Question: How often do odds reflect true probabilities in sports betting?

Research on sports betting odds has found that they often do not match the true probabilities of an event happening. Several studies have identified biases in how odds are set, such as the “favorite-longshot bias.” This bias occurs when underdogs are given worse value odds than favorites, even though the true probability of their winning may not justify such a difference (Wheatcroft, 2020; Shin, 1991; Kan, 2020).

There are several explanations for this bias. One reason is uncertainty about the available information, which can cause odds to be inaccurate (Makropoulou & Markellos, 2011). Bookmakers also use certain strategies to protect themselves against insider traders, which may contribute to the bias (Shin, 1991). In addition, bettors often make irrational decisions, which can affect the odds in ways that don’t reflect the true probabilities (Kan, 2020). Some researchers argue that odds may overreact to recent team performance, causing the odds to shift more than they should (Wheatcroft, 2020), while others point out that adjustments need to be made for the bookmaker’s margin, which can distort the odds (Clarke, 2016).

Although betting markets are generally considered efficient in predicting outcomes (Kan, 2020), discrepancies between psychological and true probabilities still exist (Griffith, 1949). Additionally, the method used to convert odds into probabilities can affect the results. Simple normalization techniques may produce biased probabilities, leading to incorrect conclusions (Štrumbelj, 2016). These biases can result from the structure of the betting market, the costs of acquiring information, or the behavior of the bettors themselves (Terrell & Farmer, 1996; Kan, 2020).

Bettors often make irrational decisions, which can affect the odds in ways that don’t reflect the true probabilities

Summary of: Kan, 2020

Anecdote

Have a story to share? Write to us at research@bettingresearch.org if you have a related, personal experience you would like to see placed here and share with the community.

Articles Cited

• “E. Wheatcroft (2020): Profiting from overreaction in soccer betting odds, https://doi.org/10.1515/jqas-2019-0009
  • The paper demonstrates the existence of a bias in soccer betting odds where teams that have performed above expectations in recent matches tend to be assigned poorer value odds than teams that have performed below expectations, and this bias can be exploited to generate a sustained and robust profit.”
• “Vasiliki Makropoulou, Raphael N. Markellos (2011): Optimal Price Setting in Fixed‐Odds Betting Markets Under Information Uncertainty, https://doi.org/10.1111/j.1467-9485.2011.00557.x
  • The paper develops a model of optimal pricing for fixed-odds betting markets that accounts for information uncertainty, and uses this model to explain the favorite-longshot bias observed in these markets.”
• “S. Clarke (2016): Adjusting true odds to allow for vigorish, –
  • The paper presents a mathematical model for adjusting probabilities from a predictive model to account for the bookmaker’s margin, known as the “”vigorish””, in a way that allows for the fact that the margin on outsiders is usually greater than on favorites.”
• “R. M. Griffith (1949): Odds adjustments by American horse-race bettors., https://doi.org/10.2307/1418469
  • The paper examines the relationship between the odds set by horse-race bettors and the actual probability of horses winning, in order to gain insights into the psychology of how people perceive and interpret probabilities.”
• “Erik Štrumbelj (2016): A Comment on the Bias of Probabilities Derived From Betting Odds and Their Use in Measuring Outcome Uncertainty, https://doi.org/10.1177/1527002513519329
  • The paper analyzes methods for deriving probability forecasts from betting odds and finds that the commonly used basic normalization (BN) method produces biased probabilities, and that the differences between probabilities produced by different methods can lead to contradictory conclusions when used to measure outcome uncertainty. The paper also provides evidence against the reported bias of bookmakers favoring better supported teams.”
• “H. Shin (1991): Optimal Betting Odds against Insider Traders, https://doi.org/10.2307/2234434
  • The paper examines the optimal pricing decision of a bookmaker facing a group of gamblers, some of whom have superior information, and finds that the bookmaker should follow a “”square root rule”” in setting betting odds, which results in the odds understating the winning chances of favorites and exaggerating the winning chances of longshots.”
• “A. Kan (2020): Efficiency, Bias, and Decisions: Observations from a Sports Betting Exchange, –
  • The paper examines the efficiency of sports betting markets, finding that they serve as good predictors of true outcomes but exhibit a bias where favorites are undervalued and longshots are overvalued, which the authors attribute to behavioral factors like prospect theory rather than information or transaction costs.”
• “D. Terrell, A. Farmer (1996): Optimal Betting and Efficiency in Parimutuel Betting Markets with Information Costs, https://doi.org/10.2307/2235361
  • The paper presents a model of parimutuel betting markets that explains several empirical observations, such as market odds failing to accurately predict outcomes and longshots earning lower expected values, as consequences of the track’s takeout and the presence of informed bettors who purchase true probabilities of events.”

Insufficient Detail?

At times it is difficult to answer the question as there are not enough relevant published journal articles to relate. It could be that the topic is niche, there’s a significant edge (and researchers prefer not to publish), there is no edge or simply no one has thought to investigate.