Optimal strategies for the uninformed player

This paper provides a counterpart of the well-known recursive formula for finitely repeated zero-sum games with lack of information on one side from the point of view of the uninformed player. Applying the results to a specific game that has already been discussed by Mertens and Zamir, for the first time a nontrivial sequence of values and optimal strategies is explicitly computed. This allows to recover their startling result on the appearance of the normal law in such models by entirely different arguments.