Optimal stopping games under winning probability maximization and player's weighted privilege (Q2731714)

A game version of the best-choice problem is studied. \(m\) players sequentially observe i.i.d. random variables \(X_1,\dots, X_n\). Each player stops at one random value and wants to maximize the probability that his chosen value is larger than those of the others. Weights are assigned to the players, and if several players stop at the same random variable, the probability for any of them to get it is proportional to his weight. The optimal strategy is derived for \(m=2\) and discussed for \(m=3\). In the latter case equal weights are considered, and the game is modified by prescribing the strategy for the two remaining players after one has obtained his chosen value. Thus the problem for \(m\geq 3\) is still open. A no-information version of this game for two players is also treated.