Maths Olympiad Prep

Library / /308 of 383

Combinatorics Difficulty 8.9 Shortlist Prove it IMO

We are given an infinite deck of cards, each with a real number on it. For every real number xx, there is exactly one card in the deck that has xx written on it. Now two players draw disjoint sets AA and BB of 100100 cards each from this deck. We would like to define a rule that declares one of them a winner. This rule should satisfy the following conditions:
1. The winner only depends on the relative order of the 200200 cards: if the cards are laid down in increasing order face down and we are told which card belongs to which player, but not what numbers are written on them, we can still decide the winner.
2. If we write the elements of both sets in increasing order as A={a1,a2,,a100}A=\{a_{1}, a_{2}, \ldots, a_{100}\} and B={b1,b2,,b100}B=\{b_{1}, b_{2}, \ldots, b_{100}\}, and ai>bia_{i}>b_{i} for all ii, then AA beats BB.
3. If three players draw three disjoint sets A,B,CA, B, C from the deck, AA beats BB and BB beats CC, then AA also beats CC.
How many ways are there to define such a rule? Here, we consider two rules as different if there exist two sets AA and BB such that AA beats BB according to one rule, but BB beats AA according to the other.

Solution

Answer. 100100.

Solution 1. We prove a more general statement for sets of cardinality nn (the problem being the special case n=100n=100, then the answer is nn). In the following, we write A>BA>B or B<AB<A for "AA beats BB".

Part I. Let us first define nn different rules that satisfy the conditions. To this end, fix an index k{1,2,,n}k \in \{1,2, \ldots, n\}. We write both AA and BB in increasing order as A={a1,a2,,an}A=\{a_{1}, a_{2}, \ldots, a_{n}\} and B={b1,b2,,bn}B=\{b_{1}, b_{2}, \ldots, b_{n}\} and say that AA beats BB if and only if ak>bka_{k}>b_{k}. This rule clearly satisfies all three conditions, and the rules corresponding to different kk are all different. Thus there are at least nn different rules.

Part II. Now we have to prove that there is no other way to define such a rule. Suppose that our rule satisfies the conditions, and let k{1,2,,n}k \in \{1,2, \ldots, n\} be minimal with the property that
Ak={1,2,,k,n+k+1,n+k+2,,2n}Bk={k+1,k+2,,n+k}. A_{k}=\{1,2, \ldots, k, n+k+1, n+k+2, \ldots, 2n\} \prec B_{k}=\{k+1, k+2, \ldots, n+k\}.
Clearly, such a kk exists, since this holds for k=nk=n by assumption. Now consider two disjoint sets X={x1,x2,,xn}X=\{x_{1}, x_{2}, \ldots, x_{n}\} and Y={y1,y2,,yn}Y=\{y_{1}, y_{2}, \ldots, y_{n}\}, both in increasing order (i.e., x1<x2<<xnx_{1}<x_{2}<\cdots<x_{n} and y1<y2<<yny_{1}<y_{2}<\cdots<y_{n}). We claim that X<YX<Y if (and only if - this follows automatically) xk<ykx_{k}<y_{k}.

To prove this statement, pick arbitrary real numbers ui,vi,wiXYu_{i}, v_{i}, w_{i} \notin X \cup Y such that
u1<u2<<uk1<min(x1,y1),max(xn,yn)<vk+1<vk+2<<vn, u_{1}<u_{2}<\cdots<u_{k-1}<\min(x_{1}, y_{1}), \quad \max(x_{n}, y_{n})<v_{k+1}<v_{k+2}<\cdots<v_{n},
and
xk<v1<v2<<vk<w1<w2<<wn<uk<uk+1<<un<yk,x_{k}<v_{1}<v_{2}<\cdots<v_{k}<w_{1}<w_{2}<\cdots<w_{n}<u_{k}<u_{k+1}<\cdots<u_{n}<y_{k},
and set
U={u1,u2,,un},V={v1,v2,,vn},W={w1,w2,,wn}.U=\{u_{1}, u_{2}, \ldots, u_{n}\}, V=\{v_{1}, v_{2}, \ldots, v_{n}\}, W=\{w_{1}, w_{2}, \ldots, w_{n}\}.
Then
- ui<yiu_{i}<y_{i} and xi<vix_{i}<v_{i} for all ii, so U<YU<Y and X<VX<V by the second condition.
- The elements of UWU \cup W are ordered in the same way as those of Ak1Bk1A_{k-1} \cup B_{k-1}, and since Ak1>Bk1A_{k-1}>B_{k-1} by our choice of kk, we also have U>WU>W (if k=1k=1, this is trivial).
- The elements of VWV \cup W are ordered in the same way as those of AkBkA_{k} \cup B_{k}, and since Ak<BkA_{k}<B_{k} by our choice of kk, we also have V<WV<W.
It follows that
X<V<W<U<Y X<V<W<U<Y
so X<YX<Y by the third condition, which is what we wanted to prove.

