First, the 17-week season provided participants with quick, frequent, and unambiguous feedback over a significant (and nonarbitrary) duration of time, and thus it provided an ideal context for evaluating the effect of experience on optimism. Second, NFL fans’ preferences for their favorite teams are strong and often held with a degree of intensity unlikely to be generated by incentives offered in the laboratory. Third, a number of alternative explanations for the effects of desirability, such as those implicating team strength and familiarity, can be controlled methodologically and statistically. Finally, unlike predictions in other emotionally important domains, football predictions offer the benefit of objective benchmarks—both ex ante and ex post—against which the accuracy of predictions can be evaluated.Their results:
We found that people are optimistic in their predictions—they judge preferred outcomes to be more likely than nonpreferred outcomes. We extended this observation in two important ways. First, we showed that optimism persists despite extensive experience—football fans are as optimistic after 4 months of feedback as they are after 4 weeks of feedback. Second, we found that desirability fueled this optimistic bias. Using four distinct tests, and a wide variety of control variables, we found that optimistic predictions were robust and uniquely related to the desirability of the outcome.