Betting Odds Research Articles

Pricing in response to new information: The case of betting markets

AbstractMarkets are information aggregators. But how do they incorporate new data into their pricing? We examine the response of prediction markets to a novel information shock in a quasi‐natural experiment: How did the absence announcements of elite soccer players influence the betting odds of affected matches? Analyzing the first four statistical moments of 117,174 odds from 32 bookmakers, we identify initial inertia followed by a lagged reaction that we cannot reason with learning. Our findings raise questions about how bettors and bookmakers incorporate new information into their beliefs. It has broader implications regarding information processing in markets.

Comparing Two Methods for Testing the Efficiency of Sports Betting Markets

Sports betting markets can be considered strongly efficient if expected returns on all possible bets on an event are equal. If this form of efficiency holds, then there is a direct mapping from betting odds into probabilities of outcomes of sporting events. We compare two regression-based methods for testing this form of efficiency that have been used in previous research: One that uses normalized probabilities as the explanatory variable for event outcomes and one that uses the inverse of the decimal odds. We show that the normalized probability method produces good tests of the null hypothesis of strong market efficiency but that the inverse odds method does not, with results biased against finding favorite-longshot bias. We illustrate this finding using large datasets of bets and outcomes for tennis and soccer and also with realistic simulations.

Returns on complex bets: evidence from Asian Handicap betting on soccer

PurposeThe Asian Handicap is a way to bet on soccer matches where payouts depend on an adjustment to the score that favors the weaker team. These bets can feature the possibility of all or half the bet being refunded and this makes the calculation of their expected return more complex than for traditional betting on a home win, away win or draw. We examine the behavior of odds in this market.Design/methodology/approachIn addition to a using well-known publicly available source of information on Asian Handicap betting odds – which provides the average odds across a range of bookmakers – we have also sourced a large dataset of Asian Handicap odds offered by an individual bookmaker.FindingsWe show that bettors systematically lose more money on Asian Handicap bets where refunds are not possible than when it is possible to obtain a half refund. We also show that bets with the possibility of a full refund have the lowest loss rates. We demonstrate that this pattern of differences in loss rates across bets is predictable based on the odds quoted. This pattern could represent preferences, with gamblers disliking bets featuring potential refunds, but we argue the evidence points more towards gamblers incorrectly calculating expected loss rates.Originality/valueDespite being one of the world's largest betting markets, there has been almost no previous research on the properties of the Asian Handicap soccer betting. Our finding of clear differences in returns on simultaneously available bets on the same team is also a new anomaly previously undocumented in any research on sports betting.

Hockey babies: National Hockey League outcomes and short-term fertility spikes in Canada

Previous research shows that mood-altering events, such as sports results, elections outcomes and natural disasters, impact fertility, crime rate, and investor behaviour. In this paper, we use recent Canadian Birth Statistics and the National Hockey League (NHL) results to examine the link between sports events and short-term fertility spikes. In addition, using betting odds, we differentiate among unexpected wins, unexpected losses, and expected outcomes, as a test for the relevance of the Prospect Theory to this context. Our dataset is a daily panel, following all the seven Canadian Census Metropolitan Areas with a team in the NHL, from 2008 to 2019. In our panel fixed effects estimations, we account for the day of the week and all statutory federal and provincial holidays of Canada. In the estimations with raw NHL results, the coefficients of interest ultimately lose their statistical significance. In the specifications accounting for expectations, we find a statistically significant association with fertility for unexpected wins, and statistically insignificant coefficients for unexpected losses as well as expected NHL outcomes. In light of the results, we propose the greater likelihood of pre-game social gatherings on sports nights and celebratory sex after the euphoria of an unexpected win, as the possible channels of impact.

AI-based betting anomaly detection system to ensure fairness in sports and prevent illegal gambling

This study develops a solution to sports match-fixing using various machine-learning models to detect match-fixing anomalies, based on betting odds. We use five models to distinguish between normal and abnormal matches: logistic regression (LR), random forest (RF), support vector machine (SVM), the k-nearest neighbor (KNN) classification, and the ensemble model—a model optimized from the previous four. The models classify normal and abnormal matches by learning their patterns using sports betting odds data. The database was developed based on the world football league match betting data of 12 betting companies, which offered a vast collection of data on players, teams, game schedules, and league rankings for football matches. We develop an abnormal match detection model based on the data analysis results of each model, using the match result dividend data. We then use data from real-time matches and apply the five models to construct a system capable of detecting match-fixing in real time. The RF, KNN, and ensemble models recorded a high accuracy, over 92%, whereas the LR and SVM models were approximately 80% accurate. In comparison, previous studies have used a single model to examine football match betting odds data, with an accuracy of 70–80%.

