IMO 2014 Solution Notes

Evan Chen《陳誼廷》
15 April 2024

This is a compilation of solutions for the 2014 IMO. The ideas of the
solution are a mix of my own work, the solutions provided by the competition
organizers, and solutions found by the community. However, all the writing
is maintained by me.
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 2014/1, proposed by Gerhard Woeginger (AUT) . . . . . . . . . . . 3
1.2 IMO 2014/2, proposed by Tonči Kokan (HRV) . . . . . . . . . . . . . . . 4
1.3 IMO 2014/3, proposed by ALi Zamani (IRN) . . . . . . . . . . . . . . . . 6

2 Solutions to Day 2 9
2.1 IMO 2014/4, proposed by Giorgi Arabidze (GEO) . . . . . . . . . . . . . 9
2.2 IMO 2014/5, proposed by Gerhard Woeginger (LUX) . . . . . . . . . . . . 11
2.3 IMO 2014/6, proposed by Gerhard Woeginger (AUT) . . . . . . . . . . . 12

1
IMO 2014 Solution Notes [Link], updated 15 April 2024

§0 Problems
1. Let a0 < a1 < a2 < · · · be an infinite sequence of positive integers. Prove that
there exists a unique integer n ≥ 1 such that
a0 + a1 + a2 + · · · + an
an < ≤ an+1 .
n

2. Let n ≥ 2 be an integer. Consider an n × n chessboard consisting of n2 unit squares.

A configuration of n rooks on this board is peaceful if every row and every column
contains exactly one rook. Find the greatest positive integer k such that, for each
peaceful configuration of n rooks, there is a k × k square which does not contain a
rook on any of its k 2 unit squares.

3. Convex quadrilateral ABCD has ∠ABC = ∠CDA = 90◦ . Point H is the foot
of the perpendicular from A to BD. Points S and T lie on sides AB and AD,
respectively, such that H lies inside triangle SCT and

∠CHS − ∠CSB = 90◦ , ∠T HC − ∠DT C = 90◦ .

Prove that line BD is tangent to the circumcircle of triangle T SH.

4. Let P and Q be on segment BC of an acute triangle ABC such that ∠P AB =

∠BCA and ∠CAQ = ∠ABC. Let M and N be points on AP and AQ, respectively,
such that P is the midpoint of AM and Q is the midpoint of AN . Prove that BM
and CN meet on the circumcircle of 4ABC.

5. For every positive integer n, the Bank of Cape Town issues coins of denomination n1 .
Given a finite collection of such coins (of not necessarily different denominations)
with total value at most 99 + 12 , prove that it is possible to split this collection into
100 or fewer groups, such that each group has total value at most 1.

6. A set of lines in the plane is in general position if no two are parallel and no three
pass through the same point. A set of lines in general position cuts the plane into
regions, some of which have finite area; we call these its finite regions. Prove that
for all sufficiently large n, in any set of n lines in general position it is possible to
√
colour at least n lines blue in such a way that none of its finite regions has a
completely blue boundary.

2
IMO 2014 Solution Notes [Link], updated 15 April 2024

§1 Solutions to Day 1
§1.1 IMO 2014/1, proposed by Gerhard Woeginger (AUT)
Available online at [Link]

Problem statement

Let a0 < a1 < a2 < · · · be an infinite sequence of positive integers. Prove that there
exists a unique integer n ≥ 1 such that
a0 + a1 + a2 + · · · + an
an < ≤ an+1 .
n

Fedor Petrov presents the following nice solution. Let us define the sequence

bn = (an − an−1 ) + · · · + (an − a1 ) .

Since (ai )i is increasing, this sequence is unbounded, and moreover b1 = 0. The problem
requires an n such that
bn < a0 ≤ bn+1
which obviously exists and is unique.

3
IMO 2014 Solution Notes [Link], updated 15 April 2024

§1.2 IMO 2014/2, proposed by Tonči Kokan (HRV)

Available online at [Link]

Problem statement

Let n ≥ 2 be an integer. Consider an n × n chessboard consisting of n2 unit squares.

A configuration of n rooks on this board is peaceful if every row and every column
contains exactly one rook. Find the greatest positive integer k such that, for each
peaceful configuration of n rooks, there is a k × k square which does not contain a
rook on any of its k 2 unit squares.

The answer is k = n − 1 , sir.

√ 

First, assume n > k 2 for some k. We will prove we can find an empty k × k square.
Indeed, let R be a rook in the uppermost column, and draw k squares of size k × k
directly below it, aligned. There are at most k − 1 rooks among these squares, as desired.

Now for the construction for n = k 2 . We draw the example for k = 3 (with the
generalization being obvious);

r
r
r
r
r
r
r
r
r
To show that this works, consider for each rook drawing an k × k square of X’s whose
bottom-right hand corner is the rook (these may go off the board). These indicate
positions where one cannot place the upper-left hand corner of any square. It’s easy to
see that these cover the entire board, except parts of the last k − 1 columns, which don’t
matter anyways.

