Mechanism Design Based on Beliefs about Responsive Play ( Position Paper )

In general, identifying a solution concept only incomplete ly specifies a mechanism design problem. The designer must consider which among a multiplicity of solutions is likely to be playe d, as well as the possibility that actual play will not correspond to any solution. Given that actual play is the ultimate determiner of a mechanism’s success, we advocate that designers embrace the cor responding forecasting problem and evaluate candidate mecha nisms with respect to belief distributions over players’ respons e. Solution concepts can play a useful role in delimiting and structurin g belief distributions. We propose that membership of prospective s rategy profiles in various solution classes be treated as evidence b earing on their likelihood of play. Flexible solution classes, for example based on approximate equilibrium, degree of dominance, or s afety level, provide natural measures (e.g., distance from equil ibri m) that can be employed in defining belief distributions.