Junior Tay

The Schliemann Defence

move by move

www.everymanchess.com
 About the Author

Junior Tay is a FIDE Candidate Master and an ICCF Senior International Master. He is a for-
mer National Rapid and Cairnhill Open Champion and has represented Singapore in inter-
national events, including the 1995 Asian Team Championships. He is a frequent opening
surveys contributor to New in Chess Yearbook and also writes articles for CHESS magazine.
He has been a chess trainer, author and editor for the past three years, after working as a
school teacher for seventeen years.

Also by the Author:

The Benko Gambit: Move by Move
lvanchuk: Move by Move
The Old Indian: Move by Move
 Contents

About the Author 3

Bibliography 5
Introduction: What is the Schliemann Defence 7

1 Trivial Tries: 4 0-0, 4 Íxc6 and 4 exf5 28

2 Central Counter: 4 d4!? 58
3 Delightful or Dullsville: 4 d3 98
4 Cloudy Classical: 4 Ìc3 fxe4 5 Ìxe4 d5 246
5 Tactical Tartakower: 4 Ìc3 fxe4 5 Ìxe4 Ìf6 283
6 Odds and Ends 367
7 Schliemann Stumpers! 374

Solutions 385
Index of Variations 395
Index of Games 399
 Introduction
What is the
Schliemann Defence?

The Schliemann is an opening variation of the Ruy Lopez which occurs after 1 e4 e5 2
Ìf3 Ìc6 3 Íb5 f5!?. Here Black immediately challenges the central white e-pawn with a
flank thrust, seeking to demolish it instantly. By eschewing development in favour of such
blatant aggression, it is not surprising that Black frequently finds himself the e5-pawn
down, especially when he has to complete development himself in the process. The open-
ing of the f-file and, often, the attainment of the bishop pair (after White plays Íxc6) does,
however, afford him chances to attack and complicate.
Black’s king is usually the more exposed one, though the tactical nature of the position
provides him with a truckload of tactics to get by, especially at lower levels. I tend to see the
Schliemann Defence as akin to the Vienna Game (1 e4 e5 2 Ìc3 Ìf6 3 f4), where White
tries to scuttle the black centre on move three.

Why play the Schliemann?

b If you are aiming for a street brawl when handling Black and seek to unbalance the
game as early as possible, the Schliemann is right up your alley. Black ups the ante
on move three with Freddie the f-pawn and plays to wrestle the centre from White,
as well as to open the f-file. On the other hand, if you prefer to aim for solid equal-
ity as Black in the opening, the Schliemann is definitely not for you.

b It’s not so easily to deal with Black’s initiative in practical play. In this computer
age, engines might be able to show the way to a White plus versus the Schliemann.
Fortunately, most mortals do not have the prodigious memory required to recall
what the engine proposes to reach that plus, and even when they do it is another
issue to win from there. On many occasions, I go wrong in the Schliemann and give

7
The Schliemann Defence: Move by Move

my opponent a big edge, but the complex nature of the positions still gives me
chances to turn the tables. It is interesting to note that, in MegaBase, the Schlie-
mann has the second best percentage for Black among the defences to the Ruy Lo-
pez, with White scoring 52.8% (up to 2016). If you’re interested, the Smyslov Sys-
tem (3...g6) is top of the class with 52.7%.

b Top players have given the Schliemann a go, even against world-class opponents.
For example, Andreikin, Aronian, Carlsen, Ivanchuk and Radjabov have punted the
Schliemann on several occasions. Special mention must be made of Radjabov, who
has used it successfully to reach a plus score versus opposition ranging from Elo
2560 to 2800. His Schliemann even held 2700+ opposition to an even score, as well
as securing three draws against the current World Champion, Magnus Carlsen.
Other notable Schliemann users are GMs Ivan Sokolov and Roeland Pruijssers and
the late IM/CC-GM Josef Boey (all Dutch), as well as the Bulgarian GM Ventzislav
Inkiov.

b I guess personal experience is the best way to relate the viability of the system.
Please bear with my explanation of how this line became my favourite Black open-
ing.

