Heterogeneous Noise and Stable Miscoordination

Coordination games feature two types of equilibria: pure equilibria, where players successfully coordinate their actions, and mixed equilibria, where players frequently experience miscoordination. We investigate learning dynamics where agents observe the actions of a random sample of their opponents. First, we show that when all agents have the same sample size, whether it is small or large, their behavior converges to one of the pure coordinated equilibria. By contrast, our main results show that stable miscoordination often persists when some agents make decisions based on small samples while others rely on large samples. (JEL C72, C73, C78)