf2+ 28 'ith2 (28 ..tgl lbxh3+ example: 28 'ir'xb6 l2Jxh3! 29 gxh3
29 gxh3 ";i'g3+ 30 ~hl 'ilxh3+ 31 lld3, or 28 'ifxa6 .:.Xe2 29 'ti'xe2 g4
..tgl 'ii'g4+!, etc., is hopeless for 30 'ii'e3 (30 ~gl gxh3!) 30 ... gxh3
White) 28 ... t2Jxe4or 28 ... dxe4 29 31 'ii'g3+ 'ii'xg3+ 32 ~xg3 lbd3 33
.ilel e3, emerging with an extra gxh3 l2Jxb2 34 'itif4 f6, with an easy
pawn. win in the knight ending. Inciden-
This is another good moment to tally, with this order of moves, just
recall Tarrasch's view that you as in Dolmatov's line that we con-
have to resort to combinations in sidered earlier, Black deprives the
order to make up for your previous opponent of his best defence -1:lfl !.
errors. I should point out, inciden- 26 4Jxe4? (D)
tally, that an over-the-board deci- White misses a golden oppor-
sion to go in for such intricate tunity. He had to play 26 lhe4!
tactical complications is not all t2Jxe427 'ii'c6 'illg3+ (after 27 ...:cts
that difficult if you apply the pro- 28 ll)xe4, the d2-bishop is de-
cess of elimination - if you satisfy fended - this is why White had to
•yourself at the outset that all other take with the rook, not the knight)
tries are unpromising. Strange 28 ~gl l:.d8 29 .l2Jxe4. Then
though it may seem, Sasha saw 29 ...'ife5 is answered by 30 'i!i'xb6
that he would gain nothing from 'ir'd4+ 31 .ile3, while on 29 ...'ifr'b8
24 ...lbf'2+ 25 ~h2 .ilxe4, but none- (which Chemosvitov wanted to
theless quickly rejected 24 ...l:.d8. play), White has 30 ..i.e3 ! b5 31 a5
Such, evidently, was the psycho- h6 32 "ir'xa6 l:dl + 33 ~f2 'ii'e5 34
logical effect of the mistake he t2Jc3 with equality.
 Analysis of a Game 245

In endgames a rook often proves Of course it is better to play with

stronger than a bishop and knight the queens on. The white king is
if it manages to break into the en- exposed and can easily come un-
emy camp if\ good time, attack the der attack. Chemosvitov illustrates
queenside pawns, and set up a this verdict with the following sam-
passed pawn in that part of the ple variations: 33 ... h6 34 b4 (34
board. However, in this case the 'ii'xa6 'ifxb2) 34 ...:.d6 35 'ii'b8+
endgame is still a long way off, and @h7 36 b5 axb5 37 axb5 f5! 38
anyway it is Black's queenside :c4 e5 (38 ... 'ifd5!, with the fear-
pawns, not White's, that will come some threat of 39 ...1We5+,is much
under attack first. simpler - Dvoretsky) 39 :C2 'YWf4+
26 tbxe4 40 g3 l:td2+ 41 :xd2 'ii'xd2+ 42
27 'ifc6 'Wg3+ <li>gl'ife3+ 43 ~g2 e4 44 'it'f4?!
28 'iitgl 'iff2+ 'ir'e2+ 45 @gl 'ii'xb5.
29 @h2 'Wg3+ Instead Black d,ecided to go into
Repeating moves helps to save a rook endgame. This was a false
thinking time. assessment of the position! Even if
30 'iitgl 'Wf2+ the ending is a win, it is clear that
31 <li>h2 :.ds exploiting Black's advantage would
32 l:.xe4 'ifxd2 have been much simpler with
33. ii'xb6 (D) queens on the board.
Black has a sound extra pawn. 33 ii'd3?
Ought he to exchange queens or 34 l:te3 'Wd6+
retain them? 35 'ifxd6 .Il.xd6 (D)
246 Analysis of a Game

An interesting rook endgame How? For example, with 40 <it'g3

has arisen. In examining it, I came l::tb7 41 @f4 @d6 42 g4 and then
across quite a few questions which 42 ... f6 43 g5 <ii.ic544 .l:te3, or
Chemosvitov either ignores or an- 42 .. .c.t,c543 .l:tc3+ '.t>xb544 <be5,
swers unconvincingly in his notes. intending Ac8-g8.
36 ... g6?
A very strange move, on which
Sasha makes no comment at all. It
is clear that Black will have to ad-
vance his kingside pawns. Then
why not push a pawn two squares
right away - why waste a tempo?
The play may become sharp, a race
may begin, in which every tempo
will count. In this instance, Black's
delay may not alter the verdict on
the position. Later,however, a simi-
lar error will do so.
Of course 36 ...g5 !? was stronger
36 l:b3 than the text-move. Also 36 ...f5!?
White is preparing a4-a5 and looks no less attractive, aiming to
l:b6. This idea is attractive, but he bring the king to f6 and then play
should also have considered the ...h5-h4 and ...e5-e4. For example:
more conventional plan starting 37 a5 (37 l:.b7 a5!?) 37 ... (j;f7 38
with 36 b4 (the rook will station it- l::.b6q;e7 ! (gaining another tempo)
self behind the passed pawn). Af- 39 l:.b7+ c.t,f6 40 J:tb6 l:d2 41 b4
ter 36 ...~f8 37 b5 axb5 38 axb5 l:d4.
'l;e7 39 :b3, Black must avoid 37 :bS+
39 ...citd7? 40 b6 ciic8 on account Here Chemosvitov observes: "It
of 41 :C3+ ~b8 42 l:.c7. The right looks as if White had an immediate
move is 39 ... l:.b6, after which draw with 37 a5 l:td5 38 J:tb8+
Black should win, though it is not ~g7 39 b4 ~f6 40 l:tb6 l:d4 41
that simple. If his king goes to c5, lha6 l:txb4 42 l%a7(D)."
White replies l:.c3+; hence Black Roughly the same position can
will need to sacrifice a tempo with arise in many variations, and its
...l:.b6-b 7. Meanwhile White will
/
evaluation is important for a cor-
be preparing kingside counterplay. rect understanding of the entire
 Analysis of a Game 247

<itc5 44 ltxd6 'it>xd6 45 g4 ! fxg4

46 hxg4 ~c5 4 7 'iii>g2the pawn
ending is drawn) 42 b4 l:tb5 43
l:ta7 l:txb4 44 l:txa6 l:ta4 45 l:ta8
<it>f446 a6 e5 47 a7 e4 (47 ... g5 48
g3+ ~e4 49 1:[g8 or 48 ... ~f3 49
l:tf8) 48 l:tg8 l:.xa7 49 l:txg6 e3 50
:.e6, with a draw.
This line is interesting but not
forced. Right at the end, in place of
49 ... e3?, Black has the much
B stronger 49 ... <aii>e3!,
after which I
am not convinced that White can
endgame. ls it drawn? The black save himself - the e-pawn is just
rook is excellently placed to the too dangerous. On the other hand it
rear of the passed a-pawn. I don't is not entirely clear why White
see how Wlili,tecan oppose the ad- used up two tempi advancing his
vance of the black kingside pawn a-pawn before going after the g6-
armada. For example, 42 ... h5 43 pawn. In answer to 45 ...<iPf4,either
a6 l:ta4 44 <l.'g3 h4+ 45 ~f3 e5 46 46 .J:.a6 e5 4 7 ltxg6 or 46 l:.e8 e5
l:ta8 (46 <aii>e3lla2) 46 ...<aii>f547 a7 47 g3+ seems indicated.
l:ta3+ 48 <iPf2and then 48 ...g5 and
49 ...~f4. Note that the f-pawn is
well placed on its original square;
after 49 l:tf8 l:txa7 it is defended by
the rook. A typical mistake would
be 48 ... f6?, when .J:.f8 becomes a
much stronger answer to ...<aii>f4.
37 ... <lig7
38 a5 l:.d2 (D)
Another plan is to bring the
black king into the centre, but then
Black would have to give up one or
two kingside pawns. Chemosvi-
tov's analysis goes: 38 ...~f6 39
l:lb6 <it>e540 l:tb7 f5 41 lhh7 l:.d5 39 <lig3 'itf6
(after 41...<aii>d442 l:lb7 e5 43 l:tb6 40 b4 l:.b2
248 Analysis of a Game

41 l:tb6 ~e5? 43 lhh7

Pushing the kingside pawns. This is the consequence of
with 41.. .h5 or first 41...gS. would Black's omission of ...h5. The out-
be more natural. Incidentally, after come of the game is now in doubt.
...h5. the king's move to e5 gains in 43 ... ~e4!? (D)
strength, for when the white rook
goes to the seventh Black simply
replies ... f5, and the h-pawn will
not be en prise.
42 l:tb7
Chemosvitov comments: "Both
players missed the two-move varia-
tion 42 l:txa6 l:.xb4 43 Aa7 with a
draw." Well, we have a.lready dis-
cussed a similar position; in fact
after43 ...~6. Black should proba-
bly win. Another attractive line is
43 ... fS 44 1:txh7 lla4 45 l:ta7 g5
followed by ...l:.a3+.
When analysing endgames it is How is White to defend this end-
very important to make a correct game? Straightforward tries are
assessment of the key positions thar unsuccessful, for example:
can arise from different variations. a) 44 l:t.a7? f4+ 45 ~h4 ltxg2
Sasha misjudged one such posi- with an easy win.
tion, and his entire understanding b) 44 l:t.g7? f4+ (better than
of this endgame was distorted. 44 ...lhb4 45 .:Xg6 ~e5 46 h4) 45
I would point out that White has 9.>h2 f3 46 lbg6 e5 intending
no time to bring his king to the 47 ...f2.
queenside. A typical variation is: c) According to Chemosvitov's
42 q.>f3f5 43 ~e3 g5 44 ~d3 f4 45 analysis, 44 h4 ! would have saved
'iPc3 ltxg2 46 l:txa6 f3 47 Aa8 (47 the game. White frees the _h3-square
~d3 can be met by 4 7 ...l:.b2 or for his king. There could follow:
47 ...hS) 47 ...hS 48 a6 l:ta2 49 ~b3 44 ...f4+ 45 'ifiih3e5 46 l:a7 ~e3 47
l:.xa6! 50 l:t.xa6 f2 51 l:t.al g4 52 .:Xa6 e4 48 ltxg6 ~f2 49 a6 e3 50
hxg4 hxg4 53 b5 ~d5 ! 54 l:tc 1 g3 lli6 e2 51 l:txf4+ ~e3 52lli8 el'iV
55 b6 g2 56 b7 f1 'iV57 b8'ii' 'ii'xc 1. 53 lle8+ 9.>f'254 Axel 9.>xel and
42 ... / f5 then 55 9.>g4or 55 a7 lla2 56 ~g4.
 Analysis of a Game 249

44 l:.h4+? special positions which, although

A decisive loss of time. not perhaps arising by force, are of
44 ..• 'it>e3 interest in themselves. One such
45 l:r.c4 e5 position is analysed by Chemosvi-
46 l:r.c3+ (D) tov.
Or 46 !k6 f4+ 47 <it>h2e4
(threatening 48 ... f3) 48 <it>gll:r.bl+
49 <it>h2l:r.xb4 followed by ... ~f2
and ... e3.