About four years after I learned chess (1986), I often joined my chesspal Chia Keng San
in the home of Wang Mong Lin (a CC Senior IM and former British Universities Champion)
for overnight blitz games. Chia was very annoyed and frustrated at Wang’s frequent usage
of the Ruy Lopez Exchange Variation (3 Íb5 a6 4 Íxc6), spoiling his Marshall Attack prepa-
ration. One day, he came across a New in Chess magazine game where British GM Jonathan
Speelman, facing imminent defeat and elimination from the World Championship Candi-
dates Quarter-finals, deployed the Schliemann against Jan Timman.

Example 1
J.Timman-J.Speelman
Candidates (7th matchgame), London 1989

1 e4
Timman had previously only used 1 d4 in the match. But if his idea was to surprise his
opponent, the tables were swiftly switched when Speelman chose something quite un-
common (for him) on move one.
1...e5 2 Ìf3 Ìc6 3 Íb5 f5!?

8
 Introduction: What is the Schliemann Defence

W________W
[rDb1kgn4]
[0p0pDW0p]
[WDnDWDWD]
[DBDW0pDW]
[WDWDPDWD]
[DWDWDNDW]
[P)P)W)P)]
[$NGQIWDR]
W--------W
Speelman seldom plays 1...e5, let alone the Schliemann Defence. In the previous two
years, he had tried 3...Íc5, 3...Ìf6, and the Closed Ruy with 3...a6 4 Ía4 Ìf6 5 0-0 Íe7 6
Îe1 b5 7 Íb3 d6. But dire situations – one point down with two games left – require dras-
tic measures. I will gloss over the next twelve moves as the theory will be explained later in
the book.
4 Ìc3 fxe4 5 Ìxe4 d5 6 Ìxe5 dxe4 7 Ìxc6 Ëg5 8 Ëe2 Ìf6 9 f4 Ëxf4 10 Ìe5+ c6 11 d4
Ëh4+ 12 g3 Ëh3 13 Íc4 Íe6 14 Íg5 0-0-0 15 0-0-0 Íd6 16 Ìf7
W________W
[WDk4WDW4]
[0pDWDN0p]
[WDpgbhWD]
[DWDWDWGW]
[WDB)pDWD]
[DWDWDW)q]
[P)PDQDW)]
[DWIRDWDR]
W--------W
White wins the bishop pair after this ‘fork’. Curiously, both Stockfish and Houdini prefer
this move too. The result of the exchange is that Black’s slight congestion has been allevi-
ated, while White’s remaining light-squared bishop doesn’t seem to have much influence.
16...Íxf7 17 Íxf7 Îhf8 18 Íc4 Îde8
Structurally, this looks like a Caro-Kann or 2...Ëxd5 Scandinavian, where Black has got-
ten the e-pawn moving after trading off f-pawns. Black’s activity accords him good

9
The Schliemann Defence: Move by Move

chances, even if White still has a small plus.

19 d5!?
Many annotators have deemed this to be an error; in actuality it is by no means bad,
though it does represent the turning point of the game. Timman had used a lot of time
prior to making this move and his logic was probably to prevent Black from getting ...Ìd5
in with a comfortable position. 19 Îhf1 is the main move, as we will see in Chapter Four.
19...c5!
This reduces the scope of White’s light-squared bishop. The d5-pawn is firmly blockaded
and, if White trades on d6, Black’s knight will become a stronger piece than White’s re-
maining bishop.
20 Îhf1 Êb8 21 Íf4!?
Speelman attached a dubious mark to this move in ChessBase Magazine. Again it is not
objectively that bad if combined with the idea of removing Black’s blockader of the passed
d-pawn.
21...Îd8
W________W
[WiW4W4WD]
[0pDWDW0p]
[WDWgWhWD]
[DW0PDWDW]
[WDBDpGWD]
[DWDWDW)q]
[P)PDQDW)]
[DWIRDRDW]
W--------W
Speelman keeps the pressure on by targeting the d-pawn indirectly.
22 Íg5?!
Perhaps a prelude to repeating moves, supposing Black replied 22...Îde8. Instead, the
‘Speelwolf’ takes the opportunity to increase the tension by adding a ...b7-b5 outflanking
possibility.
Stronger was 22 Íxd6+! Îxd6 23 Íb3 Îdd8 24 c4 Ëd7 25 Êb1 h6 26 Îde1 Îfe8 27 Ëe3
Ëd6 28 Ía4 Îe5 and it is not easy for either player to improve their position. Black cannot
get ...b7-b5 in easily without compromising his king’s safety, and White cannot target the
e4-pawn (with Íc2) without relinquishing his hold on the b5-square. Computers rate this
position as slightly better for White, perhaps on the basis of a spatial edge and protected
passed d-pawn, though for humans it is tough to play either side!
22...a6!

