Solving the Problem of Inadequate Scoring Rules for Assessing Probabilistic Football Forecast Models

Constantinou, Anthony C.; Fenton, Norman

doi:10.1515/1559-0410.1418

Cited by 50 publications

(31 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We explained why it was the most rational scoring rule of those that have been proposed and used for football outcomes in (Constantinou & Fenton, 2012a). In general, this scoring rule represents the difference between the observed and forecasted cumulative distributions in which a higher difference leads to a higher penalty (Wilks, 1995), which is subject to a negative bias that is strongest for small ensemble size (Jolliffe & Stephenson, 2003).…”

Section: Accuracy Measurementmentioning

confidence: 99%

pi-football: A Bayesian network model for forecasting Association Football match outcomes

Constantinou

Fenton

Neil

2012

Knowledge-Based Systems

116

View full text Add to dashboard Cite

A Bayesian network is a graphical probabilistic belief network that represents the conditional dependencies among uncertain variables, which can be both objective and subjective. We present a Bayesian network model for forecasting Association Football matches in which the subjective variables represent the factors that are important for prediction but which historical data fails to capture. The model (pi-football) was used to generate forecasts about the outcomes of the English Premier League (EPL) matches during season 2010/11 (but is easily extended to any football league). Forecasts were published online at www.pi-football.com prior to the start of each match. In this paper, we demonstrate that a) using an appropriate measure of forecast accuracy, the subjective information improved the model such that posterior forecasts were on par with bookmakers' performance; b) using a standard profitability measure with discrepancy levels at ≥ 5%, the model generates profit under maximum, mean, and common bookmakers' odds, even allowing for the bookmakers' built-in profit margin.Hence, compared with other published football forecast models, pi-football not only appears to be exceptionally accurate, but it can also be used to 'beat the bookies'.

show abstract

Section: Accuracy Measurementmentioning

confidence: 99%

pi-football: A Bayesian network model for forecasting Association Football match outcomes

Constantinou

Fenton

Neil

2012

Knowledge-Based Systems

116

View full text Add to dashboard Cite

show abstract

“…Subsequent analyses have documented the existence of mispricing as well as the apparent existence of pure arbitrage possibilities in European fixed-odds football betting markets [3][4][5][6][7][8][9]. There are numerous papers that have investigated market efficiency in other sports betting markets such as horse racing, National Football League, greyhounds, National Hockey League, Major League Baseball and National Basketball Association (see Sauer [10] and Williams [11] for surveys).…”

Section: Introductionmentioning

confidence: 99%

Systematic Positive Expected Returns in the UK Fixed Odds Betting Market: An Analysis of the Fink Tank Predictions

Buraimo

Peel

Simmons

2013

IJFS

View full text Add to dashboard Cite

Abstract:We report striking evidence of semi-strong inefficiency in the UK fixed-odds football betting market using a reputable newspaper tipster which offers probabilities of match outcomes rather than simple result indicators. Betting on the Fink Tank probabilities of home wins across 10 bookmakers, when there are positive expected returns, would have generated positive returns in each of the seasons from 2006-07 to 2011-12 for a variety of different betting strategies. These returns could have been enhanced by employing the best odds from a greater number of bookmakers. However, the fact that pure arbitrage bets have existed for years and appear to last for several hours or days suggest they are in practice not exploitable to a magnitude that poses any threat to bookmakers.

show abstract

“…One common choice is specifying the variance, namely, adding the following conditions, r 1 /ω 2 . However, because variances are quite small, the values of r 1 and r 2 are almost in (30,150)-this suggests estimate change rates are quite close to zero and lead to no improvement.…”

Section: The Choice Of R 1 and Rmentioning

confidence: 99%

“…The BS can be regarded as the special case of an RPS with two forecast categories [28]. The RPS is particularly appropriate for evaluating probability forecasts of ordered variables [29,30] explained that RPS was the most rational scoring rule of those that have been proposed and used for football outcomes. For a single forecast the RPS is defined as…”

Section: Rank Probability Score (Rps)mentioning

confidence: 99%

A Bayesian In-Play Prediction Model for Association Football Outcomes

2020

View full text Add to dashboard Cite

Point process models have made a significant contribution to the prediction of football association outcomes. It is conventionally the case that defence and attack capabilities have been assumed to be constant during a match and estimated against the average performance of all other teams in history. Drawing upon a Bayesian method, this paper proposes a dynamic strength model which relaxes assumption of the constant teams’ strengths and permits applying in-match performance information to calibrate them. An empirical study demonstrates that although the Bayesian model fails to achieve improvement in goal difference prediction, it registers clear achievements with regard to the prediction of the total number of goals and Win/Draw/Loss outcome prediction. When the Bayesian model bets against the SBOBet bookmaker, one of the most popular gaming companies among Asian handicaps fans, whose odds data were obtained from both the Win/Draw/Loss market and over–under market, it may obtain positive returns; this clearly contrasts with the process model with constant strengths, which fails to win money from the bookmaker.

show abstract

Solving the Problem of Inadequate Scoring Rules for Assessing Probabilistic Football Forecast Models

Cited by 50 publications

References 27 publications

pi-football: A Bayesian network model for forecasting Association Football match outcomes

pi-football: A Bayesian network model for forecasting Association Football match outcomes

Systematic Positive Expected Returns in the UK Fixed Odds Betting Market: An Analysis of the Fink Tank Predictions

A Bayesian In-Play Prediction Model for Association Football Outcomes

Contact Info

Product

Resources

About