Decision-Making with Risk and Uncertainty: The Case of Horse Racing

Abstract

Gambling provides valuable data for understanding human decisionmaking under conditions of risk and uncertainty. When decisions depend on elements of risk and uncertainty it is crucial to learn how they influence choice, for otherwise we will continue to misinterpret or at best inadequately evaluate the decision-making process. Horse racing offers a complete and accurately documented example of this type of human behavior. Writing in this JOURNAL, Griffith (1949) and McGlothlin (1956) demonstrated that horse-race betting is biased, i.e., the amount wagered on favorites is less than proportionate to their real winning chances while the opposite is true for horses with smaller chances of winning. Or, said differently, the ex ante subjective probabilities assigned by bettors to the favorites are less than the ex post objective winning probabilities, and the opposite is true for the other horses where subjective probabilities exceed objective probabilities. Subsequently, several nonpsychologists reconfirmed the bettor bias (Fabricand, 1965; Weitzman, 1965; Seligman, 1975; and Mukhtar, 1977). The analyses of bettor bias have focused uniquely on the public's betting behavior. However, like other professional sports, horse racing has its experts who aspire to outperform the public's predictions