A Critical Assessment of the Evolution of Standard Game Theory

In this chapter, we adopt a history of economic thought approach to highlight the methodological problems encountered by standard game theory in its treatment of strategic reasoning and the consequences this has on the nature of players’ beliefs. We show that whatever contributions in the history of standard game theory; of von Neumann and Morgenstern of Nash of the refinement program in the 1980s and of epistemic game theory; by focusing only on the search for a solution concept, these contributions reduce strategic rationality to the existence of a solution. They reduce a game to a simple “black box,” regardless of how the game is played. We show that such focus on solution concepts is related to the norms of research of the mathematical community. We show in particular that this leads to serious difficulties in the epistemic program of game theory, which aims to define the knowledge and belief structure of players compatible with the concepts of solution. Epistemic game theory is grounded on the idea that to guarantee the existence of a solution, players’ beliefs must be a priori coherent and correct. From that prospect, we show that players’ beliefs are not individual and subjective mental states; regardless of the nature of the players’ beliefs a priori, they are by construction the result of their reasoning and not an element intervening in this reasoning. This means that players’ reasoning about other players’ choices or beliefs is excluded from the analysis of epistemic games, as well as strategic rationality.