Abstract
We examine the influence of bettor behavior in sports gambling markets and the resulting creation of exploitable betting opportunities for patient bettors. Specifically, we build on past research on behavioral bias as a predictor of bettor behavior and explore how this behavior can result in market inefficiencies. Using data from National Football League games taking place between 2007-2019, we find that bettor decision-making is erroneously influenced by recent performance of teams. This bias creates profitable betting opportunities for those less subject to recency bias, and are surprisingly greater for the more prudent, patient bettor. Our findings confirm the need for additional research examining the influence of psychology and behavioral biases on individual decision making and how these factors can influence market efficiency.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
Some researchers (e.g., Levitt 2004; Kain and Logan 2014; Paul and Weinbach 2012) note that bookmakers may choose, in some instances, to offer lines that seek to maximize profit (by tempting bettors to take the bookmaker’s perceived, poor side of a wager) rather than seek a riskless, commission-only model. Regardless of bookmaker intention, our results herein seek to describe a contrarian-type wagering opportunity holistically made possible by market participants.
We also considered omitting Week 17 data as incentives to team performance in the final week of the season vary widely. This omission did not substantively change our findings.
There are exceptions to this structure. Most commonly, for example, our opening line data in week X for teams participating in a week X-1 Monday Night Football game are not available until Tuesday of week X. There are never intervening games played between the opening and closing lines by either team in our sample observation games. While injuries are common in sport, particularly in a high contact sport such as professional football, the NFL attempts to disseminate information publicly, regarding injuries through the Physically Unable to Perform, or PUP list, and the Injured Reserve, or IR. Deadlines are established for teams to release their PUP and IR lists to reduce the uncertainty associated with key players’ participation. It is rather rare for the bettors or oddsmakers to be unaware of a key injury, and, at least until some clarity emerges on player participation, there is usually a delay in the issuance of the opening line of a game.
For robustness, we also consider whether the time since a team last played is material to our results. For example, each NFL team has one ‘open’ week on its schedule each year so that its prior performance in, e.g., Week 8 might have actually occurred in Week 6 rather than Week 7. Games also occur on Thursdays and Mondays each NFL week, as well as occasionally on Saturdays toward the end of each season. As we find no material differences in our results when considering the different times elapsed since teams’ prior games we omit the formal results herein.
For example, if Denver covered its prior game spread by 10, and Seattle failed to cover its prior game spread by 1, we say both Denver and Seattle contribute to the PriorGameATS value of 11. Games where Team B failed to cover its prior spread and Team A covered its prior spread, thus, exclusively make up the “both sides contributing” subsample. Alternatively, as an example, when Atlanta covered its prior game spread by 6, and Buffalo covered its prior game spread by 5, only one team (Atlanta) contributed to the PriorGameATS value of 1.
The Team B’s on which we wager, in order to implement our reversal strategy, need not be underdogs in their games. In fact, nearly half of Team B’s are actually favorites when we wager on them. We consider, in omitted results, whether the favorite/underdog status of Team B affects our results and find the distinction to be immaterial.
One estimate is that less than 0.5% of bettors are ‘professional,’ so that they might be considered ‘informed investors’ in a more traditional financial sense (Bluth 1997). Their wiseguy/professional budget for wagering is, on average, considerably larger than a typical bettor; however, the overall public interest in the NFL games appears so large as to limit the correcting influence of wiseguys on inefficiency in many games. Other games have very limited interest from the public so that wiseguys indeed move the price to a more efficient point, sometimes quickly after the opening line is issued.
In fact, some such hedge fund managers sought to ride the tech stock bubble, i.e., take the opposite approach to trading against inefficiencies. In the equity environment (where there is no analog to the end of an NFL game, which sets the value of a wager as final) the demand of the naïve public for tech stocks was even able to keep the price of some never-profitable stocks inflated for years. In extreme cases, hedge fund managers that traded correctly against the tech bubble even went insolvent due to client withdrawals before their assertions could be proven correct.
To further consider the impact of specific levels of PriorGameATS on line movement, as well as our strategy performance, in unpublished results we reconstruct the findings of Table 3, Panel B for the various levels of PriorGameATS (and consider alternative segmenting of the sample, e.g., 5-point windows, 7-point windows, etc.). Results follow the general pattern of profitability, with greater profitability at closing lines, as would be expected given the results of Table 1 and Table 2.
