Philip Ramirez – “Popularity, Beauty and Mispricing in Tennis Betting Markets” – PhD Seminar

Abstract:

Behavioral biases, such as the favorite-longshot bias, can lead to market inefficiencies manifesting as systematic mispricing in betting markets. Using bookmaker odds for tennis matches, as well as match results and player rankings – ability markers for performance – we attempt to uncover behavioral biases that contribute to mispricing in tennis betting markets, i.e. beauty and popularity bias. We introduce a deep learning facial recognition model to provide subjective beauty scores for hundreds of the top female tennis players in the Women’s Tennis Association. With popularity measured with daily Wikipedia article views, preliminary results include significant popularity bias and “beauty” bias that could explain some mispricing.