Final round

Dutch Mathematical Olympiad

Friday 15 September 2017

Solutions
1. Version for klas 5 & klas 4 and below
An integer for which the digits (from left to right) are c1 , c2 , . . . , ck will be denoted by c1 c2 . . . ck .

(a) Let n = c1 c2 . . . ck be an even-steven integer greater than 9 (hence k > 2). We will prove
that n is indeed the sum of two oddball integers.
Let p = c2 . . . ck and q = c1 0 . . . 0 (with k − 1 digits equal to 0). As n is even-steven, we
have c2 > c1 > 1, hence p and q do not start with the digit 0. Moreover, it is clear that
n = p + q. Hence, it suffices to prove that p and q are oddball.
The integer q is oddball, because all digits except for the first are equal to 0. The digits of
p are the same digits as the digits of n, except for the first digit c1 . However, the digits
that were at an even position in n are at an odd position in p and vice versa. Therefore,
also p is oddball.
(b) The integer n = 109 is oddball, but it is not the sum of two even-steven integers. We will
prove this by contradiction. Suppose that n = p + q for some even-steven integers p and q.
We will show that this leads to a contradiction.
First observe that the integers 100 to 108 are not even-steven. Hence, both p and q must
be smaller than 100, and hence both are also greater than 9. In other words, p and q have
exactly two digits. Suppose p = ab and q = cd. The equation p + q = 109 now yields
b + d = 9 (because b + d < 19) and a + c = 10. Hence, b + d < a + c, which implies that
either b < a or d < c (or both). In the first case p is not even-steven and in the second case
q is not even-steven. This contradicts the assumption that p and q are even-steven.
We conclude that the oddball integer 109 cannot be written as the sum of two even-steven
integers.

1. Version for klas 6

An integer for which the digits (from left to right) are c1 , c2 , . . . , ck will be denoted by c1 c2 . . . ck .

(a) Let n = c1 c2 . . . ck be an oddball integer greater than 9 (hence k > 2). We will show that n
is indeed the sum of two oddball integers.
If c2 > 1, then we can write n as the sum of the following two oddball integers: 10 . . . 0
(k − 2 zeros) and c1 (c2 −1)c3 . . . ck .
If c2 = 0 and c1 > 2, then we can write n as the sum of the following two oddball integers:
10 . . . 0 (k − 1 zeros) and (c1 −1)c2 c3 . . . ck .
If n = 10 . . . 0, then we can write n as the sum of the following two oddball integers: 1 and
9 . . . 9 (k − 1 nines).
The last case is the case in which c1 = 1, c2 = 0 and not all digits c3 , . . . , ck are equal to 0.
Let ct be a digit unequal to 0, with t > 3 as small as possible. Hence, n = 10 . . . 0ct . . . ck
with ct > 1.
Because n is oddball and ct−1 = 0 < 1 6 ct , we find that t must be odd. We can now write
n as the sum of the integers 10 . . . 0 (k − 1 zeros) and m = ct ct+1 . . . ck . Because t is odd,
the digits at the odd positions of m are also at odd positions of n. Therefore, these digits
are greater than or equal to their neighbouring digits (because n is oddball), which yields
that m is oddball.
 (b) The integer n = 109 is oddball, but it is not the sum of two even-steven integers. We will
prove this by contradiction. Suppose that n = p + q for some even-steven integers p and q.
We will show that this leads to a contradiction.
First observe that the integers 100 to 108 are not even-steven. Hence, both p and q must
be smaller than 100, and hence both are also greater than 9. In other words, p and q have
exactly two digits. Suppose p = ab and q = cd. The equation p + q = 109 now yields
b + d = 9 (because b + d < 19) and a + c = 10. Hence, b + d < a + c, which implies that
either b < a or d < c (or both). In the first case p is not even-steven and in the second case
q is not even-steven. This contradicts the assumption that p and q are even-steven.
We conclude that the oddball integer 109 cannot be written as the sum of two even-steven
integers.

