Comment on 'Making Decisions in Large Worlds' by Ken Binmore

Abstract

In this brilliant, provocative new piece, KEN BINMORE continues in his efforts to "debayes" game theory. The current paper, given as the ADRES 2006 lecture in Marseilles, is part of a broad research agenda, where Professer BINMORE aims at refounding game the, by questioning some of its basic tenets. In this lecture, he focusses attention on the inadequacy of Bayesian reasoning to handle decision making and strategic interaction in large worlds. The distinction between small and large worlds in SAVAGES Foundation of Statistics, is hardly format or precise. A small world is a situation where the decision maker understands all the contingencies, and Bayesian decision making applies; a large world is characterized by uncertainty (about the events and their likelihood), and ambiguity. Clearly, the tools and methods of Bayesian statistics are inadequate to handle decision-making in large worlds.