Learning and Equilibrium Selection in 2x2 Games: Some Simulation Results