Market evidence against widespread point shaving in college basketball

Abstract

This paper proposes using market-based information to investigate the likelihood of widespread point shaving in college basketball games with strong favorites. Information embodied in pre-game money lines facilitates the calculation of the market's expectation that the game will end with a strong favorite winning but not covering the point spread, a result deemed suspicious in previous studies. Additional market-based information embodied in second-half lines reveal how the market's expectation about second-half play differs from pre-game expectations. Applying our methodology to college basketball reduces previous estimates of the percentage of games thought to be consistent with potential point shaving to insignificant levels. Our approach can apply to other sports in which strong favorites are common.