In his view, this position is a

case of mutual zugzwang. It is not
hard to see that White loses if it is
his move (1 g3 f3; 1 h4 g4; 1 l:.g8
46 ... 'it>e2 .:Xa7 2 lhg5 ~f2 3 l'lf5 f3 4 gxf3
47 l:.c6 lbb4 e3). However, I cannot agree with
48 ltxg6 Sasha's conclusion that with Black
Or 48 l:.e6 f4+ 49 ~g4 f3+ 50 to move the game is a draw. White's
i;t,g3 f2 51 .:Xe5+ i;t,n. king is too awkwardly placed. I
48 f4+ would play 1...l:la2 2 <it>gllta6! 3
49 <iti>h2 e4 i;t,hI (3 'it>h2 J:al) 3 ... g4! 4 hxg4
50 :.Xa6 e3 i;t,{25 gS i;t,gJ with inevitable mate.
51 .:lb6 lh4 There is also another, more
52 a6 <it>f2 striking solution: I...<it>f2 2 l:.e8
0-1 g4! 3 h4 (3 hxg4 l:.xa7) 3 ... l:.a3!!
(3 ... f3! is also strong: 4 a8'W
When analysing complex end- l:.hl+!!) 4 'ili>hl (if 4 a8'if or 4
ings we sometimes come across .:lxe4, Black plays 4 ....:lh3+! 5
250 Analysis of a Game

gxh3 g3+) 4 ...lZ.xa7 S l:lxe4 lZ.al+ and went into an ending instead of
6 'wii>h2
g3+ 7 'wii>h3
f3 8 lZ.f4l:.hl + 9 playing for mate. Furthermore,
~g4~xg2. any technician would surely have
Let us sum up. In the second pushed his g-pawn two squares in
half of the game, Chemosvitov's one go, not just one. He would also
play was decidedly shaky and even have chosen the right moment to
his annotations, albeit thorough, advance the h-pawn; Sasha left it
were none too convincing. Two se- on h7, where it perished.
rious failings came across dis- 2) An unsure grasp of rook
tinctly here: endings. Some typical ideas and
1) Weakness in exploiting an plans escaped Chernosvitov's at-
advantage. We recall that in a won tention; his general assessments
position Black needlessly allowed and specific recommendations of-
his opponent to complicate. Sub- ten proved erroneous.
sequently he didn't even make the Sasha would be well advised to
effort to fathom the complexities do some work on rook endings -
that had arisen. Finally he failed to better still, on the theory and tech-
solve the problem of exchanging, nique of the endgame as a whole.
16 Some Achievements of Our
Pupils
Artur Yusupov

As has become traditional, at the 5 cxd5 d6

end of our book we present exam- 6 e4 g6
ples of achievements by students 7 f4 ~g7
of our school (their ages are given 8 e5
in brackets). The young people This is characteristic of Maxim
have played and annotated a whole Boguslavsky's style - right from
range of interesting games; some the opening, White chooses the
of them will be offered here for the sharpest line.
reader's judgement, with minor 8 ... dxeS
amendments to the analysis. The The alternative is 8 ... lbfd7 at
writer had a difficult problem of once.
selection; after all, nearly every 9 fxe5 lbfd7
young player produces good ex- 10 e6 lbf6?
amples of attacking chess. Nor is This is a serious mistake al-
this surprising; attack, risk and ready. According to theory, a bet-
imagination are naturally associ- ter line is 10...fxe611 dxe6'ife712
ated with youth. But let the games lbd5 'ifxe6+ 13 'ife2 1i'xe2+ 14
speak for themselves. .._xe2 0-0 15 lbc7 lbc6 16 lbxa8
lbb4 17 lbf3 lbc2+ 18 ~d 1 lbxa I
Boguslavsky (14) - Lepin 19 .i.c4+ ~h8 20 J:el, as in a game
Moscow 1989 Lputian-Magerramov, USSR 1979.
Modem Benoni At this point, according to Maia
Chiburdanidze and Eduard Gufeld,
l"d4 ~6 20 ... ~b6 21 ~xb6 axb6 22 J:e7
2 c4 e6 would have led to an unclear posi-
3 lbc3 c5 tion.
4 dS exdS 11 .i.bS+ ~8
252 Some Achievements of Our Pupils

1 l...We7 is also dangerous in quite good too) 20 ...<iii>xf8

21 tbe6+
view of 12 JH4 fxe6 13 d6+ wn .i.xe6 22 .i.xb7, winning material.
14 tbf3. 16 t2Jf7 (D)
12 ltif3 fxe6
12 ... a6 looks a little more pre-
cise, although after 13 .ile2 fxe6
14 0-0 exd5 15 tbg5 Wg8 (not
15 ... .itf5?? 16 l:.xf5; Black also
loses with 15... d4 16 iib3 'Wd7 17
.i.g4) 16 .i.c4 b5 17 .Jtxd5+ tbxd5
18 tbf7 White has a very promis-
ing position.
13 0-0 exdS
14 tbg5 Wg8
It was worth thinking about
14... h6, although in that case, after B
15 'ii'xd5 'ifxd5 16 tbxd5 hxg5 17
~xg5, White regains his piece and 16 ... .i.d4+?
keeps the initiative in the ending. In a difficult position Black
Now, however, the king's rook goes wrong and is mated elegantly.
is shut in the comer, and White has The unfortunate thing is that a fine
a chance to carry on his attack vir- queen sacrifice which Boguslav-
tually unhindered. sky had prepared remains on the
15 ltixd5 sidelines. Black's best is 16...'iie7;
15 JTLc4!?b5 16 .itxd5+ ©xd5 then after 17 ©h6+ ilxh6 18
17 li:Jf7 is not bad either; the idea is ~xd5+ ile6 (bad alternatives are
to answer 17...'ild7 with 18 ©h6+ 19 .:n+'ti'xf720 ilxh6+
18...c;f.>g7
3Lxh6 19 .i.xh6, with decisive and 18...'i¥e619 .itxh6) 19'ifxe6+!!
threats. ~xe6 20 .i.xh6 a fantastic position
15 ... arises, where White has only one
When defending your king, you piece for the queen and yet the ad-
should usually try to exchange vantage is on his side - because
queens. In the present case, this Black' s rook is shut out of play and
would have brought no particular his king is in a mating net. The fol-
relief: 15...'ii'xcl.516 'ifxd5+ ~xd5 lowing variations, based on Bo-
17 .i.c4 ild4+ 18 <iii>hl~g7 19 guslavsky 's analysis, support this
.itxd5 l:.f8 20 .J:r.xf8(20 ©f7!? is paradoxical conclusion:
 Some Achievements of Our Pupils 253

a) 20 .. .'iff7 loses at once to 21 4 t2Jc3 g6

Jtc4. 5 d4 Jtg7
b) 20 ... t2Jd7 21 l:tael 'it'd5 (or 6 Jte2 0-0
21... 'Wxe1 22 Jtc4+) 22 l:te7 'it'd4+ 7 0-0 t2Jbd7
23 'it>hl t2Jf6 (23 ... t2Jb6 24 l:te8+) The Schlechter Variation of the
24 lhf6 and mate is inevitable. Griinfeld has arisen by transposi-
c) 20 ... t2Jc6 21 l:tael and now: tion. Black's last move is consid-
cl) 21...'ii'xa2 22 l:te4! (22 ered inaccurate, since White can
Jtxc6? bxc6 23 l:te7 'ii'xb2 24 l:.ef7 now exchange on d5 without wor-
is weak.er because of 24 ... 'Wal !) rying about the black knight going
22 ... 'iff7 23 Jtc4 'ifxc4 24 l:.xc4 to c6 - its best square in this line.
t2Jd4 25 g4, and despite his two ex- Thus, in Botvinnik-Blau, Tel-Aviv
tra pawns Black's position is unen- Olympiad 1964, White gained a
viable. clear plus after 8 cxd5 cxd5 9 'ii'b3
c2) 21...'iid5 22 b3 t2Ja5 (if e6 10 a4 b6 11 Jtd2.
22 ... t2Je5, then 23 J:txe5) 23 l:te7 8 b3 e6
'ii'd4+ 24 ~h 1 and Black has no In a game against Kantsler, Tep-
answer to the threat of 25 Jle8. litsky found an antidote to another
c3) 21. ..'if f7 !? 22 b3 (22 ltxf7 of Black's possibilities; on 8 ... b6,
is weaker: 22 ... <;,txn 23 Jtc4+ there followed 9 a4 a5 10 cxd5
~f6 24 l:tfl+ ~e5 25 Jlg7+ 'it>e4 t2Jxd5 11 t2Jxd5 cxd5 12 ~a3 .:es
and Black equalizes) 22 ... t2Ja5 (if 13 l:.cl Jla6 141'.b5 ! with the bet-
22 .. .'ifxfl+ 23 l:txfl t2Je5, White ter game.
has 24 ltel a6 25 .i.fl t2Jf7 26 9 'ii'c2 l:.e8
~c4, tying up all the black pieces) 10 ~b2 aS
23 .i.e2 'ii°d7 24 .:tdI 'ii'xd 1 25 11 ltadl t2Jh5
l:txdl, continuing the king-hunt. 12 Jta3!
17 'Wxd4! cxd4 Up to now White has just been
18 t2Jh6+ 1-0 developing his pieces sensibly. At
this point he reacts to his oppo-
Teplitsky (14)- Parutin nent' s plans and takes prophylactic
Tashkent 1989 measures against ...f5, which would
Griinfeld Defence now be answered by 13 .i.d6, con-
trolling the dark squares.
1 t2Jf3 d5 12 ... b6
2 c4 c6 Black revises his plan, but his
3 e3 t2Jf6 knight turns out to be poorly
 254 Some Achievements of Our Pupils

placed on the edge of the board (Dr 16...b5 !?. After 17 g4 b4 18 lcie4,
Tarrasch's famous dictum inevita- he should avoid 18... bxa3 in view
bly comes to mind!). White achieves of 19 lcid6 llb8 20 lcixe8 'ii'xe8 21
a good game by simple means: he .i.xc4, but might try a recommen-
prepares a central advance. dation ofDvoretsky's to muddy the
13 lU'el i.b7 waters: 18...lbhf6!? 19 exf6 lbxf6
14 e4 llc8 20 lcixf6+ 'ii'xf6 21 .i.cl c5.
15 'ii'd2 dxc4 (D) 17 exf6!?
Interesting play, although the
quiet 17 bxc4 would also have en-
sured White the better chances.
17 ... lohxf6
18 .i.xc4
This move leads to great com-
plications. White gives up two
bishops for a rook and pawn. The
consequences of such an exchange
are usually very hard to assess cor-
rectly. In many cases, particularly
in the middlegame, the two pieces
w prove stronger, since they can cre-
ate more threats against the enemy.
16 e5! In this position, Teplitsky rightly
An unconventional decision. took into account the activity of his
Such moves are very easy to miss. major pieces - which will seize the
White is now threatening to win a only open file - and the weakening
piece with 17 g4. The 'automatic' of Black's castled position; these
16 bxc4 would give Black more factors outweigh the potential
chances of counterplay after 16...eS strength of the bishops, which at
17 d5 .i.f8 !? (but not 17...~5? 18 present are dozing .
.i.xc5 bxc5 19 d6 .llb8 20 lcia4 18 ... bS
with a decisive plus for White). 19 .i.xe6+ lhe6
16 ... f5 20 lhe6 b4
16...cxb3 would be met by 17 g4 21 'ii'e2!? bxa3
b5 18 lcie4 b4 19 lcid6 bxa3 20 If 21...bxc3, then 22 l:.e7 c2 23
~xb7 'ile7 21 lbd6, with advan- 'ifxc2 with an attack.
tage. Black probably had to opt for 22 .llel .i.f8
 Some Achievements of Our Pupils 255

