The full moon effect and the efficiency of point-spread betting in the NFL

Abstract

This study examines the efficiency of point-spread betting lines of the National Football League, relative to the observed outcome of the games, using data on all regular season games between 1978 and 2007. In addition to replicating and extending the existing literature on the home-underdog bias as well, as the late-season bias, new and significant contributions have been made. This study identifies bias in the lines that is occurring as a result of the lunar cycle. In particular, the home team is being consistently and significantly understated on the day before the full moon creating possible arbitrage opportunities. The models present a full-moon bias that is most relevant the day before the full moon. The bias also changes during the course of a season, and the time span covered by the data set.