10
 Introduction: What is the Schliemann Defence

All of a sudden, Black threatens to advance on the queenside and lock in the light-
squared bishop. Timman responded less than ideally to the pressure exerted by the wily
Englishman.
23 Íxf6?!
This is where Timman, in seeking trades, starts losing thread of the game. As the posi-
tion opens up, the difference in the strength of the bishops become more apparent.
It was more prudent to restrain Black on the queenside. Hence 23 a4 is better, when the
game might continue 23...Ëd7 24 b3 Îde8 25 Êb1 h6 26 Íe3 with equal chances, even if
White’s light-squared bishop is looking more and more like a duck.
23...gxf6
Keeping more tension by not allowing a rook trade after 23...Îxf6.
24 Ëxe4
Technically speaking, there is nothing wrong with this move either, but it requires
White to play with great accuracy as the black pieces now become very active. In particular,
Black’s bishop increases in power once the h2-pawn is munched, and the difference be-
tween the two bishops’ strengths will soon be apparent.
24...Ëxh2
W________W
[WiW4W4WD]
[DpDWDWDp]
[pDWgW0WD]
[DW0PDWDW]
[WDBDQDWD]
[DWDWDW)W]
[P)PDWDW1]
[DWIRDRDW]
W--------W
25 Îh1?!
Losing a pawn, since White cannot afford to capture on h7. Instead, the ever-creative
Speelman recommended 25 Îf3!, with the idea of Íxa6, and this can very well occur after
25...Îfe8 (or 25...Íxg3 26 Íxa6) 26 Ëd3 Îg8 27 Íxa6 bxa6 28 Ëxa6 Îxg3 29 Îxf6 Îg1 30
Ëb6+ Êc8 31 Ëc6+ Êb8 32 Ëb6+ with a draw by perpetual check.
25...Ëxg3
Not just snagging a pawn, this also frees the f4-square for the black bishop.
26 Îxh7??
This is the move that actually loses the game for White, whose back rank is left exposed
after Black’s reply. White has no choice but to grovel a pawn down with 26 Îdf1; for exam-

11
The Schliemann Defence: Move by Move

ple, 26...Îh8 27 Îh5 Îde8 28 Ëf3 Ëxf3 29 Îxf3 Íe5 30 Íf1 h6 31 c3 and so on.
26...Îfe8
W________W
[WiW4rDWD]
[DpDWDWDR]
[pDWgW0WD]
[DW0PDWDW]
[WDBDQDWD]
[DWDWDW1W]
[P)PDWDWD]
[DWIRDWDW]
W--------W
Now White is well and truly lost: ...Îe1 is coming, followed by ...Íf4+, nudging the king
away from the defence of the d1-rook, while Black threatens to embarrass the white bishop
with ...b7-b5 and ...c5-c4 as well.
27 Ëf5
Nothing else is any better: 27 Ëd3 Îe3 28 Ëf1 Îde8 is curtains for White, and 27 Ëh4
Ëf3! wins at least a piece; e.g. Îf1 Ëe3+ 29 Êd1 (or 29 Êb1 Îe4 and the bishop drops)
29...b5 30 Íd3 c4 and the bishop has nowhere to go without allowing ...Ëe2+, snagging
the f1-rook.
27...b5
There goes the bishop. 27...Îe1 28 Îh1 Îde8 29 Îf1 b5 also wins.
28 Íf1 Îe1 29 Ëh5 Ëf4+ 30 Êb1 Ëxf1 0-1
Timman resigned, thus allowing Speelman to draw level at 3½-3½ (though Timman re-
covered to win the next game and the match).