22 ... ttlf8 is bad because of 23 25 'We6+ c:i;h8

.:.e7 .:lc7 24 'ili'c4+. 26 l:.xd7 l:.e8
23 lbgS ~b4 (D) 27 lhh7+!
Black overlooks his opponent's Destroying Black's last illu-
striking reply. However, 23 .. .l:tc7 sions. White now obtains a deci-
could similarly be met by 24 l:td6! ! sive material plus, and Teplitsky
1'.xd6 25 'fre6+ c:l;g7 26 'fixd6 confidently conducts the game to
~g8 27 'ii'e6+ c:l;g728 'ili't7+ c:i;h6 victory.
29 loe6 'ifeS 30 'i'g7+, with a win- 27 l2Jxh7
ning attack. 28 'ii'xe8 l2Jxg5
29 'it'xg6 l2Jf7
30 .:le3 'iig7
31 'if.e6 ~d6
32 l:th3+ ~g8
33 lbe4 'ii'xd4
34 lbr6+ <MS •
35 1We8+ c:l;g7
36 l:th7+ 'it.>xf6
37 'iixf7+ 1-0

Boguslavsky (15) - Morozevich

Moscow Junior
Championship 1990
King's Indian Defence
24 l:td6!!
Freeing e6 for the decisive inva- 1 d4 lof6
sion by the white queen. 2 c4 g6
24 ... 11i'f8 3 loc3 .i.g7
After 24 ... ~xd6 25 W'e6+~g7 4 e4 0-0
26 'ibd6 (but not 26 "fit7+? c:i;h6 5 lof3 d6
27 lbe6, on account of 27 ...ltg8), 6 .ile2 eS
Black is helpless against the con- 7 0-0 'iVe8
certed pressure of the white pieces; This line is rarely seen; its chief
there is no· adequate defence against merit is that the game now leaves
the threats of 27 l2Je6+ or 27 l:te7+. the well-trodden paths.
Black's attempt to counter-attack 8 dxeS dxeS
in the'e-file_ proves futile. 9 .i.e3
256 Some Achievements of Our Pupils

In answer to 9 ltJd5, Bogus- 11 b4 fS?!

lavsky gives 9 ...ltJa6 10 ltJxf6+ Premature activity. With incom-
hf6 l 1 il.e3 "fle712 lbd2 lbc5 13 plete development, it is extremely
b4 l2Je6 14 cS l:.d8 15 'ifc2 .i.gS, dangerous to open the game.
with an equal game. The last word 12 exf5 gxf5
spoken on this system by theory is 13 .i.h6 iixh6
9 b4!?, played by Ivanchuk against 14 'ii'xh6 'it'g6
Judit Polgar at Novgorod 1996. 14.. Jli6 was worth considering.
After the continuation 9 ...c6 10 b5 15 'ife3 a5
'iie7 11 a4 l1d8 12 il.a3 'ii'e8 13 16 a3 l1e8
'ifb3 il.g4 14 a5 a6 15 bxa6 ltJxa6 17 c5 'it'g7? (D)
16 'ii'xb7 the game turned clearly This is a serious mistake. Al-
in White's favour. though behind in development,
9 ... , lbrd7?! Black makes another move with a
Theory prefers 9 ...b6, probably piece already in play. He should
not without reason. After 10 ltJd5 have continued with 17...f4, and if
lba6 White has tried out various 18 'ii'e4 (or 181i'd2 e4 with coun-
continuations, but has not gained terplay) 18 ... lbf6 19 jixg6+ hxg6
the advantage with any of them: 20 1'.c4+ rj;g7 21 lbg5, then (as
a) 11 1i'c2 lbg4 12 .il.d2 c6 13 Dvoretsky has shown) 21. ..i&.fS
l2Je3 ltJxe3 14 iixe3 lbb4 with un- (preparing ...ltJa6) 22 tjJf7 il.d3! 23
clear play, Gostisa-Kupreichik, iixd3 '3;;xf7 with a defensible po-
Belgrade 1988. sition.
b) 11 il.g5 ltJxe4!? 12 il.e7 c6
13 .i.xf8 'ii'xf8 gives Black good
compensation for the sacrificed
exchange, Vucicevic-Kupreichik,
Belgrade 1988.
c) The line 11 ltJd2!? lbd7 12
ii'a4 .i.b7 leads to an unclear po-
sition, Dreev-Gelfand, European
Under-20 Championship, Arnhem
1988/9.
10 'fid2
The immediate 10 b4!?, followed
by ~5 and lbd2-c4, is also not bad. w
10 ::. c6
 Some Achievements of Our Pupils 257

18 Jlc4+ <;;tbS Of course there is no doubt about

19 ltlgS l:te7 White's advantage. yet it is often
It is too late now for 19...f 4? on hard to decide whether the moment
account of 20 ltlf7+. for tactical operations has come.
20 ltle6 'ii'f6 Here White had a good chance to
Boguslavsky intended to an- improve his position further with
swer 20 ...'ii'f7 with 21 li:Jc7 'iixc4 23 l:tad 1. S~ll, the exchanging
22 ~xa8 axb4 23 axb4 'iixb4 24 combination which Boguslavsky
~b6 (24 ltla4 is also possible); a carries out, and which is reminis-
sample continuation is 24 ... ~xb6 cent of the ideas of the previous
25 cxb6 f 4 26 °i¥d2 'if xb6 27 l:ta8 game, is also good: White begins
.C.d7 (27 ... 'ilic7 28 J:txb8) 28 'ii'c2 the tactics just when his oppo-
'iic7 29 :Cl b5 (29 ...~a6 also fails nent's pieces are least prepared for
to save him: 30 li:Jb5 'if d8 31 ~a7 co-ordinated action.
:C7 32 l:r.dl 'ii'e8 33 'ii'd2) 30 ~b5 23 ... cxd5
'ifb7 31 lh.b8 'i¥xb8 32 'i!ixc6, and 24 ~xdS f4
Black's position collapses. This intermediate move in no
21 li:Jc7 l:.a7 way alters the verdict on the posi-
22 bS! tion. The black centre pawns re-
Proceeding to the decisive ac- main under fire from the major
tion. pieces.
22 ... a4(D) 25 'iid2 'ii'gS
26 ltlxe7 'i¥xe7
27 :Cel
Black is unable to unravel his
tangle of pieces on the queenside.
His pawn-hunting excursion can
be explained by the good old prin-
ciple of being hanged for a sheep
rather than a lamb. Boguslavsky
quickly finishes the game with en-
ergetic play.
27 ... 'ii'xc5
28 l:.acl 'iVxa3
w If 28 ...'ii'f8, White simply plays
\
29 Jle6 ~b6 30 Jlxc8 ~xc8 31
23 ltl7d5!? lhe5 and wins.
 258 Some Achievements of Our Pupils

29 .i.e6 'ilirs therefore prepares to unpin his e-

30 'ilic3 pawn.
30 llxe5 is possible too. 7 ... ~6
30 ... b6 Zviagintsev-Frog. Moscow 1989
Nor does 30 ...'ii'c5 save him. in illustrates what can happen if
view of 31 'ft'al 'ji'f8 32 J:.xe5. Black renounces the fight for the
31 .:XeS 'ilif6 initiative: 7 ... lhbd7 8 g4 .i.g6 9
32 llgS! lc!xc4 e6 10 .i.g2 .i.b4 11 0-0 0-0
A nice stroke. though not com- 12 a5! lc!d5 13 'ii'b3 b5 14 axb6
plicated. As taking the queen al- lc!7xb6 15 e4 lc!xc3 16 bxc3 .i.e7
lows immediate mate. Black can 17 f4!. White has successfully car-
only try to postpone this inevitable ried out his plan to seize the centre.
fate by a couple of moves. 8 g4 .i.g6
32 h6 9 .i.g2 lob4
33 llg8+ c.fi>h7 10 0-0
34 'ji'd3+ 1-0 10 e4 would be answered by
10...'ifxd4.
Zviagintsev (13) - Nachev 10 ... .i.c2!?
Voskresensk 1990 Betterthan i0 ...lt)d7?! 11 l2)xc4
Slav Defence with advantage to White, Gelfand-
Khuzman, USSR 1987.
1 d4 dS 11 'ilid2 .i.b3
2 c4 c6 12 loe4!
3 ~f3 -~f6 Obscure complications arise
4 l2x3 dxc4 from 12 'iif4!? h6 13 .i.e3, Levitt-
5 a4 .i.g4 Flear, British Championship, Ply-
A somewhat risky system. mouth 1989. Against 12 a5, with
which Nikolic has championed in the unambiguous threat of pushing
a number of games. Without a the pawn further, current theory
doubt. White has to play most en- recommends 12 ...e6, ignoring the
ergetically to cast doubt on it. threat. After 13 a6 'ifc7. 14 axb7
6 ~ .i.hS 'ilixb7. Campos Moreno-Rogers,
7 h3! Manila Olympiad 1992, White still
The most unpleasant variation has to prove that he has enough
from Black's point of view. White compensation for the sacrificed
wants to seize the centre with his pawn.
pawns at a favourable moment, and 12 ... loc2
 Some Achievements of Our Pupils 259

