Weakly-Bayesian and Consistent Assessments

Abstract

A weakly-Bayesian assessment is computed applying Bayes rule at positive probability information sets. We characterize the set of extensive-forms for which the sets of weakly-Bayesian and consistent assessments coincide. In doing so we disentangle the different restrictions imposed by consistency across information sets. We apply this knowledge to strengthen weakly-Bayesian assessments and to derive conditions for equivalence with consistency that can be useful in economic applications.