2. Version for klas 4 & below

Since AED is an isosceles triangle, angles ∠EDA and ∠DAE D C
are equal. In turn, these angles are equal to angles ∠EBF and
∠BF E (alternate interior angles). This implies that triangle
BF E is isosceles as well, with |BE| = |EF |.
Comparing triangles ABE and DF E, we see that |AE| = |DE|
F
and |BE| = |F E|. Since ∠BEA and ∠F ED are a pair of
opposite angles, they have the same size. It follows that triangles E
ABE and DF E are congruent (SAS).
From the fact that ABE and DF E are congruent it follows that A B
|DF | = |AB|. Since ABCD is a parallelogram, we also have
|AB| = |CD|. It follows that triangle CDF is isosceles as well
(with apex D).
Triangle ADB is also isosceles. Since ∠F CD = ∠BAD (because ABCD is a parallelogram), we
deduce that triangles CDF and ADB are similar. In particular, ∠CDF = ∠ADB holds.

2. Version for klas 5 & klas 6

Since AED is an isosceles triangle, angles ∠EDA and ∠DAE D C
are equal. In turn, these angles are equal to angles ∠EBF and
∠BF E (alternate interior angles). This implies that triangle
BF E is isosceles as well, with |BE| = |EF |.
Comparing triangles ABE and DF E, we see that |AE| = |DE|
F
and |BE| = |F E|. Since ∠BEA and ∠F ED are a pair of
opposite angles, they have the same size. It follows that triangles E
ABE and DF E are congruent (SAS).
From the fact that ABE and DF E are congruent it follows that A B
|DF | = |AB|. Since ABCD is a parallelogram, we also have
|AB| = |CD|. It follows that triangle CDF is isosceles as well
(with apex D).
On the one hand, this implies that ∠F CD = ∠DF C. On the other hand, we know that triangle
DBC is isosceles (since |BD| = |AD| = |BC|), which implies that ∠F CD = ∠CDB = 2·∠CDF
since DF is the angle bisector of ∠CDB.
Altogether, we have ∠DF C = ∠F CD = 2 · ∠CDF . Since the angles in any triangle sum to 180
degrees, we also know that

180◦ = ∠DF C + ∠F CD + ∠CDF = 5 · ∠CDF.

From this, it follows that ∠CDF = 15 · 180◦ = 36◦ , and hence ∠F CD = 2 · ∠CDF = 72◦ . Since
triangle DBC is isosceles, also ∠CDB = 72◦ holds. Using alternating interior angles, we now
find that ∠ABD = ∠CDB = 72◦ .
3. Let the scores of the six teams be s, s + 2, s + 4, s + 6, s + 8, and s + 10. Let T be the total
number of awarded points, so that T = 6s + 30. It follows that the total number of points is a
multiple of six.
The number of games played equals 6·5
2 = 15. Let g be the number of games that ended in a
draw. A game that ends in a draw results in 1 + 1 = 2 awarded points and every other game
results in 3 + 0 = 3 awarded points. Therefore, the total number of awarded points equals
T = g · 2 + (15 − g) · 3 = 45 − g.
From T = 45 − g it follows that 30 6 T 6 45 because the number of draws satisfies 0 6 g 6 15.
Since T is a multiple of six, this leaves the following possibilities: T = 30, T = 36, and T = 42.
If T = 30, we have g = 45 − 30 = 15. But then all games must have ended in a draw and all
teams must have the same score. Hence, the case T = 30 is ruled out.
If T = 36, then g = 45 − 36 = 9 and s = T −30 6 = 1. The six scores are therefore 1, 3, 5, 7, 9, 11.
The team that scored 1 point must have lost 4 games (and played one draw). The team
that scored 3 points must have lost at least 2 games (at most 3 games were not lost). The
team that scored 11 points must have won at least 3 games (otherwise the score is at most
3 + 3 + 1 + 1 + 1 = 9), so apart from the teams with scores 1 and 3 at least one other team has
lost a game. In total, at least 4 + 2 + 1 = 7 games ended in a loss for some team, contradicting
the fact that 15 − 9 = 6 games did not end in a draw. This rules out the case T = 36.
Finally, we consider that case T = 42. The six scores are 2, 4, 6, 8, 10, 12 and we have g = 3.
Since the total number of points obtained from won games is a multiple of three, the six teams
must have received at least 2, 1, 0, 2, 1, 0 points from draws, respectively. In total, exactly 2 · 3 = 6
points are awarded in games that ended in a draw. Hence, since 2 + 1 + 0 + 2 + 1 + 0 = 6, the
six teams have received exactly the mentioned numbers of points from draws. In particular, the
team ending in the fourth place (with 6 points), was involved in 0 draws and must have won
exactly two games.