12...fue4? 13 'ii'xb4 ~d6 is bad Ji.f4+ ~xc6 23 'ii'xc4+; l 8 ... ~b4

because of the striking retort indi- 19 Ji.f4) 19 ~xc6+ 'ii'xc6 20 Ji.xd5
cated by Gelfand and Kapengut: Ji.xd5 21 lhd5+ ~c8 22 :td3 !.
14 'ii'xb7!! and White wins.
13 ~5
If 13 loxc6, then 13...1ib6! (Gel-
fand and Kapengut).
13 ... 'ir'xd4
13... ~d5, as played in Bellon-
Pomes, Playa de Aro 1994, requires
further tests.
14 ~b7 ~dS
14...1i'xd2? is a mistake because
of 151'.xc6+! 1Wd716 ~xd7 ~xd7
17 .:td I f5 18 .:txd7 ~xal 19 .:td8+
9-il 20 .:Xa8, Khenkin-Sapis, Len- w
ingrad 1989. ,
15 WgS! 16 .ae3!!
Zviagintsev's pre-game prepa- A most unpleasant move for
ration sets Black some unpleasant Black to have to face. It turns out
problems. The alternative 15 ~c6 that capturing the bishop is bad;
'ii'xd2 16 Ji.xd2 is less dangerous 16... ~dxe3?? loses at once to 17
on account of 16...e6 (Gelfand and Ji.xc6+, while after 16 ...lbcxe3 the
Kapengut), with a roughly equal f-file is opened and White obtains
game. an extremely dangerous attack: 17
Now White threatens both 16 fxe3 and then 17...f6 18 l:txf6 gxf6
~xc6 and 16 'iff5. 19 'ii'h.5+ (Horvath) or 17...'ir'xb2 18
15 ... e6!? (D) llabl 1i'c3 19 .:xt7, when 19...1'.c2
Against the uncompromising is answered by 20 'ifxg7! ~xg7 21
15... f6!?, Zviagjntsev intended 16 ~d6+ rl;d8 22 ~xc6#.
1i'h5+ g6 17 ~xg6 hxg6 181Wxh8 16 ... 'ifxb2
. g5 19 h4! with the initiative. If in- 17 .t.cS!
stead 15... 'ifb6, then according to White increases the pressure.
his analysis a strong reply is 16 17 ... hc5
'iff5 'ifxb7 17 'ifxf7+ ~d8 18 .:tdl 17... ~xal? loses to 18 hd5
c3 (18 ...~xal? 191'.xd5 Ji.xdl 20 cxd5 19 ~d6+. In reply to 17...h6,
1'.xc6 1Wxc6 21 ~xc6+ rl;c7 22 White had prepared 181i'f4 !! f6 19
260 Some Achievements of Our Pupils

'ii'e4 'iixe5 20 ~g6+ 'it'd7 21 ~f7+ 'it'xc6 24 ~xd5+ exd5 25 l:tf6+

<;!i>c8 22 ltia5 l:tb8 23 ~xf8, with a and Black is helpless.
very powerful attack. 19 ltixc5 0-0-0!
18 'ifxg7! (D) In such a position it is easy to
overlook something. Thus, after
19... ltixal? 20 ltixe6! 'iia3! 21
~xd5 cxd5 22 ltic7+ ~e7 (or
22 ... @d8 23 'iif6+ @c8 24 ~c6)
23 'ii'g5+ f6 24 'ii'g7+ Black loses
at once.
20 l:tabl ! 'i¥d4
Again best; 20 ...'ifc3 is weaker
due to 21 ltie4, and so is 20 ...'it'a3
in view of 21 ltixb3 cxb3 22 ltixc6.
21 lticd7?! (DJ
Against stubborn defence it can
B be very hard to play the attack
faultlessly. A simpler line was 21
This zwischenzug, with its mate ltixb3! cxb3 22 l:.xb3, when the
threat, is the point of White's play. contrasting situations of the two
18 ... lU8 kings would surely tell. The opera-
Relatively the best continuation. tion that White has in mind gives
A queen sacrifice with 18... 'ibe5 his opponent unexpected saving
19 1fxe5 .i.d4 would bring no re- chances.
lief, because the threats created by
the queen in conjunction with the
knight would be too strong. For in-
stance, 20 'ifd6 .i.xal 21 'ihc6+
'itf8 22 ltid6 'itg7 23 ~xd5 exd5
24 lt:)f5+ would end in mate.
The insertion of 18...~xf'2+ only
opens up another line of attack: 19
'it.>h2!(19 l:xf2 'ifxal+ 20 'it.>h2
l:tf8 21 :.Xf7 leads to unnecessary
complications after 21...'1Wxe5+)
19 ...l:f8 20 ltid6+ 'it.>e
7 21 ltixc6+
'itxd6 22 'it'xb2 ~c5 23 'ii'e5+
 Some Achievements of Our Pupils 261

21 ... lt:)c3? following game is a case in point -

Black thinks he has to return victory was gained without any
blow for blow. With his king ex~ outward show, by methods that
posed, this leads to a quick catas~ look perfectly simple. Yet such·
trophe. Of course, 21...J:.xd7? is simplicity is deceptive; finding a
bad: 22 'itxf8+ .l:.d8 23 'ii'xd8+! quiet move to strengthen your po-
'iii>xd824 lt:)xc6+. The sole defen- sition is often harder than striking
sive possibility lay in 21.. ..:tfeS ! ! a tactical blow.
22 e3!? (22 .i.xd5 exd5!) 22 ...'ifc3
23 lt:)c5!? l:e7 !, and if 24 lt:)e4, Nikitin - Makariev (15)
then 24 ...f5 !. True, White would CIS Junior Championship,
have good chances of victory in Junnala 1992
this line too. He would continue 25 King's Indian Defence
"f//xe7Vxe5 26 'ifc5!? (26 Vg5 is
also worth considering) 26 ... fxe4 1 d4 lt:)f6
27 'ir'xc6+ lt:)c7 28 .i.xe4 with the 2 c4 g6
better game, 1forexample: 28 ...l:.d5 3 lt:)c3 iJ..g7
29 iJ..xdS 'ti'xdS30 •xdS lcixdS 31 4 e4 d6
.:Xb3 cxb3 32 l:bl lt:)c3 33 J:[xb3 s f3 0-0
.!LJxa434 l:b5. 6 .i.e3 lt:)c6
22 lt:)xf8! lt:)xbl 7 lt:)ge2 a6
23 .i.xc6 'fih6 8 a3
23 ... 'iff4? is no good in view of The main line is the natural 8
24 lt:)xe6. 'illd2, but the move chosen here,
24 'illxf7 c3 preparing qt.ieenside play, also con-
24 ...lt:)d4 is relatively better, but tains some poison. Ilia Makariev,
is adequately met by 25 .i.d7+ however, is well prepared for this
~b8 26 l:l.xbl. The move played tum of events, and plays as theory
loses even more material. recommends.
25 .i.d7+ 1-0 8 .i.d7
9 b4 'ti'b8!?
When we talk about attack, this Black prepares ... b5 without
naturally conjures up thoughts of hurrying, since in this line it is sim-
beautiful ~ombinations and finely pler for White to react to his oppo-
calculated sacrifices. · In many nent's operations than to carry out
cases, though, the attacker suc- his own plan. An immediate 9 ... b5
ceeds without any brilliancy. The would be met by 10 cxb5 axb5 11
262 Some Achievements of Our Pupils

d5 ~e5 12 ~d4, and if 9 ...e5, then the open file in his hands would as
10 d5 ~e7 11 g4 lbe8 12 c!cicl f5 yet play no particular role.
13 c!cib3, Korchnoi-Stein, USSR 11 ... bxc4 (D)
Championship, Leningrad 1963.
10 1!i'd2
If 10 lbcl, the reply 10 ...b5 is
quite in order, since after 11 cxb5
axb5 the b5-pawn is indirectly de-
fended (12 ixb5 c!cixb4). If 10 d5
~5 11 ~d4, then l l ...c6 12 dxc6
( 12 f4 is met by bringing a knight
to g4; 12 ie2 cxd5 13 cxd5 :cS)
12...bxc6 13 ie2 a5, with counter-
play.
10 ... b5
If the pawns are now exchanged
on b5, the threat of capturing on b4
arises. Black can also play Boleslav- 12 b4
sky 's recommendation: 10....:CS!? On 12 :bl, Black was planning
(this move is very useful if White 12...e5 13 d5 ~7 (13 ... lbe7, aim-
removes his knight from e2, since ing for ...c!cie8and ...f5, is also play-
...e5 then gains in strength) 11 g3 able) 14 a4 c6, with counterplay.
b5 12 c5 a5 13 l:.bl e614 ~f2 'iib7 It was worth thinking seriously
with equal chances. about 12g5!?. Then 12...~Swould
11 g4 be met, not by 13 f4, in view of
A bold decision, but White is 13...e5 ! 14 dxe5 ig4 15 exd6 (or 15
conducting his attack on too broad lbg3 dxe5 16 f5 c!cid4)15...c!cixd6,
a front. On the other hand, con- but by 13 h4! e5 14 d5 (14 h5!?)
tinuations like 11 d5? c!cie512 cxb5 14... ~e7 15 h5 with unpleasant
c!cic4,11 cxb5 axb5 12 d5? ~xb4 threats. Black would probably have
and 11 lbcl e5 are none too im• to opt for 12... c!cih5!? 13 c!cig3! e5
pressive. A playable alternative is 14 lbxh5 gxh5 (14 ... exd4? 15
11 c5!? a5 12 l:r.bl axb4 13 axb4, c!cixg7dxe3 16 'ifh2) 15 lbd5 (15
maintaining his central position. dxe5!?) 15... exd4 (15 ... c!cixd4?16
Organizing counterplay for Black ixd4 exd4 17 c!cif6+) 16 ~f6+,
would then be a good deal more and now, on Dvoretsky's advice:
complicateo ( 13...dxc5 14 bxc5); 16...~h8 17 ixd4 ( 17 lbxd7?
 Some Achievements of Our Pupils 263

dxe3) 17... ~xd4 18 'ifxd4 i.e6,

with prospects of queenside coun-
terplay with ... c5 or ... a5.
12 ... hS
It was essential to put a brake on
White's kingside advance. The un-
expected 12...a5 13 b5 ~b4? ! is too
pretty to be true. White can choose
between the unpretentious 14 axb4
axb4 15 :XaS bxc3 (15 ...'iixaS?!
16 ~a2) 16 ~xc31i'xa8 17 .i.xc4
with the better game, and the more
refined 14 ~c 1 !? threatening l:tb 1.
13 gS ~h7 15 ... e5
14 l:tbl ?! (D) Black starts the central counter-
This move seems to hand the attack.
initiative to--the opponent. Instead, 16 fxeS?
14 f4!?, threatening 15 d5, is more 16 d5? is a mistake in view of
consistent. If 14...a5, then accord- 16... exf4. Now, however, Black's
ing to Makariev's analysis White central position is strengthened
should continue 15b5 ~a716l:tbl even more. It was time to think
(16 a4 c6!) 16...11fe817 a4 ~cS 18 about development and prepara-
~g3 ~b6 19 f5. It definitely tion for castling with 16 i.g2.
makes sense for Black to sacrifice 16 dxeS
a pawn by 14 ... e5!? 15 dxe5 i.g4 17 dS ~a7
16 exd6 i.f3, with a double-edged 18 a4
game. Parrying the obvious threat of
14 ... 1i'd8! ...~b5, White allows a more sinis-
By modestly bringing his queen ter plan. However, it is hard to rec-
back, Black prepares counterplay ommend anything more suitable.
in the centre. Neither 18 i.xa7 :Xa7 (threaten-
15 f4 ing ... f6), 18 i.c5 l:teS nor 18 d6
Black would answer 15 d5 with cxd6 19 •xd6 iob5 would bring
15... lbe.5 16 ~d4 e6! 17 f4 (17 White any comfort. The pawn Black
dxe6 c5!) 17 ... exd5 18 ~xd5 (18 happened to snatch on c4 is look-
fxe~ dxe5) 18... ~g4 with numer- ing more and more like a healthy
ous threats. extra one ...
264 Some Achievements of Our Pupils

18 ... ~c8! 23 hxg5 'ife7

The knight heads for d6, where 24 'ii'd2 l:f 4!
it will not only defend the c4-pawn The decisive blow! After Black's
but also pressurize White's centre. 'quiet' 20th move his attack pro-
Black's advantage is growing. ceeds on oiled wheels.
19 ~g3 ~d6 25 ..i.xhS gxhS
20 ..i.e2 (D) 26 ~s l:lg4
27 lhg4 .i.xg4
28 ~g3 :rs
29 g6 ~r6
Jo ..t.cS ~s
0-1

To conclude my story of achieve-

ments by our youngsters, I would
like to give another, more recent,
game by Vadim Zviagintsev. It was
declared the best game in Informa.-
tor 62.

