We study a dynamic process where agents in a network interact in a Prisoner’s Dilemma. The network not only mediates interactions, but also information: agents learn from their own experience and that of their neighbors in the network about the past behavior of others. Each agent can only...

Conventional wisdom usually suggests that agents should use all the data they have to make the best possible prediction. In this paper, it is shown that agents may make better predictions by discarding old data if their model is mis-specified. The applicability of the results to some economic...