Abstract
Mathematical models have been shown to be no more successful in predicting the outcomes of professional football games than professional oddsmakers. Study I demonstrated that the strategy used by the oddsmakers can be approximated by a simple linear model which subtracts average points allowed on defense from average points scored on offense and adjusts for quality of opponent. In Study 2, football fans were asked to judge the quality of nine football teams on the basis of their average offense and their average defense. The nine teams were selected to vary systematically on these dimensions and to complete the cells of a 3 x 3 factorial design. Only half of the subjects were also told from which professional teams the data were obtained. When fans do not know the names of the teams, they tend to use the information similarly to oddsmakers. However, when team identities are also given, fans tend to use the information differently—making it less likely they will be successful in predicting outcomes.
Get full access to this article
View all access options for this article.
