Refining Nash Equilibrium by Bayesian Iterative Conjectures Approach

Abstract

Bayesian equilibrium by iterative conjectures (BEIC) analyzes games with players forming their conjectures of what other players will do iteratively starting with first order uninformative conjectures (or prior distribution functions) and updating their conjectures iteratively with game theoretic reasoning until a convergence of conjectures is achieved, and this process of conjectures formation and updating itself is a common knowledge. The BEIC is a refinement of Nash equilibrium and narrows down the set of equilibrium, normally to a unique one. The paper compares how BEIC fares as a refinement of Nash Equilibrium with other refinements, including payoff-dominance, risk-dominance, iterated admissibility, subgame perfect equilibrium, Bayesian Nash equilibrium, perfect Bayesian equilibrium and the intuitive criterion.