Learning across games

► We model a process of learning across games. ► We show that equilibrium selection can be very different with learning across games compared to learning in a single game. ► We show that the approach can explain several well known experimental results.