Forecasting number of corner kicks taken in association football using compound Poisson distribution

This article presents a holistic compound Poisson regression model framework to forecast number of corner kicks taken in association football. Corner kick taken events are often decisive in the match outcome and inherently arrive in batch with serial clustering pattern. Providing parameter estimates with intuitive interpretation, a class of compound Poisson regression including a Bayesian implementation of geometric-Poisson distribution are introduced. With a varying shape parameter, the corner counts serial correlation between matches is handled naturally within the Bayesian model. In this study, information elicited from cross-market betting odds was used to improve the model predictability. Margin application methods to adjust market inefficiency in raw odds are also discussed.

A comparative evaluation of Elo ratings- and machine learning-based methods for tennis match result prediction

Elo ratings-based methods, including the recently proposed Weighted Elo method, have been found to perform well when forecasting tennis match results, however, whether they can outperform machine learning (ML) has not been established. In this study, a comparative evaluation of the two types of methods is conducted using the Sports Result Prediction CRISP-DM experimental framework. The first full year of mens ATP tennis data (2006), in a dataset containing matches from 2005 to 2020, was set to be the initial training set and 1 year of data was incrementally added to this set to predict 14 test years, from 2007 to 2020. Features were ranked based on their average rank across five feature selection techniques. It was found that, of the five ML models, Alternating Decision Trees (ADTrees) and Logistic Regression achieved higher accuracies than Elo ratings and similar accuracies to predictions derived from betting odds. Furthermore, ADTrees show potential in this domain, with solid performance achieved with an interpretable decision tree that allows for variation in the average betting odds difference threshold.

Broadcasting demand for Formula One: Viewer preferences for outcome uncertainty in the United States

ABSTRACT We explore the determinants of broadcasting demand for Formula One racing in the United States of America and pay specific attention to the relationship between outcome uncertainty and television viewership figures. Using implied probabilities derived from betting odds, we offer an approach to measuring outcome uncertainty that differs from those currently established in the Formula One broadcasting demand literature. We do not find any evidence to support the uncertainty of outcome hypothesis. On the contrary, viewers display a preference for less closely contested races. We find that scheduling and accessibility to live races are important determinants of viewership. Our results have practical implications and can guide the sport's administrators and television broadcasters seeking to understand a growing national market for F1 broadcasting.

Does the Game Matter? Analyzing Sponsorship Effectiveness and Message Personalization in Sport Live Broadcasts

This study aims to increase the effective use of in-stadium sponsor message placement by analyzing the influence of various run-of-play characteristics on television viewers’ visual attention allocation. Sports broadcasts constitute one potential platform for sponsors to place personalized messages. However, literature still questions the effectiveness of in-stadium sponsor messages, and the influence of game-related factors on viewers’ visual attention has received little consideration in this context. In addition, researchers call for more reliable and realistic measures concerning the effective evaluation of sponsorship-linked marketing. Therefore, this study uses real-time adaptions (eye-tracking, in-play betting odds, etc.) utilizing live soccer broadcasts as one of the first. Data were analyzed second by second (n = 100,298) using generalized linear mixed models. Results indicate significant associations of several run-of-play characteristics with viewers’ visual attention to sponsor messages depending on the characteristic, the games’ degree of suspense, and playing time. Findings provide hands-on advice for practitioners to enhance sponsor message placement during live broadcasts.

Does Foreign Investment Affect Sporting and Managerial Efficiency? The Case of English Football Clubs

Abstract: The number of foreign owners in football clubs has recently increased in Europe and especially in England. This study aims to analyze whether clubs owned by foreign investors show different parameters of sporting and managerial efficiency. We use data on expected performance (from betting odds), wage costs, and the results of football clubs in the English Premier League (EPL) and English Football League Championship (EFL) over 13 seasons. We estimate different measures of efficiency using three alternative methods: expected vs. actual performance index, nonparametric conditional data envelopment analysis (DEA), and stochastic frontier, while considering foreign ownership as an exogenous factor. Our results suggest that there are not significant differences between domestic- and foreign-owned clubs.

A framework of interpretable match results prediction in football with FIFA ratings and team formation.

