Implementation In Quantal-Response Equilibia

 

Norovsambuu Tumennasan

 

   We study the classical implementation problem when players are prone to make mistakes. To capture the idea of mistakes, Logit Quantal-Response Equilibrium (LQRE) is used and we consider a case where players are almost rational, i.e., the sophistication level of players, λ, approaches infinity. We show quasi-monotonicity, a small variation of Maskin monotonicity, and no worst alternative conditions are necessary for restricted Limiting LQRE (LLQRE) implementation. Moreover, these conditions are sufficient for LLQRE implementation if there are more than three agents and social choice rule satisfies no veto power.