A Simple Adaptive Procedure Leading to Correlated Equilibrium

We propose a simple adaptive procedure for playing a game. In this, players depart from their current play with probabilities that proportional to measures of regret for not having used other strategies (these measures are updated every period). It is shown that our adaptive guaranties that with probability one, the sample distributions play converge to the set of correlated equilibria of the game. To these regret measures, a player needs to know his payoff function the history of play. We also offer a variation where every player only his own realized payoff history (but not his payoff function).