4
IMO 2014 Solution Notes [Link], updated 15 April 2024

It remains to check that n ≤ k 2 also all work (omitting this step is a common mistake).
For this, we can delete rows and column to get an n × n board, and then fill in any gaps
where we accidentally deleted a rook.

5
IMO 2014 Solution Notes [Link], updated 15 April 2024

§1.3 IMO 2014/3, proposed by ALi Zamani (IRN)

Available online at [Link]

Problem statement

Convex quadrilateral ABCD has ∠ABC = ∠CDA = 90◦ . Point H is the foot of the
perpendicular from A to BD. Points S and T lie on sides AB and AD, respectively,
such that H lies inside triangle SCT and

∠CHS − ∠CSB = 90◦ , ∠T HC − ∠DT C = 90◦ .

Prove that line BD is tangent to the circumcircle of triangle T SH.

¶ First solution (mine). First we rewrite the angle condition in a suitable way.

Claim — We have ∠AT H = ∠T CH + 90◦ . Thus the circumcenter of 4CT H lies

on AD. Similarly the circumcenter of 4CSH lies on AB.

Proof.

]AT H = ]DT H
= ]DT C + ]CT H
= ]DT C − ]T HC − ]HCT
= 90◦ − ]HCT = 90◦ + ]T CH.

which implies conclusion.

A
T

P
OD
S

B H D

Let the perpendicular bisector of T H meet AH at P now. It suffices to show that PAP
H
is symmetric in b = AD and d = AB, because then P will be the circumcenter of 4T SH.
To do this, set AH = 2Rbd
and AC = 2R.
Let O denote the circumcenter of 4CHT . Use the Law of Cosines on 4ACO and
4AHO, using variables x = AO and r = HO. We get that
d
r2 = x2 + AH 2 − 2x · AH · = x2 + (2R)2 − 2bx.
2R

6
IMO 2014 Solution Notes [Link], updated 15 April 2024

By the angle bisector theorem, PAP

H = HO .
AO

The rest is computation: notice that

d
r2 − x2 = h2 − 2xh · = (2R)2 − 2bx
2R
where h = AH = 2R ,
bd
whence
(2R)2 − h2
x= d
.
2b − 2h · 2R
Moreover,
1 r2
   
1 2 2
− 1 = R − b .
2 x2 x x
Now, if we plug in the x in the right-hand side of the above, we obtain
!
d d
2b − 2h · 2R 2b − 2h · 2R
 
2 2h d
−2hdR + bh2 .


2 2 2 2
· 2R − b = 2 2 2
b−h·
4R − h 4R − h (4R − h ) 2R

Pulling out a factor of −2Rh from the rightmost term, we get something that is symmetric
in b and d, as required.

¶ Second solution (Victor Reis). Here is the fabled solution using inversion at H.
First, we rephrase the angle conditions in the following ways:
• AD ⊥ (T HC), which is equivalent to the claim from the first solution.
• AB ⊥ (SHC), by symmetry.
• AC ⊥ (ABCD), by definition.
Now for concreteness we will use a negative inversion at H which swaps B and D and
overlay it on the original diagram. As usual we denote inverses with stars.
Let us describe the inverted problem. We let M and N denote the midpoints of A∗ B ∗
and A∗ D∗ , which are the centers of (HA∗ B ∗ ) and (HA∗ D∗ ). From T ∗ C ∗ ⊥ (HA∗ D∗ ),
we know have C ∗ , M , T ∗ collinear. Similarly, C ∗ , N , S ∗ are collinear. We have that
(A∗ HC ∗ ) is orthogonal to (ABCD) which remains fixed. We wish to show T ∗ S ∗ and
M N are parallel.

C∗
T

B H D

M N

A∗ C
T ∗ S∗

7
IMO 2014 Solution Notes [Link], updated 15 April 2024

Lot ω denote the circumcircle of 4A∗ HC ∗ , which is orthogonal to the original circle
(ABCD). It would suffices to show (A∗ HC ∗ ) is an H-Apollonius circle with respect to
M N , from which we would get C ∗ M /HM = C ∗ N /HN .
However, ω through H and A, hence it center lies on line M N . Moreover ω is orthogonal
to (A∗ M N ) (since (A∗ M N ) and (A∗ BD) are homothetic). This is enough (for example,
if we let O denote the center of ω, we now have r(ω)2 = OH 2 = OM · ON ). (Note in
this proof that the fact that C ∗ lies on (ABCD) is not relevant.)

