Learning to signal: Analysis of a micro-level reinforcement model (Q1004397)

In a two player game theoretic setup, the authors consider the situation in which player one can observe a private signal, and player two's objective is to guess the signal correctly. Player one can transmit information about the signal, using an urn scheme, in which correct transmission, i.e. successful coordination, is reinforced. The authors show, that the urn scheme achieves asymptotically optimal coordination.