Both Chia and I were instantly hooked on this arcane way of dealing with the Ruy Lopez,
and he did get quite a few blitz points off Wang with it. As for me, this was what occurred in
my first tournament game with the Schliemann against an experienced National Master.

Example 2
L.O.Choong-J.Tay
Cairnhill Open, Singapore 1991

1 e4 e5 2 Ìf3 Ìc6 3 Íb5 f5 4 Ìc3 fxe4 5 Ìxe4 d5

Tartakower’s 5...Ìf6 is the other main line.

12
 Chapter Five
Tactical Tartakower:
4 Ìc3 fxe4 5 Ìxe4 Ìf6

Rather than push his d-pawn at once, as in the previous chapter, by playing 5...Ìf6
Black concentrates on rapid development. White has three ways to respond.
With 6 Ìxf6 Ëxf6 7 0-0, White plays for swift development too. Then Games 34-36 see
Black replying with 7...Ìd4, after which White can decide between a quick draw (as shown
in the notes to Game 34) or playing for a blitzkrieg assault on the black king following a
queenside fianchetto (as in Games 35 and 36). In the latter case Black’s resources seem
adequate provided he plays the counter-intuitive ...Êd8 when White checks on the e-file.
GM Sokolov prefers 7...Íe7, which is examined in Game 37. I am rather doubtful of
Black’s ability to generate sufficient play if White just sits on the isolated e-pawn after 8
Íxc6 dxc6 9 Ëe1, so I have included 8...bxc6!? as a viable alternative. This follows the cur-
rent preference for the b-pawn capture in the 7 Ëe2 Íe7 8 Íxc6 variation, where White
plays simply to win the e-pawn. Here Black often has to navigate a queenless middlegame
a pawn down, but the extra central pawn and possibility of developing the light-squared
bishop to a6 after 8...bxc6 (Games 39-40) offer more counterplay than after 8...dxc6 (Game
38), where Black’s main aim is to hold a fortress with opposite-coloured bishops.
White’s third option is to play 6 Ëe2 first, waiting for 6...d5 before exchanging with 7
Ìxf6+ gxf6, after which he can damage Black’s pawn structure with 8 d4 Íg7 9 dxe5 0-0
10 e6 or 10 Íxc6 bxc6 11 e6. This is given a work over in Games 43 and 44. White has some
unorthodox and trappy sidelines too: 8 Ìd4!? is discussed in Game 42, while 7 Ìxe5!? and
7 Ìeg5!? are scrutinized in Game 41.

Game 34
Li Ruifeng-J.Tay
World Mind Games, Internet (rapid) 2014

283
The Schliemann Defence: Move by Move

1 e4 e5 2 Ìf3 Ìc6 3 Íb5 f5 4 Ìc3 fxe4 5 Ìxe4 Ìf6 6 Ìxf6+ Ëxf6

The advantage for White of taking on f6 at once is that, unlike in the 6 Ëe2 variation,
Black is virtually obliged to recapture with the queen, since gambling with 6...gxf6? is too
risky. After 7 d4! Íg7 (if 7...e4 8 Ìg5! Íb4+ 9 c3 0-0 10 Ìxe4, White is already winning,
Y.Quesada Perez-S.Palit, Barcelona 2012; or similarly 7...d6 8 Ìg5! fxg5 9 Ëh5+ Êd7 10 d5
Ëf6 11 Íxg5 Ëf6 12 dxc6+ bxc6 13 Íd3 e4 14 Íe2, N.Ristic-D.Bokan, Serbian Team Cham-
pionship 2002) 8 dxe5 0-0 9 exf6 Ëxf6 10 0-0 d6 11 c3 Íg4 12 Íe2 Îae8 13 Ëb3+ Êh8 14
Íe3, Black has no compensation for his pawn minus, H.Faber-P.Leisebein, correspondence
2013.
7 0-0
This has been superseded by 7 Ëe2 in the past two decades. However, it remains a low-
risk line for White to play, especially if he doesn’t mind a draw.
7...Ìd4
W________W
[rDbDkgW4]
[0p0pDW0p]
[WDWDW1WD]
[DBDW0WDW]
[WDWhWDWD]
[DWDWDNDW]
[P)P)W)P)]
[$WGQDRIW]
W--------W
This seems like a very counter-intuitive move to make, doesn’t it? Instead of focusing on
development, Black is asking White to trade off his only developed minor piece and open
the e-file as well.