Solution 2. Another possible approach to Part II of this problem is induction on nn. For n=1n=1, there is trivially only one rule in view of the second condition.

In the following, we assume that our claim (namely, that there are no possible rules other than those given in Part I) holds for n1n-1 in place of nn. We start with the following observation:

Claim. At least one of the two relations
({2}{2i12in})({1}{2i2in}) (\{2\} \cup \{2i-1 \mid 2 \leqslant i \leqslant n\}) \prec (\{1\} \cup \{2i \mid 2 \leqslant i \leqslant n\})
and
({2i11in1}{2n})({2i1in1}{2n1}) (\{2i-1 \mid 1 \leqslant i \leqslant n-1\} \cup \{2n\}) \prec (\{2i \mid 1 \leqslant i \leqslant n-1\} \cup \{2n-1\})
holds.

Proof. Suppose that the first relation does not hold. Since our rule may only depend on the relative order, we must also have
({2}{3i22in1}{3n2})>({1}{3i12in1}{3n}). (\{2\} \cup \{3i-2 \mid 2 \leqslant i \leqslant n-1\} \cup \{3n-2\}) > (\{1\} \cup \{3i-1 \mid 2 \leqslant i \leqslant n-1\} \cup \{3n\}).
Likewise, if the second relation does not hold, then we must also have
({1}{3i12in1}{3n})>({3}{3i2in1}{3n1}). (\{1\} \cup \{3i-1 \mid 2 \leqslant i \leqslant n-1\} \cup \{3n\}) > (\{3\} \cup \{3i \mid 2 \leqslant i \leqslant n-1\} \cup \{3n-1\}).
Now condition 3 implies that
({2}{3i22in1}{3n2})>({3}{3i2in1}{3n1}), (\{2\} \cup \{3i-2 \mid 2 \leqslant i \leqslant n-1\} \cup \{3n-2\}) > (\{3\} \cup \{3i \mid 2 \leqslant i \leqslant n-1\} \cup \{3n-1\}),
which contradicts the second condition.

Now we distinguish two cases, depending on which of the two relations actually holds:

First case: ({2}{2i12in})({1}{2i2in})(\{2\} \cup \{2i-1 \mid 2 \leqslant i \leqslant n\}) \prec (\{1\} \cup \{2i \mid 2 \leqslant i \leqslant n\}).
Let A={a1,a2,,an}A=\{a_{1}, a_{2}, \ldots, a_{n}\} and B={b1,b2,,bn}B=\{b_{1}, b_{2}, \ldots, b_{n}\} be two disjoint sets, both in increasing order. We claim that the winner can be decided only from the values of a2,,ana_{2}, \ldots, a_{n} and b2,,bnb_{2}, \ldots, b_{n}, while a1a_{1} and b1b_{1} are actually irrelevant. Suppose that this was not the case, and assume without loss of generality that a2<b2a_{2}<b_{2}. Then the relative order of a1,a2,,an,b2,,bna_{1}, a_{2}, \ldots, a_{n}, b_{2}, \ldots, b_{n} is fixed, and the position of b1b_{1} has to decide the winner. Suppose that for some value b1=xb_{1}=x, BB wins, while for some other value b1=yb_{1}=y, AA wins.