4. Version for klas 5 & klas 4 and below

(a) Let r be the remainder upon dividing n by a. We will first prove that r < n2 . If 2a 6 n,
this follows from the fact that r < a. If 2a > n, we have r = n − a (since we already knew
that a < n), which implies that r = n − a < n − n2 = n2 .
For the same reasons, the remainder upon dividing n by b is smaller than n2 .
It follows that the two remainders obtained by dividing n by a and b add up to a number
smaller than n.
(b) Let r, s, and t be the remainders upon dividing n by 99, 132, and 229, respectively. The
number n − t is a multiple of 229 and nonzero since n > 229 > t. We know that r + s + t = n,
and hence n − t = r + s. We can conclude that r + s is a positive multiple of 229. Since
99 + 132 < 2 · 229, we have r + s < 2 · 229, which implies that we must have r + s = 229.
(c) Since r 6 98 and s 6 131, the fact that r + s = 229 implies that r = 98 and s = 131.
Therefore, the number n + 1 is divisible by both 99 and 132, and hence by their least
common multiple lcm(99, 132) = lcm(9 · 11, 3 · 4 · 11) = 4 · 9 · 11 = 396. Also, from n = 229 + t
and t < 229 we deduce that n + 1 6 458. It follows that the only possibility is n + 1 = 396,
hence n = 395.
When n = 395 the three remainders are r = 98, s = 131, and t = 166, and indeed satisfy
the equation n = r + s + t.
4. Version for klas 6
(a) Let r be the remainder upon dividing n by a. We will first prove that r < n2 . If 2a 6 n,
this follows from the fact that r < a. If 2a > n, we have r = n − a (since we already knew
that a < n), which implies that r = n − a < n − n2 = n2 .
For the same reasons, the remainder upon dividing n by b is smaller than n2 .
It follows that the two remainders obtained by dividing n by a and b add up to a number
smaller than n.
(b) Let r, s, and t be the remainders upon dividing n by 99, 132, and 229, respectively. The
number n − t is a multiple of 229 and nonzero since n > 229 > t. We know that r + s + t = n,
and hence n − t = r + s. We can conclude that r + s is a positive multiple of 229. Since
99 + 132 < 2 · 229, we have r + s < 2 · 229, which implies that we must have r + s = 229.
Since r 6 98 and s 6 131, the fact that r + s = 229 implies that r = 98 and s = 131.
Therefore, the number n + 1 is divisible by both 99 and 132, and hence by their least
common multiple lcm(99, 132) = lcm(9 · 11, 3 · 4 · 11) = 4 · 9 · 11 = 396. Also, from n = 229 + t
and t < 229 we deduce that n + 1 6 458. It follows that the only possibility is n + 1 = 396,
hence n = 395.
When n = 395 the three remainders are r = 98, s = 131, and t = 166, and indeed satisfy
the equation n = r + s + t.

5. Four of the eight points are coloured black and the other four points are coloured white in the
way indicated in the figure on the left. The circle through the four black points is denoted C1
and the circle through the four white points is denoted C2 . If two points lie on a circle C, we
say that C covers that pair of points.

Circle C1 covers all pairs of black points and circle C2 covers all pairs of white points. It is easy
to check that each of the 4 · 4 = 16 pairs consisting of a white point and a black point is covered
by one of the four circles in the figure on the right. It follows that the six circles form a solution.
We will now prove that five or fewer circles do not suffice. First observe that any circle passing
through more than two black points must be equal to C1 and that any circle passing through
more than two white points must be equal to C2 . Indeed, a circle is already determined by three
points.
A circle passing through 2 or fewer black points covers at most one of the 4·3 2 = 6 pairs of black
points. A solution consisting of only five circles must therefore contain circle C1 (since otherwise
at most 5 pairs of black points are covered). In the same way we see that such a solution must
contain circle C2 .
Each of the remaining three circles in the (hypothetical) solution contains at most 2 black points
and at most 2 white points. Such a circle covers at most 2 · 2 = 4 pairs consisting of a white and
a black point. In total, the five circles can therefore cover at most 0 + 0 + 3 · 4 = 12 such pairs,
while there are 16 to be covered. The five circles can therefore not form a correct solution after
all. We conclude that the smallest number of circles in a solution is 6.

© 2017 Stichting Nederlandse Wiskunde Olympiade