Question: Why would Black allow all that?

Answer: It does seem like a bad variation of the Spanish Bird’s (3...Ìd4), but there are some
counter-arguments. Firstly, Black resolves the problem of the typically weak e5-pawn by
diverting it to the d-file after the exchange on d4. Secondly, if the white rook goes to e1,
Black will have queen and rook bearing down on the f2-pawn after he castles. Lastly, White
will not find it so easy to mobilize the rest of his queenside pieces.
The alternative 7...Íe7 is covered in Game 36.
8 Ìxd4 exd4 9 Îe1+

284
 Tactical Tartakower: 4 Ìc3 fxe4 5 Ìxe4 Ìf6

White can accelerate his development by playing 9 b3 and bringing the c1-bishop out to
a3 or b2. This dangerous plan is covered in the next two games.
9...Íe7
Here 9...Êd8!? is an interesting way to avoid the drawing variation below. Surprisingly,
it seems only to have been tried once, and Black almost got into trouble: 10 d3 c6 11 Ía4
d5 12 Ëh5 Êc7? (12...h6 is mandatory with equal chances) 13 Íg5! Ëf5
W________W
[rDbDWgW4]
[0piWDW0p]
[WDpDWDWD]
[DWDpDqGQ]
[BDW0WDWD]
[DWDPDWDW]
[P)PDW)P)]
[$WDW$WIW]
W--------W
Exercise: How can White exploit the awkward placement of the black king
and queen?

Answer: White has the powerful 14 Îe5! (it looks very tempting to play 14 Íd8+ Êxd8 15
Îe8+ Êc7 16 Ëxf5 Íxf5 17 Îxa8, but things are not so simple after 17...Íc8, cutting off
the rook) 14...Ëg6 (or 14...Ëxh5 15 Îxd5+!) 15 Íf4 Íd6 16 Îe7+ Íd7 17 Íxd6+ Ëxd6 18
Îxg7 with a big advantage.
Instead, G.Hitter-M.Lyell, Budapest 2010, continued 14 Îe8 b5 15 Íb3 Êb7, and even
here 16 Ëh4! would have kept Black on the defensive.
10 Ëe2
Alternatively, 10 Ëh5+ g6 11 Ëh6 (if 11 Ëe5 Ëxe5 12 Îxe5 c6 13 Íd3 Êf7 14 b3 Íf6 15
Îe2 d5 16 Íb2 Íd7 17 Îae1, as in G.Laketic-V.Tseshkovsky, Moscow 1992, then 17...c5
gives Black a space advantage and potential queenside expansion ideas) 11...c6 12 Íf1 (12
Ía4 Êf7 13 Íb3+ d5 14 d3 Íd7 is about equal, W.Lay-J.Tay, Singapore rapid 2012) 12...d5
13 d3 Êf7 14 Îxe7+!? (Kupreichik’s idea) 14...Ëxe7 15 Íd2 can be met by 15...Ëf6 16 Îe1
g5! 17 Ëh5+ Ëg6 18 Ëf3+ and the players agreed a draw in S.Grodzensky-U.Ploder, corre-
spondence 1994.
10...c6 11 Íd3 d5

285
The Schliemann Defence: Move by Move

W________W
[rDbDkDW4]
[0pDWgW0p]
[WDpDW1WD]
[DWDpDWDW]
[WDW0WDWD]
[DWDBDWDW]
[P)P)Q)P)]
[$WGW$WIW]
W--------W
Ho-hum. I was bracing myself for the inevitable draw now.
12 f3!?
A surprise for me.

Question: What is the point of this move? Doesn’t it weaken White’s

kingside structure?