Write Bx={x,b2,,bn}B_{x}=\{x, b_{2}, \ldots, b_{n}\} and By={y,b2,,bn}B_{y}=\{y, b_{2}, \ldots, b_{n}\}, and let ε>0\varepsilon>0 be smaller than half the distance between any two of the numbers in BxByAB_{x} \cup B_{y} \cup A. For any set MM, let M±εM \pm \varepsilon be the set obtained by adding/subtracting ε\varepsilon to all elements of MM. By our choice of ε\varepsilon, the relative order of the elements of (By+ε)A(B_{y}+\varepsilon) \cup A is still the same as for ByAB_{y} \cup A, while the relative order of the elements of (Bxε)A(B_{x}-\varepsilon) \cup A is still the same as for BxAB_{x} \cup A. Thus ABxεA \prec B_{x}-\varepsilon, but ABy+εA \succ B_{y}+\varepsilon. Moreover, if y>xy>x, then Bxε<By+εB_{x}-\varepsilon < B_{y}+\varepsilon by condition 2, while otherwise the relative order of the elements in (Bxε)(By+ε)(B_{x}-\varepsilon) \cup (B_{y}+\varepsilon) is the same as for the two sets {2}{2i12in}\{2\} \cup \{2i-1 \mid 2 \leqslant i \leqslant n\} and {1}{2i2in}\{1\} \cup \{2i \mid 2 \leqslant i \leqslant n\}, so that Bxε<By+εB_{x}-\varepsilon < B_{y}+\varepsilon. In either case, we obtain
ABxεBy+εA A \prec B_{x}-\varepsilon \prec B_{y}+\varepsilon \prec A
which contradicts condition 3.

So we know now that the winner does not depend on a1,b1a_{1}, b_{1}. Therefore, we can define a new rule <<^{*} on sets of cardinality n1n-1 by saying that A<BA<^{*} B if and only if A{a}<B{b}A \cup \{a\} < B \cup \{b\} for some a,ba, b (or equivalently, all a,ba, b) such that a<minAa<\min A, b<minBb<\min B and A{a}A \cup \{a\} and B{b}B \cup \{b\} are disjoint. The rule <<^{*} satisfies all conditions again, so by the induction hypothesis, there exists an index ii such that A<BA<^{*} B if and only if the ithi^{\text{th}} smallest element of AA is less than the ithi^{\text{th}} smallest element of BB. This implies that C<DC<D if and only if the (i+1)th(i+1)^{\text{th}} smallest element of CC is less than the (i+1)th(i+1)^{\text{th}} smallest element of DD, which completes our induction.

Second case: ({2i11in1}{2n})({2i1in1}{2n1})(\{2i-1 \mid 1 \leqslant i \leqslant n-1\} \cup \{2n\}) \prec (\{2i \mid 1 \leqslant i \leqslant n-1\} \cup \{2n-1\}).
Set A={aaA}-A=\{-a \mid a \in A\} for any ARA \subseteq \mathbb{R}. For any two disjoint sets A,BRA, B \subseteq \mathbb{R} of cardinality nn, we write ABA \prec^{\circ} B to mean (B)<(A)(-B)<(-A). It is easy to see that \prec^{\circ} defines a rule to determine a winner that satisfies the three conditions of our problem as well as the relation of the first case. So it follows in the same way as in the first case that for some ii, ABA \prec^{\circ} B if and only if the ithi^{\text{th}} smallest element of AA is less than the ithi^{\text{th}} smallest element of BB, which is equivalent to the condition that the ithi^{\text{th}} largest element of A-A is greater than the ithi^{\text{th}} largest element of B-B. This proves that the original rule \prec also has the desired form.

Want a route through all this instead of an archive? The track puts 2,000 problems in a working order, from AMC 10 level to the IMO shortlist.

Source: MathNet, licensed CC-BY-4.0. Statement reproduced verbatim; metadata (topic, difficulty) added by this project.