Sequential Parimutuel Betting in the Laboratory

This paper investigates parimutuel betting in the laboratory. Our experimental design relies on a simple sequential betting game where equilibrium strategies are characterized according to objective probabilities, the number of bettors, and publicly observable odds. The empirically well-documented phenomenon referred to as the "favorite-longshot bias" is observed in two of our three treatments. We offer a theoretical explanation of the subjects' behavior which relies on rank-dependent expected utility and pessimistic expectations about future bets.