Cifuentes -Zviagintsev (18)

20 ... 'ile8! Wijk aan 'Zee 1995
Black would like to play ...f6, Semi-Slav Defence
after which White's position on the
kingside should collapse. How- 1 d4 dS
ever, the immediate 20 ...f6? would 2 c4 e6
be answered by 21 ..i.xh5 ! or 21 3 ~f3 ~6
~xh5 !, with wholly unnecessary 4 ~c3 c6
complications. Black's subtle pro- 5 e3 ~bd7
phylactic move prepares the deci- 6 'ilc2 b6!?
sive offensive without loss of 7 ..i.e2
tempo. 7 ..i.d3 ! ..i.b7 8 0-0 is more ener-
21 'ii'a2 getic; if 8 ...dxc4? !, then 9 ..i.xc4 c5
Stopping his opponent's most 10 'ii'e2.
obvious and least dangerous threat. 7 ..i.b7
21 f(, 8 0-0 !£.e7
22 l:tgl fxgS 9 l:tdl
 Some Achievements of Our Pupils 265

In Zviagintsev's view, 9 b3 is 15 tl)h4

more precise. Cifuentes's aim of exchanging
9 0-0 Black's blockading piece will not
10 e4 dxe4 be achieved; 15 a4 at once was bet-
11 fue4 'iic7! ter. It was also worth considering
It makes sense to prevent .i.f4. the simple 15 .i.g5 or even the
12 ~c3?! (D) hyper-active 15 'if f5 !? .
This way White cannot count on 15 ... g6
any advantage. If he wanted a An immediate 15 ... .i.d6 would
comfortable game, he shouldn't have led to roughly the same situa-
have avoided exchanges. Either 12 tion, but Black didn't want to give
lbxf6+!? or the simple 12 jlg5 c5 his opponent active possibilities
would have given approximate after 16 tl)fS .i.xh2+ 17 ~h 1 .i.d6
equality. 18 ~e4 - even though the pawn
sacrifice doesn't look correct:
18 ... ll)xe4 19 'iixe4 q)f6 20 'fi'f3
.i.e5.
16 j\_h6 lU"eS
17 'ii'dl?!
At this point 17 a4 !, to restrict
his opponent's queenside play, was
imperative.
17 ... .i.d6
With his thematic move Black
counters his opponent's elemen-
tary threats of 18 d6 and 18 lbf5.
B 18 g3 bS
19 .i.f3
12 ... cS If 19 lbf5, then 19 ... b4 is un-
13 dS?! pleasant.
A risky conception, but 13 ~b5 19 ... b4
'ifb8 14 g3 cxd4 15 ~bxd4 l:te8 19 ... ll)e5 also deserved atten-
also promised White nothing. Per- tion.
haps he should have completed his 20 ll)e2
development with 13 .i.g5. White reckons on creating coun-
13 exdS terplay after a3 or lbf5.
14 cxdS a6 20 ... ll)e4
266 Some Achievements of Our Pupils

Again Black had a chance to 25 ~xf2 'Wh3 26 .t.f 4 'ii'xh2+ 27

play 20 ...loeS. The text-move al- ~g2 .i.xf 4 28 loxf 4.
lows White to bring his king's
knight into the game and strengthen
his position.
21 'ii'c2 lodf6
22 log2!
The knight is heading for e3, and
on to c4 if the opportunity arises.
22 ... 'iVd7
23 loe3
In the event of 23 .t.f 4, Black
planned to decline the bishop ex-
change by 23 ....t.fS!, with some-
what the better chances. B
23 ... l:ad8
Zviagintsev brings his last re- 24 ... ~!
serves into the battle. The combi- This sacrifice is merely the prel-
nation 23 ...loxf2? 24 <il.>xf2 'ii'h3 ude to the real combinative fire-
25 ~f 4 .t.xf4 was premature on works. The white king is drawn
account of 26 lt:)xf4 (but not 26 into the centre, into the line of fire
gxf4 :Xe3 ! 27 ~xe3 lt:)g4+). Black of the black pieces.
also gains nothing from 23 ... 'ii'h3 25 ~ lhe3! (DJ
24 .t.f4 l:tad8 25 loc4.
24 .t.g2? (DJ
White covers the h3-square, and
in some lines threatens t2-f3. How-
ever. despite its natural appear-
ance, this manoeuvre proves to
have a tactical flaw. There was.not
much promise in 24 .t.f 4 .i.fS ei-
ther, as on 25 loc4 Black has the
simple 25 ....i.xd5. White should
have followed his opponent's ex-
ample an~ brought his rook into
play, because the sacrifice on f2
doesn't yet work: 24 l:acl loxf2?
 Some Achievements of Our Pupils 267

26 1'xe3?! 30 ... J1..xd5+ 31 J:xd5 1Vxd5+ 32

If White had foreseen the fate in _.e4 was dearly weaker, though
store for him. he would surely have 30 ... .i.xf4 31 ii'e4!? 'ifxe4+ 32
preferred 26 ..txe3 log4+ 27 ~d2 <i;xe4 .ilg5 and also 30 ...f5!? were
loxh6 28 ~c 1 'fie7 - though even perfectly playable alternatives.
then the chances are with Black, 31 'ifc4 (D)
who has a pawn for the exchange There is no other defence against
and good attacking prospects. 31 ....itxd5+.
26 lbg4+
27 <iPf3• lbxh2+
28 q.,a lbg4+
29 ffi
White was probably hoping his
opponent would repeat moves;
after all, Black is a rook down!
However. Zviagintsev keeps find-
ing new attacking resources.
29 ... 'iVe6!
30 1Lf4
Although this move loses, it
should perhaps not be condemned.
Other continuations would most
probably have led to the same re- 31 ... 1fe3+!!
sult: The most elegant solution, lead-
a) 30 1i'e4 'ilfxe4+ 31 ~xe4 ing to a forced mate - although the
:CS+. prosaic 31 ....i.xf 4 would also have
b) 30 .i.cl c4! 31 'fie4 (or 31 won:
~f4 loh2+ 32 c;t>f2.i.c5+ and a) 32 gxf4 'ife3+ 33 ~xg4
mates) 31 ...1fxe4+ 32 ~xe4 lof2+ .i.c8+ 34 f5 (34 ~h4 •n+ 35 lbg3
33 ~d4 lbxdl with a material and J:e3) 34 ...J1..xf5+35 ~h4 1Vf2+ 36
positional plus. lbg3 .:.e3.
c) 30 .i.g5 J1.e7 31 ~f41Ve3+ b) 32 loxf4 ibh2+ 33 ~f2
32 ~xg4 J1.c8+ 33 loe61Vxg5+ 34 'fte3#.
~ fxe6· with a winning attack. In my view, a dual solution like
30 ·- J:e8! this hardly detracts from the aes-
TJ.µs looks the most energetic thetic effect of Zviagintsev's won-
way to continue the attack. Instead, derful attack.
268 Some Achievements of Our Pupils

32 .he3 34 ... h6+!

Mate follows even more quickly The finishing touch; 34 ...~g7
after 32 ~xg4 .i.c8+ 33 ~g5 (33 would have prolonged the game in
<it>h4.i.e7+) 33 ...l:te5+. view of 35 J:lhl (or 35 'ir'xc5).
32 lhe3+ 35 'itixh6 :es
33 <itiixg4 .i.c8+ 0-1
34 ~g5 There is no defence against mate
Or 34 ~h4 Ji.e7#. next move ( ....i.f8# or ...J:h5#).
Solutions to Chapter 1
Exercises
1. Berg-Hort, Biel 1985 Wc6, while if 2 ..tiie1 Black wins
with 2 ...c3. The game continuation
Black has the advantage as his was 2 g4 hxg4 3 fS g3! 4 fxg6 g2 5
king is more active. His basic ~f2 ~e6 6 g7 ~f7 0-1. White re-
threat is ... ~c6-b5. White's main signed in view of 7 b5 c3 8 b6 c2 9
counterchance lies in the pawn- b7 gl 'if+! 10 Wxgl cl'ii'+.
break g3-g4!, but at the present There is nevertheless a way for
moment it clearly fails. It can only White to save himself:
be carried out with the white king 1 ~el!! (D)
on the e-file. In that position the
king is within one move of stop-
ping Black from queening a pawn
on either the c-file or the g-file.
The most natural king move, 1
'it>e3?, proves weakest, since after
1...@c6 ! 2 g4 hxg4 3 f5 gxf5 4 h5
f4+! it is not hard to see that Black
will queen first. Nor does it help to
play 2 ~d2 <li'b5 3 g4 (3 <li'c3 f5)
3 ... hxg4 4 f5 g3, or 2 'it>e2'it>b5 3
g4 f5!.
The drawback to having the B
king on e3, then, is that White ex-
poses himself to the tempo-gaining 1 ... <li'c6
... f5-f4+. Realizing this, White After 1...<li'd5 2 <li'e2! we reach
played 1 <li'e2?(with a view to the the same zugzwang position as be-
drawing line l...@c6? 2 g4! hxg4 3 fore, but this time it is Black's
f5). However, after 1 ... Citi>dS!
he un- move. Now 2 ...<iti>c6is answered by
expectedly found himself in zug- 3 g4!. If instead 2 ... <li'd4, White
zwang. We already know what hap- should not play 3 g4? f5 4 gxf5
pens after 2 ~e3 'it>c6 or 2 <it>d2 gxf5 5 'iPd2 c3+ 6 <li'c2 ~c4 7 b5
 270 Solutions to Chapter 1 Exercises

~xb5 8 i;i,xc3 ,_;,c5,but simply 3

<itiid2!with a draw.
2 g4! hxg4
2 ...f5 3 gxf5 gxf5 4 ~d2.
3 fS gxfS
4 hS g3
4 ...c3 5 h6 c2 6 ~d2 leads to
similar play.
S h6 g2
6 ffi c3
7 h7 c2
8 h8._. gl'ii'+
9 ~1 cl'ii'+
10 <ifi'f2 , 'fkxf7 'flc7 or 5 .. J:k7) 4 ...l:tc7!?,
The position is drawn. and White's attack peters out (after
King and pawn endings are an 4 ...'ifc7 5 'ii'xc7 l:txc7 6 f7 he still
excellent training ground for the has some saving chances based on
technique of analysing varia- 6 ....i.f8 7 l:tcl).
tions! 3 'fkxh6 .:C7!!
Not, of course, 3 ... J:.d7? 4 f7
2. Bobrov-Platonov, Chemovtsy l:txf7 5 :Xf7. However, following
1963 the text-move, after 4 f7 Black wins
with 4...l:txf7 5 l:txf7 l:td1+ 6 ~g2
1 1i'h5 ._.g8+. White actually played 4
This occurred in the game. .i.xc2 and resigned after 4 ...'ilb7+.
1 ... .bg5 Conclusion: the tempting com-
2 .i.e4 .i.h6!! (D) bination 1 'fkh5? is unsound. White
There is no doing without this should simply have played 1 l:th5,
move; on 2 ...'fkc7, 2 ...iia7, 2 ...:C7 maintaining a strong attack.
or 2 ...l%d7White plays 3 f7.
It is very important to find out at 3. Jansa-A. Sokolov, BieUnten:.o-
once which attacking line White is nal 1985
going to choose: 31i'xh6 or 3 •rs.
If 3 1i'f5, then 3 ...l:tc7? 4 f7 White wants to play his bishop to
:Xf7 5 'ilxf7 is hopeless for Black. g5. The only question is whether to
However, he has the pretty defence do it at once or after exchanging
3 ...l%d7! ! 4 1txd7 (4 f7 .:Xf7 5 pawns on the queenside.
 Solutions to Chapter J Exercises 271