Answer: Well, apart from preventing ...Ëxf2+ when Black castles, White also gives his
queen the f2-square to attack the d4-pawn and perhaps it can even use the g3-square
later.
White can also push the pawn further with 12 f4. If Black responds with 12...Êf7, then
13 c4 dxc3 14 bxc3 Íc5+ (not 14...Îe8? 15 Ëh5+) 15 Íe3 Íxe3+ 16 Ëxe3 Íd7, as in
W.Wittmann-K.Pytel, Holzoster am See 1981, and now 17 Ëf2! Îhf8 18 g3 Êg8 19 a4, keep-
ing an eye on the a-pawn, might give White a little something. So Black should probably
send his king the other way: 12...Êd8 13 c4 (or 13 b3 a5!? 14 Íb2 Íc5 15 Ëf2 Îf8 16 Îf1
g6 17 a3 Íd7 18 Îae1 b5 19 a4 b4 with even play) 13...dxc4 14 Íxc4 Íd6 15 d3 Êc7 16
Íd2 Íd7 and I think Black is at least equal, R.Zelcic-B.Lalic, Croatian Championship, Rijeka
2009.
The other option is 12 b3 0-0 13 Ëxe7 Ëxf2+ 14 Êh1 Íh3!, when 15 gxh3 Ëf3+ 16 Êg1
Ëf2+ with perpetual check is the drawing variation.
White can play on with 15 Îg1!? Îae8 16 Ëxf8+ Ëxf8 17 Îf1 and Sokolov considers this
to be good for White, but Black seems able to hold his own despite the material disparity.
The main drawing idea is to trade a pair of rooks and harass the white king with the queen:
17...Íxg2+! 18 Êxg2 Ëe7 19 Íb2 c5

286
 Tactical Tartakower: 4 Ìc3 fxe4 5 Ìxe4 Ìf6

W________W
[WDWDrDkD]
[0pDW1W0p]
[WDWDWDWD]
[DW0pDWDW]
[WDW0WDWD]
[DPDBDWDW]
[PGP)WDK)]
[$WDWDRDW]
W--------W
20 Îf2 (20 Îf5 can also be met by 20...Îf8, or even 20...Ëe6!? 21 Îaf1 c4 22 bxc4?! dxc4
23 Ía3 g6 24 Îf8+ Îxf8 25 Îxf8+ Êg7 26 Íf1? d3 27 Îf2 Ëg4+ 28 Êh1 Ëd4 29 Êg1, when
a draw was agreed drawn in B.Haas-K.Shoup, correspondence 1995, though in fact
29...dxc2 and a queenside pawn push is now winning for Black) 20...Îf8 21 Îaf1 Îxf2+ 22
Îxf2 leaves White with rook and two bishops for the queen, but his bishops are too passive
and king too exposed for him to have serious hopes of winning. For example: 22...g6 (re-
stricting the light-squared bishop) 23 c3 Ëg5+ 24 Êf1 (or 24 Êh1 Ëh4) 24...Ëg4 25 Íc2 (25
cxd4 Ëd1+ is a draw at once) 25...dxc3 (25...Ëh3+ 26 Êe2 Ëg4+ 27 Êe1 Ëg1+ 28 Îf1 Ëxh2
29 cxd4 might offer some chances) 26 bxc3 (or 26 Íxc3 d4, keeping the dark-squared
bishop a useless one) 26...b5 27 Íc1 a5 28 Íe3 b4 29 Íxc5 (or 29 cxb4 d4 30 Íf4 d3)
29...dxc3 and although White has activated his bishops somewhat, he is still a long way
from finding a safe haven for his king, while Black now has counterplay with the central
pawns.
W________W
[rDbDkDW4]
[0pDWgW0p]
[WDpDW1WD]
[DWDpDWDW]
[WDW0WDWD]
[DWDBDPDW]
[P)P)QDP)]
[$WGW$WIW]
W--------W
287
The Schliemann Defence: Move by Move

12...Êf7
This is more natural than 12...Êd8.

Question: What’s the difference?

