We report experiments designed to test the theoretical possibility, first discovered by Shapley (1964), that in some games learning fails to converge to any equilibrium, either in terms of marginal frequencies or of average play. Subjects played repeatedly in fixed pairings one of two 3 × 3...

History offers many examples of dictators who worsened their behavior significantly over time (like Zimbabwe's Robert Mugabe), while there are also cases of dictators who have displayed remarkable improvements (like Jerry Rawlings of Ghana).  We show that such mutations can result from rational...

Models of macroeconomic learning are populated by agents who possess a great deal of knowledge of the "true" structure of the economy, and yet ignore the impact of their own learning on that structure; they may learn about an equilibrium, but they do not learn within it.  An alternative learning...

Disinflationary episodes are a valuable source of information for economic agents trying to learn about the economy.  This paper is especially interested in how a policymaker can themselves learn by disinflating.  The approach differs from the existing literature, which typically focuses on...

Why do people stall while bargaining? Why are people keen to conclude a deal quickly, only to subsequently allow delay before the pie is realised? We propose that the reason is not fully explained by discount rates in combination with agents being engaged in a signalling equilibrium with...

In models of learning by experimentation, there is a natural benchmark of myopia when the only intertemporal link is the agent`s subjective belief (signal independence). An alternative benchmark using a passive learner has been proposed when there is a further intertemporal link that directly...

The Folk Theorem for infinitely repeated games offers an embarrassment of riches; nowhere is equilibrium multiplicity more acute. This paper selects amongst these equilibria in the following sense. If players learn to play an infinitely repeated game using classical hypothesis testing, it is...

This paper considers a simple adaptive learning rule in Bayesian games where players employ threshold strategies. Global convergence results are given for supermodular games and potentital games.

The paper surveys recent work on learning in games and delineates the boundary between forms of learning that lead to Nash equilibrium and forms that lead to weaker notions of equilibrium (or none at all).