While forecasting football match results has long been a popular topic, a practical model for football participants, such as coaches and players, has not been considered in great detail. In this study, we propose a generalized and interpretable machine learning model framework that only requires coaches' decisions and player quality features for forecasting. By further allowing the model to embed historical match statistics, features that consist of significant information, during the training process the model was practical and achieved both high performance and interpretability. Using five years of data (over 1,700 matches) from the English Premier League, our results show that our model was able to achieve high performance with an F1-score of 0.47, compared to the baseline betting odds prediction, which had an F1-score of 0.39. Moreover, our framework allows football teams to adapt for tactical decision-making, strength and weakness identification, formation and player selection, and transfer target validation. The framework in this study would have proven the feasibility of building a practical match result forecast framework and may serve to inspire future studies.

The Longshot Bias Is a Context Effect

In nearly every betting market, gamblers overvalue the riskiest bets. This “longshot bias” is usually explained by a tendency to overweight low-probability events. We offer a novel explanation: contrast effects enhance the attractiveness of longshots because gambles presented in terms of their payoffs are easier to compare along the payoff dimension than along the probability dimension. Our analysis of historical betting odds suggests that contrasts between gamble payoffs can explain the longshot bias. Laboratory experiments confirm those contrast effects and show that the longshot bias disappears when gamblers consider bets in isolation or when winning probabilities are easier to compare. This paper was accepted by Manel Baucells, behavioral economics and decision analysis. Funding: The authors thank the University of Chicago’s Center for Decision Research and Richard Thaler for funding. Supplemental Material: The online supplement and data are available at https://doi.org/10.1287/mnsc.2023.4684 .

Systematic Analysis of Position-Data-based Key Performance Indicators

Abstract In the past 20 years, performance analysis in soccer has accumulated a wide variety of key performance indicators (KPI’s) aimed at reflecting a team’s strength and success. Thanks to rapidly advancing technologies and data analytics more sophisticated metrics, requiring high resolution data acquisition and big data methods, are developed. This includes many position-data-based KPI’s, which incorporate precise spatial and temporal information about every player and the ball on the field. The present study contributes to this research by performing a large-scale comparison of several metrics mainly based on player positions and passing events. Their association with team’s success (derived from goals scored) and team’s strength (estimated from pre-game betting odds) is analysed. The systematic analysis revealed relevant results for further KPI research: First, the magnitude of overall correlation coefficients was higher for relative metrics than for absolute metrics. Second, the correlation of metrics with the strength of a team is stronger than the correlation with the game success of a team. Third, correlation analysis with team strength indicated more positive associations, while correlation analysis with success is most likely confounded by the intermediate score line of a game and revealed more negative associations.

Estimating the Implied Probabilities in the Tennis Betting Market: A New Normalization Procedure: Comment

Abstract: In this note, we comment on a recent paper in the journal that proposes a new normalization procedure when converting tennis betting odds to the implied probabilities of each player winning. The new procedure is especially germane for matches in which there is a heavy favorite and where there is concern that traditional conversion methods understate the true probability of the favorite winning. However, in this comment, we argue that the new adjustment, while an interesting contribution, suffers from at least three limitations that make the procedure relatively costly while not materially improving predictions of match outcomes.

EXPLOITING THE MAIN BIASES IN AMERICAN SOCCER BETTING MARKETS.

In this paper the author will discuss the main inefficiencies in the 4 American major leagues of football (Argentina, Brazil, Mexico and USA), proposing a betting scheme, which outperforms the market with consistency over the past 9 Seasons (from 2012/2013 up to 2020/2021). Analysing the available data it was discovered how “Home advantage bias” and “Championship favourite advantage bias” play a big role in these championships and, by betting only on events, in which they were predicted favourable betting odds, it was possible to obtain an average return of +6.87% over 567 matches, using average bookmakers coefficients available on the market. The results are particularly significant, since, at least to our knowledge, there were no similar strategies available, which could guarantee such a big positive outcome for the leagues presented.