Answer: When White attacks the pawn centre with b3-b4, c2-c4, or c2-c3, it is clearly pref-
erable to have the king sitting safely on g8 rather than c7 or d8, especially if White has a
rook ready on c1.
Nevertheless, 12...Êd8!? still seems playable; for example, 13 b3 (or 13 c4 Íd6, intend-
ing 14 cxd5 Ëh4 15 g3 Íxg3! etc) 14...Íd6 14 Íb2 Îf8 15 Îac1 (or 15 Ëf2 c5 16 b4 b6 17
a4 Ëf4 18 g3 Ëxf3 19 Ëxf3 Îxf3 20 Íxh7 Êc7) 15...Íd7 16 c4 dxc3 17 Íxc3 Ëh4 18 Íe5
Îe8 19 f4 Íxe5 20 fxe5 Ëd4+ 21 Ëe3 Ëxe3+ 22 Îxe3 g6 and the position is equal.
13 b3
Note that 13 c4, as in J.Niewold-J.Duriez, correspondence 2000, can be met by 13...dxc3
14 dxc3 Îe8!, since the f3-pawn blocks Ëh5+, and 15 Íe3 Íf5 16 Ëc2 Íxd3 17 Ëxd3 Ëg6
is just equal.
13...Íd6 14 Íb2 Îf8
Preparing artificial castling.
15 Ëf2
Ruifeng starts aiming at the black centre.
15...c5 16 c3
W________W
[rDbDW4WD]
[0pDWDk0p]
[WDWgW1WD]
[DW0pDWDW]
[WDW0WDWD]
[DP)BDPDW]
[PGW)W!P)]
[$WDW$WIW]
W--------W
16...Êg8?
I tried to be too clever here, thinking that f2-f3 was a weak move because of the possi-
bility of ...Ëf4, winning the f3-pawn after losing the one on d4. The correct continuation
was 16...dxc3! 17 Íxc3 (or 17 dxc3 Êg8) 17...d4, which I casually rejected because of 18
Íc4+ Êg6, not realizing that the king is perfectly safe on g6! For instance, after 19 Íb2
Ëf4 20 g3 Ëxf3 21 Ëxf3 Îxf3 22 Íd5 Îf8 23 Îf1 Îxf1+ 24 Îxf1 Íf5 25 Íxb7 Îe8, Black

288
 Tactical Tartakower: 4 Ìc3 fxe4 5 Ìxe4 Ìf6

has the more active pieces in an otherwise equal game.

17 cxd4 Ëf4 18 g3 Ëxf3 19 Ëxf3 Îxf3 20 Íf1
I only now realized that I was in serious trouble! The c5-pawn is in danger after Ía3
and I also need to guard the d5-pawn after Íg2, and then there is the matter of my abys-
mal development.
20...Íd7 21 Íg2 Îf5
This puts the rook on a really odd square, but what choice did I have at this stage?
22 Ía3!
No matter how Black tries to work things out, the c5-pawn cannot be saved. It is futile
to defend it with 22...b6 because of 23 dxc5 Íxc5+ (or 23...bxc5 24 d4 and the c-pawn is a
goner) 24 Íxc5 bxc5 25 Îac1 Îc8 26 Îe7 Îc7 27 Îxd7 Îxd7 28 Íh3 Îff7 29 Íxd7 Îxd7 30
Îxc5 and White has decent winning chances in this ending.
There is nothing left to do except hope that the pawn loss won’t be too significant
(some hope).
22...Íc6 23 Îe6
Gaining a tempo to connect the rooks. Just taking on c5 at once, or inserting 23 Íh3
first, is good too.
23...Íc7!
W________W
[rDWDWDkD]
[0pgWDW0p]
[WDbDRDWD]
[DW0pDrDW]
[WDW)WDWD]
[GPDWDW)W]
[PDW)WDB)]
[$WDWDWIW]
W--------W
Question: Black is clearly much worse. So why award this move an
exclamation mark?

Answer: This is my only chance of getting back into the game. The c5-pawn is lost, but how
should it be captured. 24 dxc5 straightens White’s pawn structure but shuts out his dark-
squared bishop temporarily, while 24 Íxc5 only gives White the advantage of a doubled d-
pawn. Well, the difference between these two captures will become apparent on the next
move.

