Polynomials

Adithya B., Brian L., William W., Daniel X.

4/22

§1 Introduction
Definition 1.1. A polynomial is a function in one variable of the form

P (x) = an xn + an−1 xn−1 + . . . + a1 x + a0

where n is a nonnegative integer and the constants a0 , a1 , . . . an are the coefficients of

6 0, then we say that the degree of the polynomial is n, and denote
the polynomial. If an =
this by deg P. We’ll mostly consider polynomials where the coefficients are real numbers.
Definition 1.2. A complex number z is a root of a polynomial P if P (z) = 0.
An important fact about one-variable polynomials is:

Theorem 1.3 (Fundamental Theorem of Algebra)

Any polynomial P (x) with complex coefficients has at least one complex root, and
has exactly deg P roots up to multiplicity.

Definition 1.4. We say that a polynomial Q(x) is a factor of a polynomial P (x) if there
exists a polynomial R(x) such that P (x) = Q(x)R(x).
For example, x + 1 is a factor of x2 − 1 (as x2 − 1 = (x − 1)(x + 1)) but x is not a
factor of x2 − 1. This is similar to the definition for integers: an integer b is a factor of a
if there exists a third integer c for which a = bc. The next theorem connects factors of a
polynomial with its roots:

Theorem 1.5 (Factor Theorem)

For a one-variable polynomial P (x) and complex number z, P (z) = 0 if and only if
x − z is a factor of P (x).

Proof. One direction is easy: if x − z is a factor of P (x), then

P (x) = (x − z)Q(x)

for some polynomial Q. Then

P (z) = (z − z)Q(z) = 0.

The converse is true as well: if P (z) = 0, then x − z is a factor of P (x). We’ll see why
this is true when we discuss the Remainder Theorem.

1
Adithya B., Brian L., William W., Daniel X. (4/22) Polynomials

Combining the Fundamental Theorem of Algebra and the Factor Theorem, we conclude
the following important fact: if z1 , z2 , . . . , zn are the roots of a polynomial P (counted
with multiplicity), then P can be written in the form

P (x) = c(x − z1 )(x − z2 ) · · · (x − zn )

for some constant c.

§2 Vieta’s Formulas
As we’ve seen, if r1 , r2 , . . . , rn are the roots of the polynomial P (x) = an xn + an−1 xn−1 +
. . . + a1 x + a0 , then we may write

an xn + an−1 xn−1 + . . . + a1 x + a0 = an (x − r1 )(x − r2 ) · · · (x − rn )

Let’s expand the product of the n linear factors on the right-hand side. Then we get

an xn −an (r1 +r2 +. . . +rn )xn−1 +an (r1 r2 +r1 r3 +. . . +rn−1 rn ) +. . . + (−1)n an r1 r2 . . . rn

where the coefficient of xk is (−1)n−k an times the sum of all possible products of n − k
of the roots. But we know these coefficients; they’re precisely a0 , a1 , . . . , an . Now if we
set the coefficient of xk equal to ak we see that
an−1
r1 + r2 + . . . + rn = −
an
an−2
r1 r2 + r1 r3 + . . . + rn−1 rn =
an
...
a0
r1 r2 · · · rn = (−1)n
an
These equations relating the roots of the polynomial to its coefficients are called Vieta’s
Formulas, and can help use to compute various expressions involving the roots of a
polynomial without having to compute the roots themselves.

Example 2.1 (2008 AIME II # 7)

Let r, s, and t be the three roots of the equation

8x3 + 1001x + 2008 = 0.

Find (r + s)3 + (s + t)3 + (t + r)3 .

Solution. From Vieta’s Formulas, we know the values

1001 2008
r + s + t = 0, rs + st + tr = , rst = − = −251.
8 8
From the first equation, we immediately see that r + s = −t and (r + s)3 = −t3 , so

(r + s)3 + (s + t)3 + (t + r)3 = −(r3 + s3 + t3 ).

Our goal is to express r3 + s3 + t3 in terms of r + s + t, rs + st + tr, and rst, at which

point we can substitute in numerical values and find the answer.