8
IMO 2014 Solution Notes [Link], updated 15 April 2024

§2 Solutions to Day 2
§2.1 IMO 2014/4, proposed by Giorgi Arabidze (GEO)
Available online at [Link]

Problem statement

Let P and Q be on segment BC of an acute triangle ABC such that ∠P AB = ∠BCA

and ∠CAQ = ∠ABC. Let M and N be points on AP and AQ, respectively, such
that P is the midpoint of AM and Q is the midpoint of AN . Prove that BM and
CN meet on the circumcircle of 4ABC.

We give three solutions.

¶ First solution by harmonic bundles. Let BM intersect the circumcircle again at X.

B Q P C

N M

The angle conditions imply that the tangent to (ABC) at B is parallel to AP . Let ∞
be the point at infinity along line AP . Then
B
−1 = (AM ; P ∞) = (AX; BC).

Similarly, if CN meets the circumcircle at Y then (AY ; BC) = −1 as well. Hence X = Y ,

which implies the problem.

¶ Second solution by similar triangles. Once one observes 4CAQ ∼ 4CBA, one can
construct D the reflection of B across A, so that 4CAN ∼ 4CBD. Similarly, letting E
be the reflection of C across A, we get 4BAP ∼ 4BCA =⇒ 4BAM ∼ 4BCE. Now
to show ∠ABM + ∠ACN = 180◦ it suffices to show ∠EBC + ∠BCD = 180◦ , which
follows since BCDE is a parallelogram.

9
IMO 2014 Solution Notes [Link], updated 15 April 2024

¶ Third solution by barycentric coordinates. Since P B = c2 /a we have

P = (0 : a2 − c2 : c2 )

so the reflection M ~ has coordinates

~ = 2P~ − A

M = (−a2 : 2(a2 − c2 ) : 2c2 ).

Similarly N = (−a2 : 2b2 : 2(b2 − a2 )). Thus

BM ∩ CN = (−a2 : 2b2 : 2c2 )

which clearly lies on the circumcircle, and is in fact the point identified in the first
solution.

10
IMO 2014 Solution Notes [Link], updated 15 April 2024

§2.2 IMO 2014/5, proposed by Gerhard Woeginger (LUX)

Available online at [Link]

Problem statement

For every positive integer n, the Bank of Cape Town issues coins of denomination n1 .
Given a finite collection of such coins (of not necessarily different denominations)
with total value at most 99 + 12 , prove that it is possible to split this collection into
100 or fewer groups, such that each group has total value at most 1.

We’ll prove the result for at most k − k

2k+1 with k groups. First, perform the following
optimizations.

• If any coin of size 1

2m appears twice, then replace it with a single coin of size m.
1

• If any coin of size 2m+1

1
appears 2m + 1 times, group it into a single group and
induct downwards.

Apply this operation repeatedly until it cannot be done anymore.

Now construct boxes B0 , B1 , . . . , Bk−1 . In box B0 put any coins of size 12 (clearly
there is at most one). In the other boxes Bm , put coins of size 2m+1
1
and 2m+2
1
(at most
2m of the former and at most one of the latter). Note that the total weight in the box is
less than 1. Finally, place the remaining “light” coins of size at most 2k+1
1
in a pile.
Then just toss coins from the pile into the boxes arbitrarily, other than the proviso
that no box should have its weight exceed 1. We claim this uses up all coins in the pile.
Assume not, and that some coin remains in the pile when all the boxes are saturated.
Then all the boxes must have at least 1 − 2k+1 1
, meaning the total amount in the boxes
is strictly greater than  
1
k 1− > k − 12
2k + 1
which is a contradiction.

Remark. This gets a stronger bound k − k

2k+1 than the requested k − 12 .

11
IMO 2014 Solution Notes [Link], updated 15 April 2024

§2.3 IMO 2014/6, proposed by Gerhard Woeginger (AUT)

Available online at [Link]

Problem statement

A set of lines in the plane is in general position if no two are parallel and no three
pass through the same point. A set of lines in general position cuts the plane into
regions, some of which have finite area; we call these its finite regions. Prove that for
all sufficiently large n, in any set of n lines in general position it is possible to colour
√
at least n lines blue in such a way that none of its finite regions has a completely
blue boundary.

Suppose we have colored k of the lines blue, and that it is not possible to color any
additional lines. That means any of the n − k non-blue lines is the side of some finite
region with an otherwise entirely blue perimeter. For each such line `, select one such
region, and take the next counterclockwise vertex; this is the intersection of two blue
lines v. We’ll say ` is the eyelid of v.

You can prove without too much difficulty that every intersection of two blue lines has
at most two eyelids. Since there are 2 such intersections, we see that
k

 
k
n−k ≤2 = k2 − k
2

so n ≤ k 2 , as required.
√
Remark. In fact, k = n is “sharp for greedy algorithms”, as illustrated below for k = 3:

12