Can we throw out the record books? the effects of rivalry on market efficiency in college football betting markets

Abstract

Abstract This study examines whether rivalry games contribute to the mispricing of spreads in college football betting markets. We posit that rivalries increase player motivation and amplify bettor sentiment, potentially biasing closing spreads. We suggest that college football provides a compelling environment for this analysis because college rivalries carry deeper social identity ties, with potential bettors having direct affiliations via alumni status, family tradition, and geographical proximity. Using a large sample of games from 2008 to 2018, we employ logit and linear probability models to evaluate whether rivalry games affect the likelihood that the favored team covers the spread, as well as deviations from moneyline-implied win probabilities. Our results indicate that rivalry games are associated with a lower probability that the favored team covers the spread. However, this effect is not homogeneous. We find that the impact of rivalry is influenced by conference characteristics, particularly in matchups involving non-Power Five teams. Overall, while rivalry can contribute to mispricing under certain conditions, closing spreads mostly appear to incorporate rivalry dynamics.