Friedman, Eric; Shenker, Scott - 1998
We address the problem of learning and implementation on the Internet. When agents play repeated games in distributed … environments like the Internet, they have very limited {\em a priori} information about the other players and the payoff matrix …