IMO 1972 · Problem 1

Prove that from a set of ten distinct two-digit numbers (in the decimal system), it is possible to select two disjoint subsets whose members have the same sum.

Official text of the problem. Official problem paper (PDF)

Solution

Let \(S\) be the given set. Each of its ten elements can either belong or not belong to a subset, so \(S\) has \(2^{10}=1024\) subsets.

For each subset \(T\subseteq S\), define \[ s(T)=\sum_{t\in T}t, \] with \(s(\varnothing)=0\). Since every element of \(S\) is an integer between \(10\) and \(99\), \[ 0\le s(T)\le 10\cdot99=990. \] Thus the \(1024\) subset sums take at most \(991\) different values. By the pigeonhole principle, there are distinct subsets \(A,B\subseteq S\) with \(s(A)=s(B)\).

Set \[ C=A\setminus B,\qquad D=B\setminus A. \] These subsets are disjoint, and removing the common elements gives \[ s(C)=s(A)-s(A\cap B) =s(B)-s(A\cap B) =s(D). \] Moreover, both are nonempty: because \(A\ne B\), at least one is nonempty and therefore has positive sum; their equal sums force the other to be nonempty as well.

Hence \(C\) and \(D\) are two disjoint nonempty subsets whose members have the same sum. \(\square\)

The solution was checked against John Scholes' IMO solution archive (Kalva).