In the game, White played 19 In contrast to the game con-

axb5?! axb5 ( 19 ...exf 4? 20 bxc6 is tinuation, the a-file is closed and
bad for Black) 20 .i.g5 (20 .i.d2 White keeps his extra pawn.
was worth considering, for exam- Let us see whether avoiding the
ple 20 ... b4 21 lod5 and the b4- pawn exchange gives the oppo-
pawn is attacked, or 20 ...0-0 21 nent any new possibilities. Black
.i.g5 !?) 20 ... b4! 21 .i.xf6 bxc3 can try 19 ... lt}d?, counting on 20
(otherwise 22 lt}d5 with advan- .i.xe7? lt}cS! 21 'ii'a2 (21 'ifb4a5!)
tage) 22 .i.xe7 'it'xb3 (but not 21...'.txe7 with a good game. How-
22 .. .ci;;xe7? 23 'it'xc3) 23 cxb3 ever, the intermediate exchange 20
~xe7 24 bxc3 .:U! 25 b4! (25 axb5 ! puts Black in a difficult posi-
J:lbl ..r.b8) 25 .. .ltxc3 26 J:lcl lhcl tion, for example 20 ... ~xg5 21
27 .:.Xcl 'itd7 with equality. bxc6, or 20 ... axbS 21 .i.xe7 lt}c5
Now let us look at the immedi- 22 °it'b4.
ate 19 .i.g5. He probably ought to choose
19 .i.g5! (D) 19...bxa4!? 20 'ii'xb7 .i.xb7, though
after 21 .i.xf6 gxf6 (21.. ..i.xf6 22
.:.Xd6) 22 lt}xa4 White has the bet-
ter endgame chances. For example
22 ...l:.cS 23 lt}cJ, followed by J:Cd2
and lt}dl-e3.

4. Timoshchenko-Vaganian,
USSR Championship, First League,
Baku 1977

Black's h6-pawn is en prise. He

can defend it with his king, but
B should also consider the active con-
tinuations 20 ...~3 and 20 ...'ifg3.
19 b4 Vaganian decided on a combi-
20 ~6 bxc3 nation without working out its con-
21 he7 'ifxb3 sequences to the end: 20 .. ..i.xhJ?
21. ..'irxe7 22 'ii'xc3 .i.xa4 23 21 'ifxh6 (threatening both 22
'ifa3 is no better for Black. 'ii'xh3 and 22 lt}g5) 21 ... 'it'g3 22
,\ 22 cxb3 'itxe7 'ii'xh3 'ii'xf2+ (22 ....i.xt'2+ 23 ~I
23 bxc3 ~xe 1 24 :.Xe I is unsatisfactory for
272 Solutions to Chapter 1 Exercises

Black) 23 'iii>hl'ii;g7 (23 ...'ilt'xc2 24 24 ..:ti'xc2 25 l2)g5~ (25 ....:Jis

4:\g5 !). Now the threats of 24 ...l:r.h8 26 .:Xt7+ '.tg8 27 l:r.h7!)26 ~7+
and 24 .. :ti'xc2 look dangerous, but '.tf6 27 'ihi4! l:r.h828 Jha+ 1-0.
White goes over to the counter- Black resigned due to 28 ... 'ifxf2
attack: 24 lll1 ! (D) 29 4:lh7+ and 30 'ifxf2.
From the test position, Black has
two other continuations of roughly
equal value:
a) 20...'.tg7!? 21 l::teS(21 l:tadl?
~xh3) 21 ... ..i.c6.
b) 20 .. :iVg3!? 21 4:ld4! (the al-
ternative 21 'ifxh6? fails against
21 ....i.xf2+ 22 r;i.,h1 .i.xel and then
23 ltJg5 'ii'h4 or 23 .:.xel .:r.fe8)
21 ...~4.
In either variation Black stands
worse but is quite able to defend
himself.
Solutions to Chapter 4
Exercises
1. Smyslov-Gurgenidze, USSR ~g4 ttJg6 with a clear draw. When
Championship, Tbilisi 1966 a passed h-pawn appears, Black will
be able to give up his knight for it,
45 h4! (D) provided his king can get back to b8.

2. Jochelson-Belavenets, corre-
spondence 1974-9

25 ... l:.8xe6! (D)

It is essential to fix the black

pawn on the vulnerable square h5,
so as to attack it with the bishop
and create a dangerous passed h-
pawn when the occasion arises.
Black will scarcely be able to 26 dxe6 'ii'f3! !
guard both his weaknesses, a7 and 0-1
h5. White should win. (With most of these exercises
The game continuation was 45 the answer consists of a single
~dS? h4! 46 .i.e2 lfil8 47 '.t>e4(if move, but sometimes it's a con-
White goes after the a-pawn, his tinuation of two moves or more. In
king will be shut in the comer by such cases the trainer makes the
.. .c~c7) 47 ... ~gS 48 ~dS @£6 49 reply and starts the clock again -
 274 Solutions to Chapter 4 Exercises

and so on until the full solution is 4. Miles-Makarychev, Oslo 1984

played out.)
White's queen is almost in a posi-
3. Beliavsky-Chernin, Tunis In- tion to attack the black queenside
tenonal 1985 pawns, but first he has to forestall
his opponent's kingside counter-
If 21 'ii'xc4 dxc4, Black would play. 37 ii'c6? is premature in view
stand quite well. of 37 ....a.xg3 38 lii.ixg3'iigl+, and
21 'ii'b4! (D) 37 ~fl? 'iib2 is similarly useless.
On 37 ~e2? Black has 37 .. .'ii'el!
(38 'iixe5+?? fails to 38 ....a.f6).
37 ~1!! (D)

White has in mind 2 l...~xc3 22

:Xc3 and then 22 ...'iixe2 23 l:c7
l:e7 24 1i'xd6 or 22 .. .'ii'xb4 23
ax.b4 J:lxe224 J:lc7J:lb825 J:lxb7, Threatening 38 'ifc6 with ad-
winning. He also has a plus after vantage. The game went37 ...'iib2
21...iixb4 22 axb4 ~f6 23 e3 l:e7 38 'ii'c6 'ifhl 39 iixc7 'ife4+ 40
24 ~2 g5 25 .a.cs !. ~h2 h5 41 'ii'c6 'iic2 42 gxh51i'f5
In the game followed 21 ...aS 22 43 'ifg2!? 'ifxh5 44 c5!, and White
'iixb5 lm:c3 23 'ifxc4 dxc4 24 bxc3 won.
J:lab8 25 .a.d7! l:e7 26 .a.a4 .a.ds
21 g4! g6 28 n rs 29 gxrsgxrs30 5. Pinter-Larsen.Las Palmas 1982
~ ~7 31 ~g3 'itr6 32 ~4 .a.r1
33 l:gl .i.g6 34 h4, and White suc- Obviously White should prepare
cessfully explohed his extra pawn. active operations on the kingside.
 Solutions to Chapter 4 Exercises 275

However, the hasty 25 ~f2? would the kingside and prepare to break
come up against the exchange sac- with g3-g4.
ritice 25 ...:Xg5 ! 26 fxg5 l2Jg6,mak- 25 ... Jle8 26 ~f3 l2Jd7 27 ffi
ing the position unclear. ~g6 28 :Z.hl 1;{7 29 .i.gS .aS 30
25 ~h4! (D) g4! 'iti>g8? (D) (30 ... fxg4 31 i.xg4
~e4 is more tenacious)

Putting paid to the threatened

sacrifice. If now 25 ...log6, then 26 31 ~f6 l:.r7 32 gxfS exfS 33
~f6; otherwise White will play ...ta JlhS! loxf6 34 exf6, and Black is
and Jlf3, concentrate his forces on defenceless.
Solutions to Chapter 11
Exercises
1. Forintos-Zedek, /mperia 1991 In the game, Black resigned af-
ter the further moves 18... rs 19
White needs to open the h-file for hxgS+ ~g8 20 g6 ..i.xe3+ 21 lbxe3
attack, but 17 h4? is answered by lli61-0.
17 ...g4!. The object of attack
must first be fixed. 2. Kholmov-Naumkin, Moscow
17 g4! ..i.xg4 Championship 1983
18 h4! (DJ
12 ... h6! (D)

B
w
Black is defenceless. If he plays
18...i.xf3, then 19 hxg5+ i.xhl 20 Black aims to play ...gS, not only
fi'h2+ (of course, 20 lhh 1+ also starting an offensive against the
leads to mate). king but also preparing to attack
Nor is 18...g6 any better: 19 the e5-pawn by means of ...lZ'ig6,
i.xg6 i.xf3 (19 ...fxg6 20 'ir'xg6) and ...i.g7, followed if necessary
20 hxg5+ '.Pg8 21 l:th8+ (or 21 by ...g4 or ...h5. White is powerless
'ii'h2) 21...@g7 22 l:th7+ ~g8 23 to stop this plan .
..i.xf7+ l:bt7 Z4 Wg6+. 13 h4 gS!
 Solutions to Chapter 11 Exercises 277

14 hxgS lLJg6 done by the king, which strides

Black has obtained a big advan- boldly forward.
tage. 23 ~b3! (D)
15 .i.g3 .i.e7! 16 .i.d3 l:tdg8 17
c4 dxc4 18 .i.xc4 hxgS 19 'ii'b3
lLif4 20 .:.Cdl :g6! 21 .i.xf4 gxf4
22 a4 l:r.hg8 23 ~fi q.Jb8! 24 aS
.i.c6 25 a6 :Xg2 26 be6 .i.h4 27
:d2 .i.xf3 28 .i.xg8 l:r.gl+ ! 0-1

