Learning in Games by Random Sampling ¤ Depart ment of Economics Working Paper 99-04 James W . Friedman Depart ment of Economics University of North Carolina Chapel Hill, NC 27599-3305 email: Jim_ Friedman@unc.edu Claudio Mezzetti Depart ment of Economics University of North Carolina Chapel Hill, NC 27599-3305 email: mezzet t i@email.unc.edu November 16, 1999 A bst r act We st udy repeat ed int eract ions among a …xed set of “ low rat ionality” play- ers. Each player has a status quo action. Occasionally, he randomly samples other actions and changes his status quo if the sampled action yields a higher payo¤. This behavior generates a random process, the better-reply dynamics. ¤ We thank seminar participants at Boston and Northwestern Universities for stimulating com- ments and particularly Russell Cooper, Eddie Dekel and Bob Rosenthal. We thank the associate edit or and an anonymous referee for helpful comment s. 1