Full-shape galaxy power spectra and the curvature tension

ABSTRACT With recent evidence for a possible ‘curvature tension’ among early and late universe cosmological probes, Effective Field Theories of Large Scale Structure (EFTofLSS) have emerged as a promising new framework to generate constraints on Ωk that are independent of both CMB measurements, and some of the assumptions of flatness that enter into other large-scale structure analyses. In this work we use EFTofLSS to simultaneously constrain measurements from the 6dFGS, BOSS, and eBOSS catalogues, representing the most expansive full-shape investigation of curvature to date. Using the Bayesian evidence ratio our full-shape data assigns betting odds of 2:1 in favour of curvature, indicating present measurements remain broadly compatible with both flat and curved cosmological models. When our full-shape sample is combined with Planck 2018 CMB measurements, we break the geometric degeneracy and recover a joint fit on Ωk of $-0.0037^{+0.0025}_{-0.0026}$. Using the suspiciousness statistic we find evidence for a moderate tension between Planck 2018 and our suite of full-shape measurements, at a significance of $1.86 ^{+0.13}_{-0.11} \sigma$ (p ∼ 0.06 ± 0.02). These results demonstrate the usefulness of full-shape clustering measurements as a CMB independent probe of curvature in the ongoing curvature tension debate.

On sticky bookmaking as a learning device in horse-racing betting markets

We present a stochastic dynamic model of the adjustment of betting odds by bookmakers in a horse-racing betting market. We use optimal stopping theory in a two-horse benchmark model with both informed and noise punters. A costly learning process discloses what information the informed traders possess and a risk-neutral bookmaker selects a stopping time at which the betting odds for each horse in a race are adjusted. Our main finding shows that an increased fraction of informed punters has a non-monotonic effect on the loss per trade to the bookmaker. We also find that as the fraction of noise traders goes up, the learning process is less informative, so that the decision to change the prices for each horse is taken sooner.

Female jockeys - what are the odds?

Under-representation of women persists in many industries and represents an important area of concern for society. We use a revealed preference approach to test for bias against females in an underexplored environment. Whilst much use has been made of the financial industry to examine how market prices reveal implicit views on the relative productivity of men and women, our setting offers advantages through both volume of data and unambiguity of outcome. Over a 20-year period, the effect of jockey gender on fixed price betting odds was examined in National Hunt racing. Employing censored regression to account for non-finishers we find female jockeys to be underestimated by the UK betting market. Results indicate an increasing trend for underestimation in recent years, despite growing representation and rising performance levels of female jockeys. We conclude that mistake-based discrimination and confirmation bias may be impacting efficiency in the betting market. The market might recognise some improvement in female performance but may be failing to adapt at the speed with which female jockeys are professionalising.

The effectiveness of betting odds in digital gambling advertising for sports betting

ABSTRACT Digital gambling advertising is omnipresent nowadays and leads young adults easily and directly to online gambling websites, challenging their self-control and inducing impulsive behaviors. The current study examines how including betting odds in digital gambling advertising for sports betting affects young adults’ gambling intentions and gambling attitudes toward sports betting and whether positive affective reactions toward the advertising can explain this. In addition, the moderating roles of gender and advertising format are investigated. The results of a two (advertising strategy: no odds advertising versus odds advertising) by two (digital advertising format: social media ad versus online banner) between-subjects experiment among 150 young adults (18 to 25 years) show that odds advertising (compared to an ad not including odds) led to more positive affective reactions toward the sports betting advertisement, which subsequently positively affected both gambling intentions and gambling attitudes. However, these results were only true for men and not for women. Finally, there was no moderating role of digital advertising format.

Influence of Red and Yellow cards on team performance in elite soccer

The aim of the current study is to analyze the effects of red and yellow cards on the scoring rate in elite soccer. The sample was composed of 1826 matches in the top five European leagues. All events were structured in 5-min intervals and were analyzed by means of a Generalized Linear Mixed Model with Poisson distribution, considering the presence of correlated data, where the dependent variable is represented by scoring rate. Team strength and home advantage were considered implicitly by means of a transformation of the betting odds for each game. The model also took into account the goal difference and time evolution. Overall, we found that after a sending off, each team’s scoring rate changes significantly, damaging the penalised team and favouring its opponent. When the player who is sent off belongs to the Away team, the impact of a red card is more or less maintained over time intervals. The red card effect, on the other hand, tends to fade over time when the affected team is stronger. The relative difference in scoring rates is also affected by the goal difference and the difference in booked players, being slightly lower for the team going ahead if it has more booked players. Our approach allows estimating the expected cumulative soring rate through time for various red card scenarios. Particularly if a red card is given with 30 min of remaining time, the expected impact is 0.39 goals if the guilty player is on the visiting team and 0.50 if he plays for the home team. Coaches and analysts could use this information to establish objectives for players and teams in training and matches and to be prepared for these very different scenarios of numerical superiority or inferiority.