2
Adithya B., Brian L., William W., Daniel X. (4/22) Polynomials

First, let’s cube r + s + t because we need r3 + s3 + t3 to appear:

(r + s + t)3 = r3 + s3 + t3 + 3(r2 s + s2 t + t2 r + rs2 + st2 + tr2 ) + 6rst.

Now we need to get rid of all of the terms appearing after r3 + s3 + t3 . To get rid of 3r2 s,
we can try subtracting 3(r + s + t)(rs + st + tr) (because this contains a 3r2 s term):

3(r + s + t)(rs + sr + tr) = 3(r2 s + s2 t + t2 r + rs2 + st2 + tr2 ) + 9rst.

So now we have

(r + s + t)3 − 3(r + s + t)(rs + st + tr) = r3 + s3 + t3 − 3rst.

Now all we have to do is add 3rst to both sides:

r3 + s3 + t3 = (r + s + t)3 − 3(r + s + t)(rs + st + tr) + 3rst.

From here, we just have to plug in the numbers. Since r + s + t = 0, most of the terms
vanish:
−r3 − s3 − t3 = −3rst = (−3)(−251) = 753 .
Alternatively, you may be aware of the factorization
1
r3 + s3 + t3 − 3rst = (r + s + t)((r − s)2 + (s − t)2 + (t − r)2 )
2
which makes it much easier to express r3 + s3 + t3 in terms of r + s + t, rs + st + tr, and
rst.

Example 2.2 (2013 HMMT Algebra # 5)

Let a and b be real numbers, and let r,s, and t be the roots of f (x) = x3 +ax2 +bx−1.
Also, g(x) = x3 + mx2 + nx + p has roots r2 ,s2 , and t2 . if g(−1) = −5, find the
maximum possible value of b.

Solution. From Vieta’s formulas, we know that

−a = r + s + t, b = rs + st + tr, 1 = rst

and
m = −(r2 + s2 + t2 ), n = r2 s2 + s2 t2 + t2 r2 , p = −r2 s2 t2 .
Let’s try to express m, n, and p in terms of a and b. The easiest one is p :

p = −(rst)2 = −1.

For m, we square r + s + t :

a2 = (r + s + t)2 = r2 + s2 + t2 + 2rs + 2st + 2tr = −m + 2b =⇒ m = 2b − a2 .

Finaly, for n we can square rs + st + tr :

b2 = (rs+st+tr)2 = r2 s2 +s2 t2 +t2 r2 +2r2 st+2rs2 t+2rst2 = n+2rst(r+s+t) = n−2a =⇒ n = b2 +2a.

So now we write g(x) in terms of a and b :

g(x) = x3 + (2b − a2 )x2 + (b2 + 2a)x − 1.

3
Adithya B., Brian L., William W., Daniel X. (4/22) Polynomials

We know that g(−1) = −5; when we plug this into our equation for g(x) we get

−5 = (−1)3 +(2b−a2 )(−1)2 +(b2 +2a)(−1)−1 = 2b−a2 −b2 −2a−2 =⇒ a2 +2a+(b2 −2b−3) = 0.

We seek the largest possible value of b; since a is real, we know that the discriminant of
this quadratic must be nonnegative. In particular,

22 − 4(b2 − 2b − 3) ≥ 0 =⇒ b2 − 2b − 4 ≤ 0.
√
Solving this quadratic gives us that the largest possible value of b is 1 + 5.

A second solution is possible by factoring f (x) and g(x); try the same approach as in
Example 3.2.

§3 Factorization
If we have an expression of the form P (x) = Q(x) · something, then we immediately know
that Q(x) is a factor of P (x). As we will see, this can be very useful information.

Example 3.1 (2016 AIME I # 11)

Let P (x) be a nonzero polynomial such that (x − 1)P (x + 1) = (x + 2)P (x) for every
real x, and (P (2))2 = P (3). Then P 72 = m


n , where m and n are relatively prime
positive integers. Find m + n.

Solution. Let’s look at what linear polynomials are factors of P (x). Since x − 1 appears
on the left hand side, it must be a factor of the right hand side. So we immediately know
that x − 1 is a factor of (x + 2)P (x), and so it is a factor of P (x). Similarly, x + 2 appears
on the right hand side, so it must appear on the left. Thus, we know that x + 2 is a
factor of P (x + 1), or that x + 1 is a factor of P (x). So we may write

P (x) = (x − 1)(x + 1)Q(x)

for some polynomial Q(x). Let’s plug back into the original equation:

(x − 1)x(x + 2)Q(x + 1) = (x + 2)(x − 1)(x + 1)Q(x) =⇒ xQ(x + 1) = (x + 1)Q(x).

Note that x appears on the left hand side, so x must be a factor of Q(x). As x + 1 appears
on the right hand side, it must be a factor of Q(x + 1), but this also implies that x is a
factor of Q(x). Now we can write

Q(x) = xR(x)

for a polynomial R(x). So we have

x(x + 1)R(x) = (x + 1)(x)R(x) =⇒ R(x) = R(x + 1).

Now we know that R is constant! One way to see this is to see that

· · · = R(−2) = R(−1) = R(0) = R(1) = R(2) = · · · ,

which implies that R(x)−R(0) has a root at every integer and hence is the zero polynomial.
Thus R(x) = R(0) for all real x. Suppose that R(x) is the constant polynomial c; then
we get

Q(x) = xR(x) = cx, P (x) = (x − 1)(x + 1)Q(x) = cx(x − 1)(x + 1).

4
Adithya B., Brian L., William W., Daniel X. (4/22) Polynomials

In order to find c, we can use the given numerical information about P (2) and P (3) :
P (2) = c · 2 · 1 · 3 = 6c, P (3) = c · 3 · 2 · 4 = 24c.
Since (P (2))2 = P (3), we have
2
(6c)2 = 24c =⇒ 36c2 = 24c =⇒ c =
3
(we know that
c 6= 0 because otherwise P would be the zero polynomial). Now we can
calculate P 27 :  
7 2 7 5 9 105
P = · · · =
2 3 2 2 2 4
and the answer is 105 + 4 = 109 .
It can often be useful to consider the factorization of a polynomial given by the
Fundamental Theorem of Algebra:
P (x) = an xn + an−1 xn−1 + . . . + a1 x + a0 = an (x − r1 )(x − r2 ) . . . (x − rn ).
Thus, if we are faced with expressions similar to (z − r1 )(z − r2 ) · · · (z − rn ), we can try
to relate them to P (z). For instance,
P (1)
(1 − r1 )(1 − r2 ) . . . (1 − rn ) =
an

Example 3.2 (2014 USAMO/1)

Let a, b, c, d be real numbers such that b − d ≥ 5 and all zeros x1 , x2 , x3 , and x4 of
the polynomial P (x) = x4 + ax3 + bx2 + cx + d are real. Find the smallest value the
product (x21 + 1)(x22 + 1)(x23 + 1)(x24 + 1) can take.

Solution. The condition b − d ≥ 5 is strange; it seems a bit arbitrary and disconnected

from what we wish to find. In these situations, it’s often a good idea to first ignore these
conditions and see how they can be used later on.
We wish to find (x21 + 1)(x22 + 1)(x23 + 1)(x24 + 1). We can use the same Vieta methods
as we did for Example 2.2, but with four variables this is much bashier and more painful.
Let’s try a different approach.
We know that
P (x) = (x − x1 )(x − x2 )(x − x3 )(x − x4 )
so if we factor the difference of squares, we get
(x − x21 )(x − x22 )(x − x23 )(x − x24 )
√ √ √ √ √ √ √ √
= ( x − x1 )( x − x2 )( x − x3 )( x − x4 )( x + x1 )( x + x2 )( x + x3 )( x + x4 )
√ √
= P ( x)P (− x).
Let’s try this on our given expression:
(x21 + 1)(x22 + 1)(x23 + 1)(x24 + 1) = (x21 − i2 )(x22 − i2 )(x23 − i2 )(x24 − i2 )
= (x1 − i)(x2 − i)(x3 − i)(x4 − i)(x1 + i)(x2 + i)(x3 + i)(x4 + i)
= P (i)P (−i)
= (1 − ai − b + ci + d)(1 + ai − b − ci + d)
= (1 − b + d + (c − a)i)(1 − b + d − (c − a)i)
= (b − d − 1)2 + (c − a)2

5
Adithya B., Brian L., William W., Daniel X. (4/22) Polynomials

Now it’s clear where the condition b − d ≥ 5 comes in: we have

(b − d − 1)2 + (c − a)2 ≥ (5 − 1)2 + 02 = 16.

But we’re not done yet; we need to exhibit a polynomial P (x) that achieves the value 16.
For this, we need b − d = 5 and c − a = 0 to hold. Fortunately, there’s an easy polynomial
that satisfies this condition:

(x + 1)4 = x4 + 4x3 + 6x2 + 4x + 1

and we conclude that the answer is indeed 16 .

§4 Remainder Theorem
The Remainder Theorem is a very useful generalization of the Factor Theorem. While
the result itself can be powerful, the ideas behind its proof are far more applicable to a
wide variety of problems. First, a definition:

Definition 4.1. The remainder of P (x) when it is divided by Q(x) is the unique
polynomial R(x) with deg R < deg Q such that Q(x) is a factor of P (x) − R(x). In other
words, if we write
P (x) = Q(x)S(x) + R(x)
for polynomials R(x) and S(x) with deg R < deg Q, then the remainder is R(x). For
instance, since x2 − x + 1 = x(x − 1) + 1, the remainder when x2 − x + 1 is divided by x
is 1.

Theorem 4.2 (Remainder Theorem)

The remainder when P (x) is divided by x − c is P (c).

Proof. Write
P (x) = (x − c)Q(x) + R(x),
where R(x) is our desired remainder. Since deg R < deg (x − c) = 1 we see that R must
be a constant polynomial. Suppose we plug in x = c into the above equation:

P (c) = (c − c)Q(c) + R(c) = R(c).

Since R is a constant polynomial, we see that R(x) = P (c) for all x.

Notice that this immediately implies the Factor Theorem: x − c is a factor of P (x) if and
only if the remainder R(x) = 0, which occurs if and only if P (c) = 0. The ideas behind
this proof (writing out the divison and remainder and plugging in suitable values) allow
us to tackle much more than the remainder upon division by linear polynomials.

Example 4.3 (AoPS Intermediate Algebra)

Find the remainder when x100 − 4x98 + 5x + 6 is divided by x3 − 2x2 − x + 2.

6
Adithya B., Brian L., William W., Daniel X. (4/22) Polynomials

Solution. Since the degree of x3 − 2x2 − x + 2 is 3, the remainder has degree at most 2,
so we can express it as ax2 + bx + c for constants a, b, c. Let’s write out the division:
x100 − 4x98 + 5x + 6 = (x3 − 2x2 − x + 2)Q(x) + ax2 + bx + c.
However, we note that we can actually factor x3 − 2x2 − x + 2 :
x100 − 4x98 + 5x + 6 = (x − 2)(x − 1)(x + 1)Q(x) + ax2 + bx + c.
As in the proof of the Remainder Theorem, we substitute in x = 2, 1, −1, which makes
the term containing Q(x) equal to zero:
4a + 2b + c = 2100 − 4 · 298 + 5 · 2 + 6 = 16
a + b + c = 1100 − 4 · 198 + 5 · 1 + 6 = 8
a − b + c = (−1)100 − 4 · (−1)98 + 5(−1) + 6 = −2.
We can solve this system of linear equations to get (a, b, c) = (1, 5, 2), so the remainder is
x2 + 5x + 2 .

Example 4.4 (2016 PUMaC Algebra A # 4)

Suppose that P is a polynomial with integer coefficients such that P (1) = 2, P (2) = 3
and P (3) = 2016. If N is the smallest possible positive value of P (2016), find the
remainder when N is divided by 2016.

Solution. Consider what happens when we divide P (x) by (x − 1)(x − 2). We get
P (x) = (x − 1)(x − 2)Q(x) + R(x),
where R(x) has degree 1. Then, if we plug in x = 1 and x = 2, we get R(1) = 2 and
R(2) = 3. Since R is linear, we have that R(x) = x + 1. Therefore,
P (x) = (x − 1)(x − 2)Q(x) + x + 1.
If we plug in x = 3, then we obtain 2016 = 2Q(3) + 4, which yields Q(3) = 1006. Now, if
we plugged in x = 2016, then we get P (2016) = 2015 · 2014Q(2016) + 2017. To finish, we
will utilize the following lemma.

Lemma 4.5
If P (x) is a polynomial with integer coefficients, and a, b are integers with a 6= b,
then (a − b)|(P (a) − P (b)).

Proof. Let P (x) = cn xn + cn−1 xn−1 + · · · + c1 x + c0 . Then,

P (a) − P (b) = cn (an − bn ) + cn−1 (an−1 − bn−1 ) + · · · + c1 (a − b).
Each term in this sum is divisible by a−b, and since P has integer coefficients, P (a)−P (b)
is divisible by a − b.
Now, we can use the lemma to obtain 2013|(Q(2016) − Q(3)). The minimum value of
Q(2016) such that P (2016) ≥ 0 is Q(2016) = Q(3) = 1006. With this value, we get
P (2016) = 2015 · 2014 · 1006 + 2017 ≡ (−1)(−2)(1006) + 2017 ≡ 2013 (mod 2016).

Notice that in these two problems, we never cited the Remainder Theorem; rather, we
used and generalized the ideas behind its proof.

7
Adithya B., Brian L., William W., Daniel X. (4/22) Polynomials

§4.1 Aside on Lagrange Interpolation

There are many possible problems of the form, ”P (x) is the polynomial of least degree
so that P (x) = something for x = 1, 2, . . . , n. Find P (n + 1).” We’ll give an example of
the general strategy that typically works on problems like these, then discuss a messier
alternative.

Example 4.6 (2017 HMMT Algebra and Number Theory #6)

1
A polynomial P of degree 2015 satisfies the equation P (n) = n2
for n = 1, 2, . . . , 2016.
Find b2017P (2017)c.

Solution. We know that P (x) − x12 is equal to zero at x = 1, 2, . . . , 2016, so we’d like
to say that this polynomial is divisible by (x − 1)(x − 2) · · · (x − 2016). But we can’t,
because P (x) − x12 isn’t a polynomial! Fortunately, there’s an easy fix:
1
P (x) − = 0 =⇒ x2 P (x) − 1 = 0,
x2
so x2 P (x) − 1 is a polynomial divisible by (x − 1)(x − 2) · · · (x − 2016). Since P has degree
2015, x2 P (x) − 1 has degree 2017. Thus we may write
x2 P (x) − 1 = c(x − 1)(x − 2) · · · (x − 2016)(x − k)
for some constants c, k. In other words,
x2 P (x) = c(x − 1)(x − 2) · · · (x − 2016)(x − k) + 1
Thus the right-hand side must be divisible by x2 , so it constant and linear coefficients
must both be zero. If we expand the right-hand side, then the constant term is
c(−1)(−2)(−2016)(−k) + 1 = 1 − 2016!ck
and the linear coefficient is
c · 2016! − ck((−2)(−3) · · · (−2016) + (−1)(−3) · · · (−2016) + . . . (−1)(−2) · · · (−2015))
 
1 1
= c · 2016! + ck · 2016! 1 + + · · · + = 2016!c(1 + kH2016 )
2 2016
where we let H2016 = 1 + 12 + · · · + 2016
1
for convenience. Now we set the linear coefficient
to zero:
1
2016!c(1 + kH2016 ) = 0 =⇒ k = −
H2016
and the constant coefficient to zero:
1 H2016
1 − 2016!ck = 0 =⇒ c = =−
2016!k 2016!
Thus we may write
H2016 1
x2 P (x) = 1 − (x − 1)(x − 2) · · · (x − 2016)(x + )
2016! H2016
and we get
H2016 1
20172 P (2017) = 1 − · 2016! · (2017 + ) = −2017H2016
2016! H2016
so the desired value is 2017P (2017) = −H2016 . We leave the approximation b−H2016 c =
−9 to the interested reader.

8
Adithya B., Brian L., William W., Daniel X. (4/22) Polynomials

There is actually a general formula for the polynomial in problems of this form. We
state it below.

Theorem 4.7 (Lagrange Interpolation Formula)

Let a1 , a2 , . . . an , b1 , b2 , . . . bn be real numbers. The unique polynomial P of degree
≤ n − 1 such that
P (a1 ) = b1 , P (a2 ) = b2 , . . . P (an ) = bn
is
n
X (x − a1 )(x − a2 ) · · · (x − ai−1 )(x − ai+1 ) · · · (x − an )
P (x) = bi
(ai − a1 )(ai − a2 ) · · · (ai − ai−1 )(ai − ai+1 ) · · · (ai − an )
i=1

For instance, in the polynomial from example 4.4, the unique polynomial of degree ≤ 2
satisfying P (1) = 2, P (2) = 3, P (3) = 2016 is

(x − 2)(x − 3) (x − 1)(x − 3) (x − 1)(x − 2)

P (x) = ·2+ ·3+ · 2016.
(1 − 2)(1 − 3) (2 − 1)(2 − 3) (3 − 1)(3 − 2)

From this example, you may be able to see the structure behind the formula. The ith
term of the sum is equal to bi when x = ai and is equal to zero when x is equal to any
other aj . This way, when we sum all n terms, we get a polynomial that takes the value bi
at every x = ai . The strategy we used to solve Example 4.5 is usually the much cleaner
approach, but sometimes it is not viable. In these cases, we must either use trickier
algebraic manipulations or bash with Lagrange Interpolation.

§5 Quadratics
Like any other polynomial, we can express a quadratic in terms of its roots r, s:

ax2 + bx + c = a(x − r)(x − s).

However, quadratics possess a unique property: they are symmetric. Thus we have
another method of representing a quadratic, its vertex form:

ax2 + bx + c = a(x − h)2 + k

where (h, k) is the vertex of the quadratic. These ways of representing quadratics can
come in handy in various circumstances.

Example 5.1 (2020 AIME I # 14)

Let P (x) be a quadratic polynomial with complex coefficients whose x2 coefficient is
1. Suppose the equation P (P (x)) = 0 has four distinct solutions, x = 3, 4, a, b. Find
the sum of all possible values of (a + b)2 .

Solution. Since the x2 coefficient of P (x) is 1, if r and s are the two roots of P (x) then

P (x) = (x − r)(x − s).

If P (P (x)) = 0 then we must have either P (x) = r or P (x) = s. Each of these quadratics
have exactly two solutions, so we know that two of P (3), P (4), P (a), P (b) are equal to r

9
Adithya B., Brian L., William W., Daniel X. (4/22) Polynomials

and the other two are equal to s. We consider two different cases:

Case 1. P (3) = P (4). Without loss of generality, let P (3) = P (4) = r; then we have
that a and b are the two solutions to P (x) = s. In other words, 3 and 4 are the solutions
to
(x − r)(x − s) = r =⇒ x2 − (r + s)x + (rs − r) = 0
while a and b are the solutions to

(x − r)(x − s) = s =⇒ x2 − (r + s)x + (rs − s) = 0.

By Vieta on the second equation, we see that a + b = r + s. But by Vieta on the first
equation, we have that r + s = 3 + 4 = 7. So in this case, we have (a + b)2 = 72 = 49.

Case 2. P (3) 6= P (4). In this case, let P (3) = r and P (4) = s, and assume that
P (a) = r and P (b) = s. Then 3 and a are the solutions to

(x − r)(x − s) = r =⇒ x2 − (r + s)x + (rs − r) = 0

so by Vieta 3 + a = r + s and a = r + s − 3. Similarly, 4 and b are the solutions to

(x − r)(x − s) = s =⇒ x2 − (r + s)x + (rs − s) = 0.

so by Vieta 4 + b = r + s and b = r + s − 4. Thus a + b = 2r + 2s − 7, so all we need is to

find r + s. Let’s write out the equations P (3) = r and P (4) = s :

(3 − r)(3 − s) = r =⇒ rs − 3r − 3s + 9 = r,
(4 − r)(4 − s) = s =⇒ rs − 4r − 4s + 16 = s.

When we subtract the first equation from the second equation, we can actually solve for
s!
7
r − s = (rs − 3r − 3s + 9) − (rs − 4r − 4s + 16) = r + s − 7 =⇒ s = .
2
Now we can substitute this value of s into either of our two original equations to solve
for r :  
7 7
(4 − r) 4 − = =⇒ r = −3.
2 2
Thus we have
7
a + b = 2r + 2s − 7 = 2 · + 2(−3) − 7 = −6
2
so (a + b)2 = 36 in this case.

Finally, the requested sum is 49 + 36 = 85 .

Example 5.2 (ARML 2017 Tiebreaker # 1)

Compute the least positive N such that there exists a quadratic polynomial f (x)
with integer coefficients satisfying

f (f (1)) = f (f (5)) = f (f (7)) = f (f (11)) = N.

10
Adithya B., Brian L., William W., Daniel X. (4/22) Polynomials

Solution. The first thing to note is that the given inputs are symmetric around x = 6.
Since a quadratic has an axis of symmetry, this suggests that x = 6 is our axis of
symmetry. We can confirm this as follows: If f (x) has an axis of symmetry around x = n,
then f (f (x)) does as well. Indeed, if f (x) is symmetric around x = n then

f (x) = f (2n − x) =⇒ f (f (x)) = f (f (2n − x))

which is equivalent to f (f (x)) being symmetric around x = n. Now f (f (x)) is a quartic,

so f (f (x)) − N has exactly four roots. Clearly, f (f (x)) − N is also symmetric around
x = n, so these four roots are symmetric about x = n. But these four roots are 1, 5, 7, 11,
and the only possible value for the value of n is 6.
Since f (x) has axis of symmetry x = n, we can write f (x) in vertex form:

f (x) = a(x − 6)2 + k

for nonzero a. If we substitute in x = 1, 5, 7, 11, we obtain

f (5) = f (7) = a + k, f (1) = f (11) = 25a + k.

So now we know that f (a + k) = f (25a + k) = N. But a + k 6= 25a + k, so the only way

we can have f (a + k) = f (25a + k) is if they are symmetric about 6; that is,

a + k = 2 · 6 − (25a + k) =⇒ k = 6 − 13a.

Thus we have
f (x) = a(x − 6)2 + 6 − 13a.
Now we want to compute N = f (f (1)) we see that

f (1) = 6 + 12a =⇒ N = f (f (1)) = 144a3 − 13a + 6.

Since f (x) = a(x − 6)2 + 6 − 13a must have integer coefficients, we see that a must be a
nonzero integer. Thus we want the smallest positive value of 144a3 − 13a + 6 over all
nonzero integers. We leave it as an exercise to check that this is minimized at a = 1, for
an answer of 144 − 13 + 6 = 137 .

11
Adithya B., Brian L., William W., Daniel X. (4/22) Polynomials

§6 Problems
Problem 6.1 (2017 AMC 12A #23). For certain real numbers a, b, and c, the polynomial
g(x) = x3 + ax2 + x + 10 has three distinct roots, and each root of g(x) is also a root of
the polynomial f (x) = x4 + x3 + bx2 + 100x + c. What is f (1)?

Problem 6.2 (PUMaC 2016 Algebra #2). Let f (x) = 15x−2016. If f (f (f (f (f (x))))) =
f (x), find the sum of all possible values of x.

Problem 6.3 (CMIMC 2018 Algebra #3). Let P (x) = x2 + 4x + 1. What is the product
of all real solutions to the equation P (P (x)) = 0?

Problem 6.4 (1996 AIME #5). Suppose that the roots of x3 + 3x2 + 4x − 11 = 0 are
a, b, and c, and that the roots of x3 + rx2 + sx + t = 0 are a + b, b + c, and c + a. Find t.

