We consider two state automata playing infinitely iterated two players, two strategies game, where each move can be mis-implemented (or mis-perceived) with a small error probability, and compute the payoff matrix by means of a perturbation approach.
展开▼