An Experiment on the Value of Structural Information in a 2x2 Repeated Game

In experimental studies pairs that repeatedly play the simple coordination game mutual fate control may regularly fail to coordinate when they are given little in-formation, i.e. when subjects are uninformed about the payoff matrix and feed-back is limited to their own payoff. Our experimental study shows that the provision of a small amount of structural information prior to playing the game changes subject behaviour and significantly improves performance, even though standard adaptive learning rules do not take such information into account and optimal adaptive rules do not differ much between the two treatments. Our study calls for a more intense investigation into the cognitive processing of information.