Common knowledge of rationality in epistemic-probability models requires that players hold cautious beliefs.

The extent of this conflict is nicely illustrated in Samuelson (1992). In particular, Samuelson (1992) shows that there is no epistemic-probability model[14] of the above game with a state satisfying common knowledge of rationality (where “rationality” means that players do not choose weakly dominated strategies). Prima facie, this is puzzling: What about the epistemic-probability model consisting of a single state \(w\) assigned the profile \((u, l)\)? Isn’t this a model of the above game whe