There is little evidence that college basketball teams are involving in point-shaving, according to new research coming out of the University of Maryland.
Maryland finance professor Steve Heston and Dan Bernhardt of the University of Illinois reviewed scores and betting lines from more than 40,000 of games since 1990, and concluded that previous research suggesting widespread corruption in men’s basketball is flawed.
The two researchers concluded that scoring patterns for games in which the spread increases — meaning money is being bet on the favorite — are exactly the same as when the spread shrinks. Moreover, in games where there was no spread and therefore no betting, those same scoring patterns appeared.
Their findings will appear in an upcoming edition of the academic journal Economic Inquiry.
Heston and Bernhardt’s findings refute those from University of Pennsylvania economist Justin Wolfers, who published a paper in 2006 arguing that heavily favored teams failed to cover the spread too often to be explained by just chance. In his paper, he suggested the possibility of widespread point-shaving across college basketball. (In the most common point-shaving scandals, players miss baskets or otherwise play poorly to avoid having their team cover the spread.)
The latest research from Maryland agreed with Wolfers’ findings that heavy favorites win but fail to cover more than medium favorites and underdogs. But the research differed from Wolfers’ in one key respect, in that it examined how betting lines moved according to how money was bet. The researchers theorized that if point-shaving on favored teams was occurring, there would be more money placed on an underdog, thereby shrinking the betting line. But they said there was no correlation between how a game finished and how the point spread moved in the days leading up to the game.
To account for the possibility of small-scale betting that would not cause betting lines to move, Heston and Bernhardt studied lower-profile games that bookmakers ignored (they used Winthrop vs. Gardner-Webb as an example.) Using Sagarin ratings to determine a mock point spread, the researchers concluded that scoring patterns were generally consistent across all games involving heavy favorites and weak favorites, including those where no betting took place.
So why is it that heavy favorites fail to cover more than other teams? Heston and Bernhardt say it’s not the result of point-shaving, but the nature of the way basketball is often played at the end of games. If a team is winning by a small margin, they said, that team is likely to slow down their pace of play to preserve a lead.
“The world is not conspiring against you so you lose $100 during March Madness,” Heston said in an interview. “Sports are unpredictable, and fortunately it’s not connected to the money and there’s no epidemic of corruption.”
You can check out the research here. (Warning: it’s an academic paper and slightly denser reading than your copy of Sports Illustrated.)