289
The Schliemann Defence: Move by Move

24 dxc5?
Ironically, the capture that seemingly improves his structure is the wrong option,
whereas after 24 Íxc5! White is probably winning already. The key is that the bishop not
only controls the f8-square, Black also needs to cover the e7-square; e.g. 24...Íd8 25 Îae1
Íg5 26 Î1e2 with a dominant position for White.
24...Íe5!
A sigh of relief for me. My bishop gets out of jail and soon has the best scope among the
remaining pieces. As for the bishop on a3, well, it is now a mere spectator. Of course I am
still a pawn short, but activity is just as important in rapid chess.
25 Îe1 Íd4+ 26 Êh1 Îaf8 27 b4 Îf2 28 Î6e2
I didn’t consider 28 b5!? Íxb5 29 Íxd5 during the game, but Black has enough to se-
cure a draw after 29...Îf1+ 30 Êg2 (or 30 Îxf1 Îxf1+ 31 Êg2 Îf2+ with perpetual, since 32
Êh3? loses to 32...Íd7) 30...Î8f2+ 31 Êh3 Íd7 32 g4! Íxe6 33 Îxe6 Îf3+ 34 Êh4 g5+! (not
34...Îxa3?? 35 Îe8 mate) 35 Êxg5 Êg7 36 Íb4 Îh3 37 Îe7+ Êf8 38 Îe6 Êg7 39 Îe7+ etc.
28...a6
No more b4-b5 tricks
29 Íc1 Î8f5!?
W________W
[WDWDWDkD]
[DpDWDW0p]
[pDbDWDWD]
[DW)pDrDW]
[W)WgWDWD]
[DWDWDW)W]
[PDW)R4B)]
[DWGW$WDK]
W--------W
Question: What is the point of this rook lift?

30 a3 Êf7
Answer: That’s the point, to clear the path for the king to advance, especially if White wants
to liquidate my hyperactive rooks. I hope at least to get my king sufficiently centralized to
issue threats to White’s pawns.
31 d3
The line I was roughly aiming at was 31 Îxf2 Îxf2 32 Îf1 Îxf1+ 33 Íxf1 Êf6 34 Êg2
Êe5 35 Íd3 g6 36 Êf3 Ía1 and I doubt White can make any headway; e.g. 37 Íc2 (or 37

290
 Tactical Tartakower: 4 Ìc3 fxe4 5 Ìxe4 Ìf6

Êe3 Íd4+ 38 Êe2 Ía1) 37...Êd4 38 Êg4 Êc4 39 Êg5 Íg7 and Black is alright.
31...Îxe2 32 Îxe2 Êf6!
Forward of course!
33 Íe3?!
White should try 33 g4!? Îe5 34 g5+ Êe7 35 Îxe5+ Íxe5, though Black is surely edging
closer to a draw, even with a pawn deficit.
33...Íxe3 34 Îxe3 Îf2
W________W
[WDWDWDWD]
[DpDWDW0p]
[pDbDWiWD]
[DW)pDWDW]
[W)WDWDWD]
[)WDP$W)W]
[WDWDW4B)]
[DWDWDWDK]
W--------W
I believe I have done enough to get a draw. To my surprise, White decided to liquidate
when he really shouldn’t!
35 Îf3+??
Thanks to Black’s advanced king, this rook trade leads to a lost ending for White. There
is no way to stop the black king from marching down the queenside to collect the a- and b-
pawns. Instead, 35 Êg1 Îa2 is just a drawn ending, as Black will quickly regain his missing
pawn.
35...Îxf3 36 Íxf3 Êe5
The king cannot be denied from his queenside lunch.
37 Êg2 Êd4 38 Íe2 Êc3 39 Êf3 Êb3 40 Êf4 Êxa3 41 Êe5 Êxb4 42 d4 a5 43 Íd3 h6 44 Êe6
Êc3 45 Íb1 Êxd4 46 Êf7 a4 47 Êxg7 Êxc5 48 Êxh6 a3 49 g4 d4 50 g5 d3! 51 g6 Íd5 0-1

Game 35
A.Berescu-D.Dinic
Felix Spa 2007

1 e4 e5 2 Ìf3 Ìc6 3 Íb5 f5 4 Ìc3 fxe4 5 Ìxe4 Ìf6 6 Ìxf6+ Ëxf6 7 0-0 Ìd4 8 Ìxd4 exd4 9
b3

291