The longshot bias is an observed phenomenon in sports betting where on average bettors tend to overvalue (undervalue) “long shots” and undervalue (overvalue) the favorites. It has been studied extensively in betting markets such as horse racing, American football, ice hockey and soccer. The most plausible explanation is provided by behavior finance which studies the betting patterns of several types of bettors: informed, uninformed, amateur, professional using advanced models such as utility function theory. The tennis market, however, attracted much less attention, and very little research was conducted in this area.
Unlike other sports, the tennis market presents some particularities, mostly:
- The cost of becoming an “informed” bettor is relatively low. Hence there is no uninformed, “fun” bettor who just enjoys an outing at the courts as they often do in horse racing,
- Insider information is less valuable because almost everything (e.g. players’ injuries) is visible and disseminated quickly to the public,
- Unlike team sports, tennis does not have a large number of fan clubs with permanent and fervent allegiance. Therefore the prices (odds) are less likely to be distorted by supporters’ sentiment,
- The transaction cost is low.
Among noteworthy articles, the one published by D. Forest and I. McHale, entitled “Longshot bias, insight from the betting on men’s professional tennis”, presented interesting results regarding the longshot bias in the tennis market. Using publicly available data, the authors tabulated returns of strategies that bet on favorites and underdogs from January 2001 to April 2004. They demonstrated that:
- Betting on underdogs yielded statistically significant losses,
- Betting on favorites was break-even in 2001 and made positive returns in 2002-04.
From their results, it was concluded that men’s tennis market is weak form inefficient, and a positive longshot bias exists. Therefore good values are found by betting on strong favorites. Bettors should use this result in developing their strategies.