In unreported results, inspired by Dare and Dennis (2011), we consider whether our baseline reversal strategy results differ when the strategy dictates wagering on a home team or away team in an NFL game. The results do not materially differ.
In unreported results, inspired by Shank (2019), we consider whether our baseline reversal strategy results differ when the strategy dictates wagering in an intradivision NFL game or interdivision NFL game. The results do not materially differ.
References
Bluth A (1997) EARNING IT; Life’s a Gamble. A Few People Make It a Profession. The New York Times. Retrieved from http://www.newyorktimes.com/
Borghesi R (2007) The home team weather advantage and biases in the NFL betting market. J Econ Bus 59(4):340–354
Brunnermeier MK, Nagel S (2004) Hedge funds and the technology bubble. J Financ 59(5):2013–2040
Camerer CF (1989) Does the basketball market believe in the ‘Hot Hand’,? Am Econ Rev 79(5):1257–1261
Dare WH, Dennis SA (2011) A test for the bias of inherent characteristics in betting markets. J Sports Econ 12(6):660–665
Davis JL, Mcelfresh L, Krieger K, Fodor A (2015) Exploiting week 2 bias in the NFL betting markets. J Pred Mark 9(1):53–67
Edwards W (1968) Conservatism in human information processing. In Kleinmutz, F., ed.: Formal Representation of Human Judgment (Wiley, New York)
Fodor A, DiFilippo M, Krieger K, Davis JL (2013) Inefficient pricing from holdover bias in the NFL point spread markets. Appl Financial Econ 27(17):1407–1418
Gilovich T, Vallone R, Tversky A (1985) The hot hand in basketball: On the misperception of random sequences. Cogn Psychol 17:295–314
Goetzmann W, Massa M (2002) Daily momentum and contrarian behavior of index fund investors. J Financial Quant Anal 37(3):375–389
Kain KJ, Logan TD (2014) Are sports betting markets prediction markets?: Evidence from a new test. J Sports Econ 15(1):45–63
Krieger K, Fodor A (2013) Price movements and the prevalence of informed traders: The case of line movement in college basketball. J Econ Bus 68:70–82
Lee DD, Chan H, Faff RW, Kalev PS (2003) Short-term contrarian investing – is it profitable? Yes and no. J Multl Financ Manag 13(4–5):385–404
Levitt SD (2004) Why are gambling markets organized so differently from financial markets? Econ J 114:223–246
Murtha BR (2013) Peaking at the right time: Perceptions, expectations, and effects. Organ Behav Hum Decis Process 120(1):62–72
Nichols MW (2014) The impact of visiting team travel on game outcome and biases in NFL betting markets. J Sports Econ 15(1):78–96
Paul RJ, Weinbach AP (2005) Bettor preferences and market efficiency in football totals markets. J Econ Financ 29(3):409–415
Paul RJ, Weinbach AP (2012) NFL bettor biases and price setting: further tests of the Levitt hypothesis of sportsbook behaviour. Appl Econ Lett 18(2):193–197
Paul RJ, Weinbach AP, Humphreys B (2011) Revisiting the “Hot Hand” hypothesis in the NBA betting market using actual sportsbook betting percentages on favorites and underdogs. J Gambl Bus Econ 5(2):42–56
Sauer R (1998) The economics of wagering markets. J Econ Lit 36:2021–2064
Schiereck D, De Bondt W, Weber M (1999) Contrarian and momentum strategies in Germany. Financ Anal J 55(6):104–116
Shank CA (2019) NFL Betting Biases, Profitable Strategies, and the Wisdom of the Crowd. Int J Sport Financ 14:3–12
Tversky A, Kahneman D (1974) Judgment under uncertainty: Heuristics and biases. Science 185(4157):1124–1131
Vergin RC, Scriabin M (1978) Winning strategies for wagering on National Football League games. Manag Sci 24(8):809–818
Woodland BM, Woodland LM (2015) Testing Probability in the NBA Season Wins Total Betting Market. Int J Sport Financ 10:160–174
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
About this article
Cite this article
Krieger, K., Davis, J.L. & Strode, J. Patience is a virtue: exploiting behavior bias in gambling markets. J Econ Finan 45, 735–750 (2021). https://doi.org/10.1007/s12197-021-09557-5
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s12197-021-09557-5