Author:

Abstract

There is an uncomfortable cohabitation between expected utility theory (hereafter EU theory) and paradoxes that threaten its normative stature. This essay demonstrates that the paradoxical nature of the St. Petersburg game is only true in a positive sense. That is to say, the game is counterintuitive because real individuals exhibit behaviour that diverges from rational expectations. The game’s descriptive inconsistencies have no bearing on norms established for a rational agent, and pose no challenge to the normative relevancy of EU theory.

In contrast, the inability of EU theory to prescribe a stable expectation for the Pasadena game exposes more enduring normative troubles. Possible resolutions include adopting finite states and bounded utilities, weak expectations, and statistical parameters, although shortcomings limit the applicability of each. The addition of the Altadena game illustrates the possibility of competing decision rules. In light of these observations, I advise a lexicographic approach that prioritizes and reconciles pluralistic rules as a necessary remedy to the normative ails of EU theory.