Chapter 20: Further Reading

Foundational Papers

Dixon, M. J., & Coles, S. G. (1997). Modelling Association Football Scores and Inefficiencies in the Football Betting Market. Journal of the Royal Statistical Society: Series C, 46(2), 265--280. The seminal paper introducing the Poisson regression framework for match prediction with the low-score correction factor. Every serious match prediction model is either based on or benchmarked against this work.
Maher, M. J. (1982). Modelling Association Football Scores. Statistica Neerlandica, 36(3), 109--118. The predecessor to Dixon-Coles that established the independent Poisson model with team-specific attack and defense parameters. Understanding Maher's model is essential for appreciating the Dixon-Coles contribution.
Karlis, D., & Ntzoufras, I. (2003). Analysis of Sports Data by Using Bivariate Poisson Models. The Statistician, 52(3), 381--393. Introduces the bivariate Poisson model as an alternative to the Dixon-Coles correction, allowing direct modeling of goal correlation via a covariance parameter.
Rue, H., & Salvesen, O. (2000). Prediction and Retrospective Analysis of Soccer Matches in a League. The Statistician, 49(3), 399--418. A Bayesian treatment of the Poisson match model with time-varying team strengths, using Markov chain Monte Carlo for inference.

Dendir, S. (2016). When Do Soccer Players Peak? A Note. Journal of Sports Analytics, 2(2), 89--105. A comprehensive study of aging curves across major European leagues using the delta method, with position-specific peak age estimates.
Wakelam, E., Maymin, P., & Khoshgoftaar, T. (2022). Player Performance Prediction in Football. Journal of Sports Analytics, 8(3), 207--232. A survey of machine learning methods for player performance forecasting, comparing random forests, neural networks, and ensemble approaches on large-scale event data.
Bonhomme, S., & Sauder, R. (2011). Ballers and Shot-Callers: The Effect of Experience on NBA Free-Throw Shooting. Working paper. While focused on basketball, the methodology for mixed-effects aging curves with random intercepts and slopes translates directly to soccer performance modeling.

Rossi, A., Pappalardo, L., Cintia, P., Iaia, F. M., Fernandez, J., & Medina, D. (2018). Effective Injury Forecasting in Soccer with GPS Training Data and Machine Learning. PLOS ONE, 13(7), e0201264. One of the first large-scale studies applying gradient boosting to GPS-derived training load data for injury prediction in professional soccer.
Gabbett, T. J. (2016). The Training-Injury Prevention Paradox: Should Athletes Be Training Smarter and Harder? British Journal of Sports Medicine, 50(5), 273--280. Introduces the concept of the ACWR sweet spot and danger zone, demonstrating that both undertraining and overtraining increase injury risk.

Gneiting, T., & Raftery, A. E. (2007). Strictly Proper Scoring Rules, Prediction, and Estimation. Journal of the American Statistical Association, 102(477), 359--378. The definitive treatment of proper scoring rules, explaining why log-loss and Brier score are appropriate for evaluating probabilistic forecasts.
Niculescu-Mizil, A., & Caruana, R. (2005). Predicting Good Probabilities with Supervised Learning. ICML. Demonstrates that many machine learning models produce poorly calibrated probabilities and introduces Platt scaling and isotonic regression as post-hoc calibration methods.

Sumpter, D. (2016). Soccermatics: Mathematical Adventures in the Beautiful Game. Bloomsbury. Accessible introduction covering Poisson models, betting markets, and the mathematics of match prediction.
Anderson, C., & Sally, D. (2013). The Numbers Game: Why Everything You Know About Football Is Wrong. Penguin. Covers the role of luck and regression to the mean in soccer, providing intuitive explanations of key predictive modeling concepts.
Kuper, S., & Szymanski, S. (2018). Soccernomics. 5th ed. Nation Books. Explores the economics of player transfers, including the predictability (or lack thereof) of transfer success.
Hyndman, R. J., & Athanasopoulos, G. (2021). Forecasting: Principles and Practice. 3rd ed. OTexts. The standard reference for time series forecasting methods, including exponential smoothing and ARIMA models used for player metric projection.
Gelman, A., Carlin, J. B., Stern, H. S., Dunson, D. B., Vehtari, A., & Rubin, D. B. (2013). Bayesian Data Analysis. 3rd ed. CRC Press. The authoritative textbook on Bayesian methods. Essential for understanding posterior predictive distributions and hierarchical models applied to soccer.

FiveThirtyEight Soccer SPI Model --- fivethirtyeight.com/methodology/how-our-club-soccer-predictions-work Documentation of FiveThirtyEight's public Soccer Power Index model, which combines Poisson regression with Elo-style ratings. A useful benchmark for custom match prediction models.
footballmodelling.net --- A community resource for football prediction models, including implementations of Dixon-Coles and related methods.
Friends of Tracking Data Science --- github.com/Friends-of-Tracking-Data-Science Open-source implementations of various soccer analytics models, including match prediction and player evaluation tools.

David Sumpter, "Is Maths Killing Football?" --- TEDx talk. An accessible overview of how mathematical models are used in football, including match prediction and player evaluation.
Jan Vecer, "Mathematical Analysis of Soccer" --- Columbia University course. Covers Poisson regression, Elo ratings, and betting market efficiency.

Danny Page, "Expected Goals and Match Prediction" --- A practical walkthrough of building an xG-based match prediction model in Python.
James Yorke (StatsBomb), "League Adjustment Factors" --- Discusses the challenges of comparing player performance across leagues and presents practical estimation approaches.
Ben Torvaney, "Dixon-Coles in Python" --- A step-by-step tutorial for implementing the Dixon-Coles model from scratch, including parameter estimation and prediction.

For cutting-edge research on predictive modeling in soccer, monitor: