IMO 2005 Solution Notes

Evan Chen《陳誼廷》
26 March 2024

This is a compilation of solutions for the 2005 IMO. Some of the solutions
are my own work, but many are from the official solutions provided by the
organizers (for which they hold any copyrights), and others were found by
users on the Art of Problem Solving forums.
These notes will tend to be a bit more advanced and terse than the “official”
solutions from the organizers. In particular, if a theorem or technique is not
known to beginners but is still considered “standard”, then I often prefer to
use this theory anyways, rather than try to work around or conceal it. For
example, in geometry problems I typically use directed angles without further
comment, rather than awkwardly work around configuration issues. Similarly,
sentences like “let R denote the set of real numbers” are typically omitted
entirely.
Corrections and comments are welcome!

Contents
0 Problems 2

1 Solutions to Day 1 3
1.1 IMO 2005/1, proposed by Bogdan Enescu (ROU) . . . . . . . . . . . . . . 3
1.2 IMO 2005/2, proposed by Nicholas de Bruijn (NLD) . . . . . . . . . . . . 4
1.3 IMO 2005/3, proposed by Hojoo Lee (KOR) . . . . . . . . . . . . . . . . . 6

2 Solutions to Day 2 8
2.1 IMO 2005/4, proposed by Mariusz Skalba (POL) . . . . . . . . . . . . . . 8
2.2 IMO 2005/5, proposed by Waldemar Pompe (POL) . . . . . . . . . . . . . 9
2.3 IMO 2005/6, proposed by Radu Gologan, Dan Schwartz (ROU) . . . . . . 10

1
IMO 2005 Solution Notes web.evanchen.cc, updated 26 March 2024

§0 Problems
1. Six points are chosen on the sides of an equilateral triangle ABC: A1 , A2 on BC,
B1 , B2 on CA and C1 , C2 on AB, such that they are the vertices of a convex
hexagon A1 A2 B1 B2 C1 C2 with equal side lengths. Prove that the lines A1 B2 , B1 C2
and C1 A2 are concurrent.

2. Let a1 , a2 , . . . be a sequence of integers with infinitely many positive and negative

terms. Suppose that for every positive integer n the numbers a1 , a2 , . . . , an leave
n different remainders upon division by n. Prove that every integer occurs exactly
once in the sequence.

3. Let x, y, z > 0 satisfy xyz ≥ 1. Prove that

x5 − x2 y5 − y2 z5 − z2
+ + ≥ 0.
x5 + y 2 + z 2 x2 + y 5 + z 2 x2 + y 2 + z 5

4. Determine all positive integers relatively prime to all the terms of the infinite
sequence
an = 2n + 3n + 6n − 1, n ≥ 1.

5. Let ABCD be a fixed convex quadrilateral with BC = DA and BC ∦ DA. Let

two variable points E and F lie on the sides BC and DA, respectively, and satisfy
BE = DF . The lines AC and BD meet at P , the lines BD and EF meet at Q,
the lines EF and AC meet at R. Prove that the circumcircles of the triangles
P QR, as E and F vary, have a common point other than P .

6. In a mathematical competition 6 problems were posed to the contestants. Each

pair of problems was solved by more than 25 of the contestants. Nobody solved all
6 problems. Show that there were at least 2 contestants who each solved exactly 5
problems.

2
IMO 2005 Solution Notes web.evanchen.cc, updated 26 March 2024

§1 Solutions to Day 1
§1.1 IMO 2005/1, proposed by Bogdan Enescu (ROU)
Available online at https://aops.com/community/p281571.

Problem statement

Six points are chosen on the sides of an equilateral triangle ABC: A1 , A2 on BC, B1 ,
B2 on CA and C1 , C2 on AB, such that they are the vertices of a convex hexagon
A1 A2 B1 B2 C1 C2 with equal side lengths. Prove that the lines A1 B2 , B1 C2 and C1 A2
are concurrent.

The six sides of the hexagon, when oriented, comprise six vectors with vanishing sum.
However note that
−−−→ −−−→ −−−→
A1 A2 + B1 B2 + C1 C2 = 0.
Thus
−−−→ −−−→ −−−→
A2 B1 + B2 C1 + C2 A1 = 0
and since three unit vectors with vanishing sum must be rotations of each other by 120◦ ,
it follows they must also form an equilateral triangle.

C1

B2

C2

B1

B A1 A2 C

Consequently, triangles A1 A2 B1 , B1 B2 C1 , C1 C2 A1 are congruent, as ∠A2 = ∠B2 =

∠C2 . So triangle A1 B1 C1 is equilateral and the diagonals are concurrent at the center.

3
IMO 2005 Solution Notes web.evanchen.cc, updated 26 March 2024

§1.2 IMO 2005/2, proposed by Nicholas de Bruijn (NLD)

Available online at https://aops.com/community/p281572.

Problem statement

Let a1 , a2 , . . . be a sequence of integers with infinitely many positive and negative

terms. Suppose that for every positive integer n the numbers a1 , a2 , . . . , an leave
n different remainders upon division by n. Prove that every integer occurs exactly
once in the sequence.

Obviously every integer appears at most once (otherwise take n much larger). So we will
prove every integer appears at least once.

Claim — For any i < j we have |ai − aj | < j.

Proof. Otherwise, let n = |ai − aj | 6= 0. Then i, j ∈ [1, n] and ai ≡ aj (mod n), contra-
diction.

Claim — For any n, the set {a1 , . . . , an } is of the form {k + 1, . . . , k + n} for some
integer k.

Proof. By induction, with the base case n = 1 being vacuous. For the inductive step,
suppose {a1 , . . . , an } = {k + 1, . . . , k + n} are determined. Then

an+1 ≡ k (mod n + 1).

Moreover by the earlier claim we have

|an+1 − a1 | < n + 1.

From this we deduce an+1 ∈ {k, k + n + 1} as desired.

This gives us actually a complete description of all possible sequences satisfying

the hypothesis: choose any value of a1 to start. Then, for the nth term, the set
S = {a1 , . . . , an−1 } is (in some order) a set of n − 1 consecutive integers. We then let
an = max S + 1 or an = min S − 1. A picture of six possible starting terms is shown
below.

8
7
6
5
4
3

a1 a2 a3 a4 a5 a6

4
IMO 2005 Solution Notes web.evanchen.cc, updated 26 March 2024

Finally, we observe that the condition that the sequence has infinitely many positive
and negative terms (which we have not used until now) implies it is unbounded above
and below. Thus it must contain every integer.

5
IMO 2005 Solution Notes web.evanchen.cc, updated 26 March 2024

§1.3 IMO 2005/3, proposed by Hojoo Lee (KOR)

Available online at https://aops.com/community/p281573.

Problem statement

Let x, y, z > 0 satisfy xyz ≥ 1. Prove that

x5 − x2 y5 − y2 z5 − z2
+ + ≥ 0.
x5 + y 2 + z 2 x2 + y 5 + z 2 x2 + y 2 + z 5

Negating both sides and adding 3 eliminates the minus signs:

X 1 3
≤ 2 .
cyc
x5 2
+y +z 2 x + y2 + z2

Thus we only need to consider the case xyz = 1.

Direct expansion and Muirhead works now! As advertised, once we show it suffices to
analyze if xyz = 1 the inequality becomes more economically written as
X ?
S= x2 (x2 − yz)(y 4 + x3 z + xz 3 )(z 4 + x3 y + xy 3 ) ≥ 0.
cyc

So, clearing all the denominators gives

X
x2 (x2 − yz) y 4 z 4 + x3 y 5 + xy 7 + x3 z 5 + x6 yz + x4 y 3 z + xz 7 + x4 yz 3 + x2 y 3 z 3
 
S=
cyc
X
x4 y 4 z 4 + x7 y 5 + x5 y 7 + x7 z 5 + x10 yz + x8 y 3 z + x5 z 7 + x8 yz 3 + x6 y 3 z 3

=
cyc
X
x2 y 5 z 5 + x5 y 6 z + x3 y 8 z + x5 yz 6 + x8 y 2 z 2 + x6 y 4 z 2 + x3 yz 8 + x6 y 2 z 4 + x4 y 4 z 4

−
cyc
X
x7 y 5 + x5 y 7 + x7 z 5 + x10 yz + x5 z 7 + x6 y 3 z 3

=
cyc
X
x2 y 5 z 5 + x5 y 6 z + x5 yz 6 + x8 y 2 z 2 + x6 y 4 z 2 + x6 y 2 z 4

−
cyc

In other words we need to show

X 1 1
 X
1 8 2 2 1 5 5 2

2x7 y 5 + x10 yz + x6 y 3 z 3 ≥ x y z + x y z + x6 y 4 z 2 + x6 y 5 z .
sym
2 2 sym
2 2

which follows by summing

X x10 yz + x6 y 3 z 3 X
≥ x8 y 2 z 2
sym
2 sym
1X 1X 6 4 2
x8 y 2 z 2 ≥ x y z
2 sym 2 sym
1X 7 5 1X 5 5 2
x y ≥ x y z
2 sym 2 sym

6
IMO 2005 Solution Notes web.evanchen.cc, updated 26 March 2024

1X 7 5 1X 6 4 2
x y ≥ x y z
2 sym 2 sym
X X
x7 y 5 ≥ x6 y 5 z.
sym sym

The first line here comes from AM-GM, the rest come from Muirhead.

Remark. More elegant approach is to use Cauchy in the form

1 x−1 + y 2 + z 2
≤ .
x5 + y 2 + z 2 (x2 + y 2 + z 2 )2

7
IMO 2005 Solution Notes web.evanchen.cc, updated 26 March 2024

§2 Solutions to Day 2
§2.1 IMO 2005/4, proposed by Mariusz Skalba (POL)
Available online at https://aops.com/community/p282138.

Problem statement

Determine all positive integers relatively prime to all the terms of the infinite sequence

an = 2n + 3n + 6n − 1, n ≥ 1.

The answer is 1 only (which works).

It suffices to show there are no primes. For the primes p = 2 and p = 3, take a2 = 48.
For any prime p ≥ 5 notice that

ap−2 = 2p−2 + 3p−2 + 6p−2 − 1

1 1 1
≡ + + − 1 (mod p)
2 3 6
≡ 0 (mod p)

so no other larger prime works.

8
IMO 2005 Solution Notes web.evanchen.cc, updated 26 March 2024

§2.2 IMO 2005/5, proposed by Waldemar Pompe (POL)

Available online at https://aops.com/community/p282140.

Problem statement

Let ABCD be a fixed convex quadrilateral with BC = DA and BC ∦ DA. Let

two variable points E and F lie on the sides BC and DA, respectively, and satisfy
BE = DF . The lines AC and BD meet at P , the lines BD and EF meet at Q, the
lines EF and AC meet at R. Prove that the circumcircles of the triangles P QR, as
E and F vary, have a common point other than P .

Let M be the Miquel point of complete quadrilateral ADBC; in other words, let M
be the second intersection point of the circumcircles of 4AP D and 4BP C. (A good
diagram should betray this secret; all the points are given in the picture.) This makes
lots of sense since we know E and F will be sent to each other under the spiral similarity
too.

B
A
E
P
R

Q
M
F

D C

Thus M is the Miquel point of complete quadrilateral F ACE. As R = F E ∩ AC we

deduce F ARM is a cyclic quadrilateral (among many others, but we’ll only need one).
Now look at complete quadrilateral AF QP . Since M lies on (DF Q) and (RAF ), it
follows that M is in fact the Miquel point of AF QP as well. So M lies on (P QR).
Thus M is the fixed point that we wanted.

Remark. Naturally, the congruent length condition can be relaxed to DF /DA = BE/BC.

9
IMO 2005 Solution Notes web.evanchen.cc, updated 26 March 2024

§2.3 IMO 2005/6, proposed by Radu Gologan, Dan Schwartz (ROU)

Available online at https://aops.com/community/p282141.

Problem statement

In a mathematical competition 6 problems were posed to the contestants. Each pair

of problems was solved by more than 25 of the contestants. Nobody solved all 6
problems. Show that there were at least 2 contestants who each solved exactly 5
problems.

Assume not and at most one contestant solved five problems. By adding in solves, we can
assume WLOG that one contestant solved problems one through five, and every other
contestant solved four of the six problems.
We split the remaining contestants based on whether they solved P6. Let ai denote
the number of contestants who solved {1, 2, . . . , 5} \ {i} (and missed P6). Let bij denote
the number of contestants who solved {1, 2, . . . , 5, 6} \ {i, j}, for 1 ≤ i < j ≤ 5 (thus in
particular they solved P6). Thus
X X
n=1+ ai + bij
1≤i≤5 1≤i<j≤5

denotes the total number of contestants.

Considering contestants who solved P1/P6 we have
2 1
t1 := b23 + b24 + b25 + b34 + b35 + b45 ≥ n +
5 5
and we similarly define t2 , t3 , t4 , t5 . (We have written 25 n + 15 since we know the left-hand
side is an integer strictly larger than 25 n.) Also, by considering contestants who solved
P1/P2 we have
2 1
t12 = 1 + a3 + a4 + a5 + b34 + b35 + b45 ≥ n +
5 5
and we similarly define tij for 1 ≤ i < j ≤ 5.

Claim — The number 2n+1 5 is equal to some integer k, fourteen of the t’s are equal
to k, and the last one is equal to k + 1.

Proof. First, summing all fifteen equations gives

X X
6n + 4 = 10 + 6(n − 1) = 10 + 6ai + 6bij
1≤i≤5 1≤i<j≤5
X X
= ti + tij .
1≤i≤5 1≤i<j≤5

Thus the sum of the 15 t’s is 6n + 4. But since all the t’s are integers at least 2n+1
5 = 15 ,
6n+3

the conclusion follows.

However, we will also manipulate the equations to get the following.

10
IMO 2005 Solution Notes web.evanchen.cc, updated 26 March 2024

Claim — We have

t45 ≡ 1 + t1 + t2 + t3 + t12 + t23 + t31 (mod 3).

Proof. This follows directly by computing the coefficient of the a’s and b’s. We will
nonetheless write out a derivation of this equation, to motivate it, but the proof stands
without it. P
Let B = 1≤i<j≤5 bij be the sum of all b’s. First, note that

t1 + t2 = B + b34 + b45 + b35 − b12

= B + (t12 − 1 − a3 − a4 − a5 ) − b12
=⇒ b12 = B − (t1 + t2 ) + t12 − 1 − (a3 + a4 + a5 ).
This means we have more or less solved for each bij in terms of only t and a variables.
Now
t45 = 1 + a1 + a2 + a3 + b12 + b23 + b31
= 1 + a1 + a2 + a3
+ [B − (t1 + t2 ) + t12 − 1 − (a3 + a4 + a5 )]
+ [B − (t2 + t3 ) + t23 − 1 − (a1 + a4 + a5 )]
+ [B − (t3 + t1 ) + t13 − 1 − (a2 + a4 + a5 )]
≡ 1 + t1 + t2 + t3 + t12 + t23 + t31 (mod 3)
as desired.

However, we now show the two claims are incompatible (and this is easy, many ways
to do this). There are two cases.
• Say t5 = k + 1 and the others are k. Then the equation for t45 gives that k ≡ 6k + 1
(mod 3). But now the equation for t12 give k ≡ 6k (mod 3).
• Say t45 = k + 1 and the others are k. Then the equation for t45 gives that k + 1 ≡ 6k
(mod 3). But now the equation for t12 give k ≡ 6k + 1 (mod 3).

Remark. It is significantly easier to prove that there is at least one contestant who solved
five problems. One can see it by dropping the +10 in the proof of the claim, and arrives
at a contradiction. In this situation it is not even necessary to set up the many a and b
variables; just note that the expected number of contestants solving any particular pair of
(4)n
problems is 26 = 25 n.
(2)
The fact that 2n+1
5 should be an integer also follows quickly, since if not one can improve
the bound to 2n+2 5 and quickly run into a contradiction. Again one can get here without
setting up a and b.
The main difficulty seems to be the precision required in order to nail down the second
5-problem solve.

Remark. The second claim may look miraculous, but the proof shows that it is not too
unnatural to consider t1 + t2 − t12 to isolate b12 in terms of a’s and t’s. The main trick is:
why mod 3?
The reason is that if one looks closely, for a fixed k we have a system of 15 equations in
15 variables. Unless the determinant D of that system happens to be zero, this means there
will be a rational solution in a and b, whose denominators are bounded by D. However if

11
IMO 2005 Solution Notes web.evanchen.cc, updated 26 March 2024

p | D then we may conceivably run into mod p issues.

This motivates the choice p = 3, since it is easy to see the determinant is divisible by 3,
since constant shifts of ~a and ~b are also solutions mod 3. (The choice p = 2 is a possible
guess as well for this reason, but the problem seems to have better 3-symmetry.)

12