3. A. Petrosian-Beliavsky, Riga
1973

In the Samisch Variation of the

King's Indian, as in certain other
openings, we sometimes witness
a strange s~enario. After castling
long, White conducts a queenside The advance of the white pawns
attack, advancing the pawns in cannot be stemmed. Beliavsky's
front of his own king- while Black attempts to complicate are unsuc-
attacks on the kingside. (The clas- cessful: 23 ...l:.c8 24 a4 c6 25 dxc6
sic instance of this strategy is :Xc6 26 bS dS 27 lLixdS .i.f8 28
Kotov-Szabo, Zurich Candidates Vxd3 'ii'd629 'it'b2!l:.b8 30 l:thcl,
tournament 1953.) At first sight and White won.
the actions of both players seem
paradoxical, but the explanation is 4. Liublinsky-Simagin, Moscow
simple. Each player attacks on 1939
the wing where he is stronger -
where he controls more space (as 12 .i.h8! (D)
determined by the central pawn A positional exchange sacrifice
structure) and has more pieces. - typical of this kind of structure -
In the position in front of you, which occurred several times in
White hl:!Sa clear plan- to advance Simagin's games. Black needs his
his pawns to a4 and b5, then pick bishop much more than his inac-
up the d3-pawn. However, the en- tive rook. The bishop will be use-
em)I:bishop must first be deprived ful both for defending its own king
of the c4-square. This can only be and for attacking the opponent's.
278 Solutions to Chapter 11 Exercises

for his pawn, as the actual game

continuation showed: 31 lt)d2?
~f7 32 'ii'd3 ~e6 33 lt)c4 'ilfd5! 34
~xb6+ (desperation; 34 'ii'fl tbd4
is quite hopeless for White)
34 ...axb6 35 'ifxd5 ~xd5 36 lt)xb6
~xf3 37 'itf2 .i.c6! 38 g4 l:.cf8 39
gxf5 l:.xf5+ 40 'ite3 l:.g2! 0-1.
White's best chance was to
sharpen the play by sacrificing a
piece:
31 lt)aS! (DJ

13 .hf'S 'ifxf8
14 a3?
Except when strictly neces-
sary, you should not make pawn
moves on the part of the board
where you are weaker. 14 ~4 is
better.
14 ... l:.b8
The rook co-operates splendidly
with the bishop; the two pieces ex-
ert tremendous pressure against
b2. White's position is probably B
hopeless already.
15 ~d3 c5! 16 exd5 ~xd5 17 For example, 31 ...bxaS 32 l:.b7+
lba4 i.d7! 18 •as ~xa4 19 'itd8 33 'ii'xaS+ ~e8 34 l:.xa7 with
'ifxa4 'ihi6+ 20 f 4 tbxf 4 21 l:.d2 a position difficult to assess.
~xb2+ 22 ~dl ~c3 23 .J:.f2 l:.bl#
(0-1) 6. Pcholkin-Tolonen, · Russian
Correspondence Championship
5. Van der Wiel-Larsen, Reykja- 1980-3
vik 1985
One of the most difficult ques-
In the 'normal' course of events tions in chess Is how to combine
White has no proper compensation attack and defence correctly,
 Solutions to Chapter 11 Exercises 279

avoiding both the Scylla of exces- shied away from the piece sacrifice
sive caution and the Charybdis and chose 24 .i.d3?. There followed:
of over-aggression which bor- 24 ...g6 25 l:tg3 ~b5 26 l:th3 'i!fa3 !?
ders on recklessness. (26 ... ~c3+!?) 27 Jlxb5 axb5 28
23 ... g6! (D) c3 bxc3 29 l:txd5 l:td8 30 l:.xd8
l:txd8 31 Jlcl 'ifa6 32 f4 l:tc8 33
'ifc2 exf4 34 l:txc3, with a roughly
equal game.
The variations after 24 g6 ! ! (D)
are as follows:

w
By defending against White'.s
threat g5-g6, Black maintains an
excellent position. White will have
to give very serious thought to the
threats of 24 ...dS and 24 ...~b5.
On the other hand if Black is im- 24 ... dxc4
patient and tries to carry out one of If 24 ... fxg6 25 fxg6 hxg6, then _
these threats at once, he allows 26 :Xd5 ! ~xd5 27 l:tdI l:tb5 28
White to launch a dangerous attack l:txd5 l[xd5 29 .i.b6 and White
against the king, which outweighs wins.
Black's queenside operations. For 25 1Wh4 fxg6
instance, 23 ... ~b5? 24 g6! ~a3+ 26 fxg6 h6
(24 ...~3+ 25 ~cl) 25 ~al ~xc4 Nor is 26 ...hxg6 any better: 27
26 1i'h4, threatening to give mate 'ifxc4+ ~h8 28 l:tg3.
or take on c4 with the queen. 27 .i.xh6! . gxh6
Instead Black played 23 ... dS?, 28 g7
whe~ White should have contin- and White's attack is irresisti-
ued 24 g6!!. In the game, White ble.
280 Solutions to Chapter 11 Exercises

7. Simagin-Petrosian, Moscow ltbe8 22 aS? eS 23 .i.gS? f6 24 .,tc12

1956 (threatening 25 .,lg6) 24 ...exd4 25
cxd4 llxel + 26 llxel c5! 27 a6!
17 h4! (D) cxd4 28 .,la5?!
A tempting move, but not the
best. 28 .i.e4! was stronger, for ex-
ample 28 ...bxa6 29 'ii'd3 'ii'b6 30
Jlg2! with a decisive attack.
28... b6 29 Jld2 ~e5 30 'itg2 d3
31.,la4
We have now reached the posi-
tion in the next exercise.

8. Simagin-Petrosian, Moscow
1956

B All White has to do is play 32 J:[cl,

and Black will be in a bad way in
A good example of astute pro- spite of his extra pawn. Let us see
phylaxis with the kings on oppo- how the game concluded:
site sides! •'This move looks risky, 31 ... .,lcS? 32 :et Vf7 33 b4
but is just the way to paralyse .i.d4 34 Jlb3 WJ/e735 M4 bS 36
Black's kingside attack" (Simagin). 'if c6 'ifd7 37 'ife4 Jlb6 38 J:[dl
The important thing is to deprive l:t.e839 Ax:d3 'ifc8 40 .i.f7 'IJ..e7
41
the opponent of ...h4 or ... gS. To Jlxh5 1i'c4 42 'fnt7! (42 Jlf3!?)
this end, it is even permissible to 42 ...<it>c743 J:[d2 'ii'xb4 44 'ifg8
violate the principle stated in our l:t.d745 J:t.c2+.i.c5 46 'if a8 q.>d6
comments on another Simagin 47 J:.d2+ .i.d4 48 ~e3 cit>e649
game (see exercise 4). Possessing 'ii'e8+ ~f5 50 g4+ c.te4 51 'iVa8+
two powerful bishops, White has l:tdS 52 Jhd4+ 1-0
the better chances. He threatens With 31 ...J:t.c8,Black could have
c4-c5. Instead, the immediate 17 stopped the enemy rook from oc-
c4 gS! 18 c5 1Le7 19 Jlxf4 gxf4 cupying the c-file (32 J:t.c 1? 'fi'xc 1+
(19 ...'ir'xf4) 20 loe2 is only suffi- 33 .i.xc 1 lhc 1+ 34 Wh2 ~g4+ 35
cient for equality. <if.ih3l:thl+! 36 'ii'xhl loxf2+), but
17 .. .lt:WdS 18 loe4 loxe4 19 after the reply 32 iLb5 !? his posi-
.i.xe4 ~f6 20 .i.c2 ~g4 21 g3 tion would still have been difficult
 Solutions to Chapter 11 Exercises 281

due to the exposed position of his 32 JlxbS

king and his lack of any real coun- If 32 .i.a5, then 32 ... 'ili'c6! 33
terplay. 'ii'xc6 (33 .i.xd8 can be met by
Only a player who sensed the 33 ... bxa4 or 33 ... lZJf3+ 34 c;t.,hl
full strategic danger ofBlack's po- bxa4) 33 ... ltlxc6 34 .i.xd8 bxa4!
sition could resolve to complicate 35 l:e6 (35 l:e8 lbxd8 36 l:.xd8+
with 31. ..b5. After all, he would ~c7) 35 ...Jle5 with advantage to
have to reckon with the seemingly Black.
deadly reply .i.a5. In actual fact, 32 ... 'ii'b6
that move is dubious; while win- Black needs to eliminate the
ning the exchange, White removes white a-pawn. Again the bishop
the blockade from the strong sortie to a5 is not dangerous.
passed pawn on d3. 33 ~aS?! 'ii'xbS
31 ... bS!! (DJ 34 .i.xd8 d2!
But not 34 ...'it'xa6? 35 'ifd.5, nor
34 ...ll)c6? 35 l:e8.
35 l:dl 'i!fb3
36 ft7+
The exchange of queens is
forced. Neither 36 l:.xd2? ltif3+
nor 36 'iffl? .i.c5 is any good.
36 ... 'iixb7
37 axb7 .i.b4
What is White to do now? The
threat is not only 38 ... lbc4,· but
also 38 ...ll)c6 trapping the bishop.
A good example of an opportune
counter-attack.
Index of Players
Numbers refer to pages.
When a name appears in bold, the first-named player had Black.
Otherwise the first-named player had White.

ADAMS-Yusupov, 171 CIFUENTES- Zviagintsev, 264

AHUES- Yates, 103 CIOCALTEA -Liberzon, 23
ALF.XANDER - Euwe, 12; DENISOV - Chemosvitov, 238
Marshall, 39 DOLMATOV -Flesch, 82;
ANAND- Kasparov, 233, 234, Franzoni, 79; Larsen, 86;
236; Yusupov, 162 Lerner, 76, 94;
BAGIROV - Dvoretsky, 59; Van der Sterren, 89
Suetin, 55 DV0RETSKY - Bagirov, 59;
BAREEV - Kasparov, 232 Bronstein, 187; Butnorius, 16;
BELAVENETS-Jochelson, 68 & Chekhov, 28; Georgadze, T.,
273 182; .Khramtsov, 192; Kotkov,
BELIAVSKY - Chemin, 68 & 274; 24; Mariasin, 189; Peev, 129;
Petrosian, A., 197 & 277, Tai, 63
Razuvaev,135 ENGEL - Sanakoev, 130;
BELOV - Kamshonkov, 201; Sanakoev, 145
Lempert, 204; Mitiaev, 210; EUWE - Alexander, 12;
Vaulin,199;Yachmennik,206 Nimzowltsch, 105
BERG - Hort. 37 & 269 FAHRNI - Rotlewi, 108
BLUMENFELD - Orlov, 106; FLEsCH - Dolmatov, 82
Sergeev,48;Zhivtsov,46 FORINTOS-2.edek, 197 & 276
B0BROV - Platonov, 37 & 270 FRANZONI - Dolmatov, 79
BOGUSLAVSKY-Lepin, 251; GEORGADZE,T. -Dvoretslcy, 182
Morozevich, 255 GRIGORIEV- Sergeev, 104 •
BRONSTEIN- Dvoretsky, 187 GULK0 - Yusupov, 166
BUTN0RIUS -Dvoretsky, 16 GURGENIDZE- Smyslov, 68 &
CAPABLANCA - Nimzowitsch, 64 273
CHEKHOV - Dvoretsky, 28 HORT- Berg, 37 & 269
CHERNIN - Beliavsky, 68 & 274 HOHNER- Yusupov, 159
CHERNOSVITOV- Denisov, 238 lVANCHUK- Yusupov, 156
 Index of Players 283

