Previsão de Resultados na Premier League: Comparação Sistemática de Modelos de Aprendizado de Máquina
Resumo
Este trabalho compara nove modelos de aprendizado de máquina para a previsão de resultados da Premier League inglesa (16 temporadas, 2010/11–2025/26; N = 6.080 partidas), sob validação temporal walk-forward, teste independente e bootstrap pareado em dois esquemas de reamostragem, com Log Loss como métrica principal. O protocolo detecta vazamento temporal nas features e na calibração. O Poisson superou com significância estatística os oito modelos discriminativos calibrados e duas das três especificações por força de equipe; dele permanecem indistinguíveis apenas a Regressão Logística ajustada sem calibração e a variante Dixon-Coles com decaimento temporal. No teste final obteve Log Loss de 1,0099 e acurácia de 50,92%, sendo escolhido por parcimônia e interpretabilidade.
Palavras-chave:
previsão de resultados de futebol, Premier League, aprendizado de máquina, modelo de Poisson, validação temporal, calibração de probabilidades, Log Loss
Referências
Bentéjac, C., Csörgő, A., and Martínez-Muñoz, G. (2021). A comparative analysis of gradient boosting algorithms. Artificial Intelligence Review, 54(3):1937–1967.
Bunker, R. and Susnjak, T. (2022). The application of machine learning techniques for predicting match results in team sport: a review. Journal of Artificial Intelligence Research, 73:1285–1322.
Chen, T. and Guestrin, C. (2016). XGBoost: A scalable tree boosting system. In Proc. 22nd ACM SIGKDD Int. Conf. on Knowledge Discovery and Data Mining (KDD ’16), pages 785–794. ACM.
Demšar, J. (2006). Statistical comparisons of classifiers over multiple data sets. Journal of Machine Learning Research, 7:1–30.
Dixon, M. J. and Coles, S. G. (1997). Modelling association football scores and inefficiencies in the football betting market. Journal of the Royal Statistical Society: Series C (Applied Statistics), 46(2):265–280.
Efron, B. and Tibshirani, R. J. (1994). An Introduction to the Bootstrap. Chapman and Hall/CRC, 1st edition.
Fischer, M. and Heuer, A. (2024). Match predictions in soccer: machine learning vs. Poisson approaches. arXiv preprint.
Galekwa, R. M. et al. (2024). A systematic review of machine learning in sports betting: techniques, challenges, and future directions. arXiv preprint.
Hubáček, O., Šourek, G., and Železný, F. (2022). Forty years of score-based soccer match outcome prediction: an experimental review. IMA Journal of Management Mathematics, 33(1):1–18.
Hyndman, R. J. and Athanasopoulos, G. (2021). Forecasting: principles and practice. OTexts, Melbourne, 3rd edition.
Julian, R., Page, R. M., and Harper, L. D. (2021). The effect of fixture congestion on performance during professional male soccer match-play: A systematic critical review with meta-analysis. Sports Medicine, 51(2):255–273.
Kapoor, S. and Narayanan, A. (2023). Leakage and the reproducibility crisis in machine learning-based science. Patterns, 4(9):100804.
Ke, G. et al. (2017). LightGBM: A highly efficient gradient boosting decision tree. In Advances in Neural Inf. Proc. Systems (NeurIPS), volume 30, pages 3146–3154.
Nguyen, Q. (2021). Poisson modeling and predicting English Premier League goal scoring. The New England Journal of Statistics in Data Science, 1(1):95–101.
Penn, M. J. and Donnelly, C. A. (2022). Analysis of a double Poisson model for predicting football results in Euro 2020. PLOS ONE, 17(5):e0268511.
Prokhorenkova, L. et al. (2018). CatBoost: unbiased boosting with categorical features. In Advances in Neural Inf. Proc. Systems (NeurIPS), volume 31, pages 6638–6648.
Silva Filho, T. et al. (2023). Classifier calibration: a survey on how to assess and improve predicted class probabilities. Machine Learning, 112:3211–3260.
Wheatcroft, E. (2021). Evaluating probabilistic forecasts of football matches: the case against the ranked probability score. Journal of Quantitative Analysis in Sports, 17(4):273–287.
Bunker, R. and Susnjak, T. (2022). The application of machine learning techniques for predicting match results in team sport: a review. Journal of Artificial Intelligence Research, 73:1285–1322.
Chen, T. and Guestrin, C. (2016). XGBoost: A scalable tree boosting system. In Proc. 22nd ACM SIGKDD Int. Conf. on Knowledge Discovery and Data Mining (KDD ’16), pages 785–794. ACM.
Demšar, J. (2006). Statistical comparisons of classifiers over multiple data sets. Journal of Machine Learning Research, 7:1–30.
Dixon, M. J. and Coles, S. G. (1997). Modelling association football scores and inefficiencies in the football betting market. Journal of the Royal Statistical Society: Series C (Applied Statistics), 46(2):265–280.
Efron, B. and Tibshirani, R. J. (1994). An Introduction to the Bootstrap. Chapman and Hall/CRC, 1st edition.
Fischer, M. and Heuer, A. (2024). Match predictions in soccer: machine learning vs. Poisson approaches. arXiv preprint.
Galekwa, R. M. et al. (2024). A systematic review of machine learning in sports betting: techniques, challenges, and future directions. arXiv preprint.
Hubáček, O., Šourek, G., and Železný, F. (2022). Forty years of score-based soccer match outcome prediction: an experimental review. IMA Journal of Management Mathematics, 33(1):1–18.
Hyndman, R. J. and Athanasopoulos, G. (2021). Forecasting: principles and practice. OTexts, Melbourne, 3rd edition.
Julian, R., Page, R. M., and Harper, L. D. (2021). The effect of fixture congestion on performance during professional male soccer match-play: A systematic critical review with meta-analysis. Sports Medicine, 51(2):255–273.
Kapoor, S. and Narayanan, A. (2023). Leakage and the reproducibility crisis in machine learning-based science. Patterns, 4(9):100804.
Ke, G. et al. (2017). LightGBM: A highly efficient gradient boosting decision tree. In Advances in Neural Inf. Proc. Systems (NeurIPS), volume 30, pages 3146–3154.
Nguyen, Q. (2021). Poisson modeling and predicting English Premier League goal scoring. The New England Journal of Statistics in Data Science, 1(1):95–101.
Penn, M. J. and Donnelly, C. A. (2022). Analysis of a double Poisson model for predicting football results in Euro 2020. PLOS ONE, 17(5):e0268511.
Prokhorenkova, L. et al. (2018). CatBoost: unbiased boosting with categorical features. In Advances in Neural Inf. Proc. Systems (NeurIPS), volume 31, pages 6638–6648.
Silva Filho, T. et al. (2023). Classifier calibration: a survey on how to assess and improve predicted class probabilities. Machine Learning, 112:3211–3260.
Wheatcroft, E. (2021). Evaluating probabilistic forecasts of football matches: the case against the ranked probability score. Journal of Quantitative Analysis in Sports, 17(4):273–287.
Publicado
23/09/2026
Como Citar
MOTA, Murilo Pedral; SILVA, Wilia Francisco da; ARAUJO, Victor Flávio de Andrade.
Previsão de Resultados na Premier League: Comparação Sistemática de Modelos de Aprendizado de Máquina. In: ESCOLA REGIONAL DE COMPUTAÇÃO BAHIA, ALAGOAS E SERGIPE (ERBASE), 26. , 2026, Maragogi/AL.
Anais [...].
Porto Alegre: Sociedade Brasileira de Computação,
2026
.
p. 41-50.
DOI: https://doi.org/10.5753/erbase.2026.32105.