Problem 6.5 (2001 AIME I #3). Find the sum of the roots, real and non-real, of the

2001
equation x2001 + 21 − x = 0, given that there are no multiple roots.

Problem 6.6 (2010 AMC 12A # 21). The graph of y = x6 − 10x5 + 29x4 − 4x3 + ax2
lies above the line y = bx + c except at three values of x, where the graph and the line
intersect. What is the largest of those values?

Problem 6.7 (2010 AMC 12B # 23). Monic quadratic polynomials P (x) and Q(x) have
the property that P (Q(x)) has zeroes at x = −23, −21, −17, and − 15, and Q(P (x)) has
zeroes at x = −59, −57, −51, and − 49. What is the sum of the minimum values of P (x)
and Q(x)?

Problem 6.8 (2014 HMMT Algebra #4). Let b and c be real numbers, and define
the polynomial P (x) = x2 + bx + c. Suppose that P (P (1)) = P (P (2)) = 0, and that
P (1) 6= P (2). Find P (0).

Problem 6.9 (Purple Comet 2010 #25). Let x1 , x2 , and x3 be the roots of the polynomial
x3 + 3x + 1. There are relatively prime positive integers m and n such that

m x21 x22 x23

= + + .
n (5x2 + 1)(5x3 + 1) (5x1 + 1)(5x3 + 1) (5x1 + 1)(5x2 + 1)

Find m + n.

Problem 6.10 (2010 AIME I #6). Let P (x) be a quadratic polynomial with real
coefficients satisfying
x2 − 2x + 2 ≤ P (x) ≤ 2x2 − 4x + 3
for all real numbers x, and suppose P (11) = 181. Find P (16).

Problem 6.11 (2007 AIME I # 8). The polynomial P (x) is cubic. What is the
largest value of k for which the polynomials Q1 (x) = x2 + (k − 29)x − k and Q2 (x) =
2x2 + (2k − 43)x + k are both factors of P (x)?

Problem 6.12 (2015 AIME I #10). Let f (x) be a third-degree polynomial with real
coefficients satisfying

|f (1)| = |f (2)| = |f (3)| = |f (5)| = |f (6)| = |f (7)| = 12.

Find |f (0)|.

12
Adithya B., Brian L., William W., Daniel X. (4/22) Polynomials

Problem 6.13 (USAMO 1984 #1). The product of two of the four roots of the quartic
equation x4 − 18x3 + kx2 + 200x − 1984 = 0 is −32. Determine the value of k.

Problem 6.14. Find the remainder when x28 + 1 is divided by x4 + x3 + x2 + x + 1.

Problem 6.15 (2016 PUMaC Algebra A # 7). Let SP be the set of all polynomials
P with complex coefficients, such that P (x2 ) = P (x)P (x − 1) for all complex numbers
x. Suppose P0 is the polynomial in SP of maximal degree such that P0 (1)|2016. Find
P0 (10).

Problem 6.16 (2020 HMMT Algebra and Number Theory #8). Let P (x) be the unique
polynomial of degree at most 2020 satisfying P (k 2 ) = k for k = 0, 1, 2, ..., 2020. Compute
P (20212 ).

Problem 6.17 (1984 AIME 1 # 15). Determine w2 + x2 + y 2 + z 2 if

x2 y2 z2 w2
+ + + =1
22 − 1 22 − 32 22 − 52 22 − 72
x2 y2 z2 w2
+ 2 + 2 + 2 =1
42 − 1 4 − 32 4 − 52 4 − 72
x2 y2 z2 w2
+ 2 + 2 + 2 =1
62 − 1 6 − 32 6 − 52 6 − 72
x2 y2 z2 w2
+ 2 + 2 + 2 = 1.
82 − 1 8 − 32 8 − 52 8 − 72
Problem 6.18 (IMO Shortlist 2005/N3). Let a, b, c, d, e, f be positive integers and let
S = a + b + c + d + e + f . Suppose that the number S divides abc + def and
ab + bc + ca − de − ef − df . Prove that S is composite.

13