IVK0V - Petrosian, T., 118 MILES - Makarychev, 68 & 274

JANOWSKI- Lasker, Em., 106 MITIAEV - Belov, 2/0
JANSA - Sokolov, A., 37 & 270 M0R0ZEVICH - Boguslavsky,
JOCHELS0N - Belavenets, 68 & 255
273 NACHEV - Zviagintsev, 258
KAMSH0NK0V - Belov, 201 NAJD0RF- Kotov, 14
KARPOV - Kasparov, 62 & 71 NAUMKIN- Kholmov, 197 & 276
KASPAR0V -Anand, 233, 234; NIKITIN - Makariev, 261
236;Bareev,232;Karpov,62 NIMZ0WITSCH - Capablanca, 64;
& 71 Euwe,105;Kmoch,100
KH0LM0V - Naumkin, 197 & 276 ORL0V - Blumenfeld, 106
KHRAMTS0V - Dvoretsky, 192 PARUTIN- Teplitsky, 253
KM OCH - Nimzowitsch, 100; PCH0LKIN-Tolonen, 198 & 278
Yates, 102 PEEV - Dvoretsky, 129
KOTK0V - Dvoretsky, 24 PETR0SIAN, A. - Beliavsky, 197
Korov -Najdorf, 14 &277
KRASENK0V - Lagunov, 42 PETR0SIAN,T.-lvkov,118;Sax,
LAGUN0V - Krasenkov, 42 125; Simagin, 198 & 280
LARSEN - Dolmatov, 86; Pinter, PINTER-Larsen, 69 & 274
69 & 274; Van der Wiel, 198 PLAT0N0V - Bobrov, 37 & 270
&278 RAZUVAEV- Beliavsky, 135
LASKER, EM. - Janowski, 106 REBEL 8 - Yusupov, 151
LEMPERT - Belov, 204 RIBLI - Tai, 115
LEPIN - Boguslavsky, 251 RIVAS- Mestel, 227
LERNER - Dolmatov, 76, 94; ROfLEWI -Fahrni, 108
Lukin, 22 SANAK0EV - Engel, 130; Engel,
LIBERZ0N - Ciocaltea, 23 145; Ljungdahl, 139; Maeder,
LIUBLINSKY- Simagin, 197 & 216; Sevefok, 136; Zaitsev, A.,
277 131
LJUNGDAHL- Sanakoev, 139 SAX - Petrosian, T., 125
LUKIN - Lerner, 22 SERGEEV- Blumenfeld, 48;
·MAEDER - Sanakoev, 216 Grigoriev, 1()4.
MAKARIEV - Nikitin, 261 SEVEtEK - Sanakoev, 136
MAKARYCHEV- Miles, 68 & 274 SHAMK0VICH- Simagin, 176
MARIASIN - Dvoretsky. 189 SIMAGIN- Liublinsky, 197 &
MARSHALL -Alexander, 39 277; Petrosian, T., 198 & 280;
MEsTEL - Rivas, 227 Shamkovich, 176
284 Index of Players

SMIRIN- Vogt, 147 V ASIUKOV- Tal, 110

SMYSLOV- Gurgenidze, 68 & 273 V AULIN- Belov, 199
SOKOLOV,A. -Jansa, 37 & 270 VOGT- Smirin, 147
SUETIN- Bagirov, 55 VULFSON- Tsariov, 121
TAL-Dvoretsky, 63; Ribli, 115; XIE JUN - Yusupov, 153
Vasiukov, II 0 Y ACHMENNIK- Belov, 206
TEPLITSKY- Parutin, 253 YATES-Ahues, 103; Kmoch,
TIMOSHCHENKO- Vaganian, 37 102
&271 YUSUPOV- Adams, 171; Anand,
TOLONEN- Pcholkin, 198 & 278 162; Gulko, 166; Hilbner, 159;
TSARIOV- Vulfson, 121 Ivanchuk, 156; Rebel 8, 151;
VAGANIAN- Timoshchenko, 37 Xie Jun, 153
&271 ZAITSEV,A. - Sanakoev, 131
VAN DER STERREN- Dolmatov, ZEDEK-Forintos, 197 & 276
89 ZHIVTSOV- Blumenfeld, 46
VAN DER WIEL - Larsen, 198 & ZVIAGINTSEV- Cifuentes, 264;
278 Nachev, 258

Index of Composers
Bondarenko, F. and Kuznetsov, A. 20
Bondarenko, F. and Liburkin, M. 35
Kasparian, G. 32
Wotawa, A. 19
Index of Games
Players Event Page

Alexander - Euwe Nottingham 1936 12

Alexander - Marshall Cambridge 1928 39
Anand - Kasparov World Ch (9), New York 1995 233
Anand - Kasparov World Ch ( 11 ), New York 1995 234
Anand - Kasparov World Ch ( 17), New York 1995 236
Bareev - Kasparov Linares 1992 232
Beliavsky - Chernin Tunis lnterzonal 1985 68 &274
Belov - Vaulin Katowice 1991 199
Berg-Hort Biel 1985 37 & 269
Blumenfeld - Zhivtsov Moscow Ch, Semi-final 46
Bobrov - Platonov Chemovtsy 1963 37 & 270
Boguslavsky - Lepin Moscow 1989 251
Boguslavsky - Morozevich Moscow Junior Ch 1990 255
Bronstein - Dvoretsky USSR Ch (1st League), Odessa 1974 187
Cifuentes - Zviagintsev Wijk aan Zee 1995 264
Ciocaltea - Liberzon Netanya 1983 23
Denisov - Chernosvitov Moscow Junior Ch 1991 238
Dohnatov - Flesch Bucharest 1981 82
Dohnatov - Franzoni World Junior Ch, Graz 1978 79
Dohnatov - Larsen Amsterdam 1980 86
Dolmatov - Lerner Daugavpils 1978 76
Dolmatov - Lerner Tashkent 1983 94
Dvoretsky - Bagirov USSR Ch (First league), Tbilisi 1973 59
Dvoretsky - Butnorius Dubna 1970 16
Dvoretsky - Chekhov Moscow 1974 28
Dvoretsky - Khramtsov Moscow 1970 192
Dvoretsky - Peev European Champions' Cup,
Plovdiv 1975 129
Engel - Sanakoev Romanian Jubilee Corr. 1976-9 130
Forintos - Zedek lmperia 1991 197 & 276
Georgadze, T. - Dvoretsky USSR Spartakiad, Moscow 1967 182
286 Ind.ex of Games

Jansa - Sokolov, A. Biel /nterzonal 1985 37 &270

Jochelson - Belavenets correspond.ence 1974-9 68 & 273
Kamshonkov - Belov Podolsk 199 I 201
Kasparov - Karpov World Ch (6). Moscow 1984/5 62
Kasparov - Karpov World Ch (6), Moscow 198415 71
Kholmov - Naumkin Moscow Ch 1983 197 & 276
Kmoch - Nimzowitsch Niendoif 1927 100
Km~h-Yates San Remo 1930 102
Kotkov - Dvoretsky Moscow Ch 1972 24
Lagunov - Krasenkov Dnepropetrovsk 1985 42
Lasker, Em. - Janowski World Ch (7). Paris 1909 106
Lempert - Belov Katowice 1990 204
Lerner - Lukin USSR 1977 22
Liublinsky - Simagin Moscow 1939 197 & 277
Mariasin - Dvoretsky Kiev 1970 189
Miles - Makarychev Oslo 1984 68 &274
Mitiaev - Belov Moscow 1989 210
Najdorf - Kotov Mar del Plata 1957 14
Nikitin - Makariev CIS Junior Ch, Jurmala 1992 261
Nimzowitsch - Capablanca New York 1927 64
Nimzowitsch - Euwe Karlsbad 1929 105
Orlov - Blumenfeld Semi-final, Moscow Ch 1932 106
Pcholkin - Tolonen Russian Corr. Ch 1980-3 198 & 278
Petrosian, T. -Ivkov Nice Olympiad 1974 118
Petrosian, T. - Sax Tallinn 1979 125
Petrosian, A. - Beliavsky Riga 1973 197& 277
Pinter - Larsen Las Palmas 1982 69 &274
Razuvaev - Beliavsky USSR Ch. Minsk 1979 135
Ribll-Tal Cand.idates, Montpellier 1985 115
Rivas - Mestel Marbella Zonal 1982 227
Rotlewi - Fahrni Karlsbad 1911 108
Sanakoev - Engel 10th World Corr. Ch 1978-84 145
Sanakoev - Ljungdahl 6th World Corr. Ch 1968-71 139
Sanakoev - Maeder I 0th World Corr. Ch 1979-84 216
Sanakoe-v - Sevecek 6th World Corr. Ch 1968-70 136
Sergeev - Blumenfeld Moscow Ch. Semi-final 48
Sergeev - Grigoriev Masters' Tournament, Moscow 1932 104
 Index of Games 287

Shamkovich - Simagin Leningrad 1951 176

Simagin - Petrosian, T. Moscow 1956 198 & 280
Smirin - Vogt Saltsjobaden 198819 147
Smyslov - Gurgenidze USSR Ch. Tbilisi 1966 68 & 273
Suetin - Bagirov USSR Ch. Leningrad 1963 55
Tai - Dvoretsky USSR Ch. Leningrad 1974 63
Teplitsky - Parutin Tashkent 1989 253
Timoshcbenko - Vaganian USSR Ch, First League,
Baku 1977 37 &271
Tsariov - Vulfson Moscow 1989 121
Van der Sterren - Dolmatov Amsterdam 1979 89
Van der Wiel - Larsen Reykjavik 1985 198 & 278
Vasiukov - Tai USSR Ch, Baku 1961 110
Yachmennik - Belov Smolensk 1989 206
Yates - Ahues Hamburg 1930 103
Yusupov - Adams Dortmund 1994 171
Yusupov"- Anand Linares 1991 162
Yusupov - Gulko Novgorod 1995 166
Yusupov - Hilbner 1ilburg 1987 159
Yusupov - Ivanchuk Tal Memorial, Riga 1995 156
Yusupov - Rebel 8 Action Chess match ( 13),
Ischia 1997 151
Yusupov - Xie Jun Linares 1997 153
Zaitsev, A. - Sanakoev 6th USSR Corr. Ch 1963~5 131
Zviagintsev - Nachev Voskresensk 1990 258
 Index of Openings

Numbers refer to pages.

Alekhine Defence 59 Pirc Defence

Caro-Kann Defence .i.c4 189
Panov Attack 82, 86 Austrian Attack 171
French Defence Classical 125
Classical Variation 182, 187 Queen's Gambit
Griinfeld Defence Accepted 156, 238
Russian System 176 Semi-Slav Defence 264
Schlechter Variation 253 Slav Defence 159, 258
King's Indian Attack 110 Queen's Indian Defence 71
King's Indian Defence Queen's Pawn Game 151,162
Classical, 7 ... 'ii'e8 255 Reti Opening 28, 89, 115
Samisch, 6 ...~c6 261 RuyLopez
Modem Benoni 251 Exchange Variation 153
Nimzo-lndian Defence Sicilian Defence
Rubinstein Variation 118, 166 f4 J21
Nimzowitsch/Larsen/Simagin Najdorf, 6 .i.g5 216
Attack 192 Richter-Rauzer Attack 94
Philidor's Defence 76 Taimanov Variation 79