Machine Learning-Based Identification of the Strongest Predictive Variables of Winning and Losing in Belgian Professional Soccer
Source abstract
Abstract
This study aimed to identify the strongest predictive variables of winning and losing in the highest Belgian soccer division. A predictive machine learning model based on a broad range of variables (n = 100) was constructed, using a dataset consisting of 576 games. To avoid multicollinearity and reduce dimensionality, Variance Inflation Factor (threshold of 5) and BorutaShap were respectively applied. A total of 13 variables remained and were used to predict winning or losing using Extreme Gradient Boosting. TreeExplainer was applied to determine feature importance on a global and local level. The model showed an accuracy of 89.6% ± 3.1% (precision: 88.9%; recall: 90.1%, f1-score: 89.5%), correctly classifying 516 out of 576 games. Shots on target from the attacking penalty box showed to be the best predictor. Several physical indicators are amongst the best predictors, as well as contextual variables such as ELO -ratings, added transfers value of the benched players and match location. The results show the added value of the inclusion of a broad spectrum of variables when predicting and evaluating game outcomes. Similar modelling approaches can be used by clubs to identify the strongest predictive variables for their leagues, and evaluate and improve their current quantitative analyses.
Source: crossref · Quality: baseline
Research context
OpenAlex topics: Sports Analytics and Performance · Sports Performance and Training · Sports injuries and prevention
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
Affiliations
Department of Movement and Sports Sciences, Ghent University, 9000 Ghent, Belgium
Center of Sports Medicine, Ghent University Hospital, 9000 Ghent, Belgium
Department of Electronics and Information Systems, Research Group IDLab, Ghent University—IMEC, 9052 Ghent, Belgium
References (52)
Show source-provided references
1. On the search for reliable performance indicators in game sports10.1080/24748668.2007.11868388 ↗
2. The use of performance indicators in performance analysis10.1080/026404102320675602 ↗
3. Sports performance judgments from a social cognitive perspective10.1016/j.psychsport.2006.03.007 ↗
4. Causal Attributions of Winners and Losers in Individual Competitive Sports: Toward a Reformulation of the Self-Serving Bias10.1123/jsp.6.2.184 ↗
5. Dealing With Randomness in Match Outcomes: How to Rethink Performance Evaluation in European Club Football Using Expected Goals10.1177/1527002519897962 ↗
6. The Use of Match Statistics that Discriminate Between Successful and Unsuccessful Soccer Teams10.2478/v10078-012-0015-7 ↗
7. The possession game? A comparative analysis of ball retention and team success in European and international football, 2007–201010.1080/02640414.2012.727455 ↗
8. Performance in Team Sports: Identifying the Keys to Success in Soccer10.2478/v10078-010-0035-0 ↗
9. Key team physical and technical performance indicators indicative of team quality in the soccer Chinese super league10.1080/15438627.2018.1431539 ↗
10. Statistical Analysis for the First Bundesliga in the Current Soccer Season
11. Performance indicators that predict success in an English Professional League One Soccer Team10.1080/24748668.2014.11868767 ↗
12. Success factors in football: An analysis of the German Bundesliga10.1080/24748668.2020.1726157 ↗
13. Analysis of entries into the penalty area as a performance indicator in soccer10.1080/17461391.2011.606834 ↗
14. Goes, F.R., Kempe, M., and Lemmink, K. (2019). Predicting match outcome in professional Dutch football using tactical performance metrics computed from position tracking data. Mathsport International Conference Proceeding, Propobos Publications.10.29007/4jjb ↗
15. Bialkowski, A., Lucey, P., Carr, P., Yue, Y., Sridharan, S., and Matthews, I. (2014, January 14–17). Large-scale analysis of soccer matches using spatiotemporal tracking data. Proceedings of the IEEE International Conference on Data Mining (ICDM), Shenzhen, China.10.1109/ICDM.2014.133 ↗
16. Goes, F.R., Meerhoff, R.L.A., Bueno, M.J., Rodrigues, D.M., Moura, F.A., Brink, M.S., Elferink-Gemser, M.T., Knobbe, A.J., Cunha, S.A., and Lemmink, K.A. (2020). Unlocking the potential of big data to support tactical performance analysis in professional soccer: A systematic review. Eur. J. Sport Sci., 1–16.10.1080/17461391.2020.1747552 ↗
17. State of the Art of Sports Data Visualization10.1111/cgf.13447 ↗
18. Points of Significance: Statistics versus machine learning10.1038/nmeth.4642 ↗
19. Linke, D., Link, D., and Lames, M. (2018). Validation of electronic performance and tracking systems EPTS under field conditions. PLoS ONE, 13.10.1371/journal.pone.0199519 ↗
20. High-Intensity Activity Profiles of Elite Soccer Players at Different Performance Levels10.1519/JSC.0b013e3181aeb1b3 ↗
21. The reliability of ProZone MatchViewer: A videobased technical performance analysis system10.1080/24748668.2007.11868415 ↗
22. The UEFA Champions League seeding is not strategy-proof since the 2015/16 season10.1007/s10479-020-03637-1 ↗
23. The Effect of Seeding on Tournament Outcomes: Evidence From a Regression-Discontinuity Design10.1177/1527002520955212 ↗
24. Hastie, T., Tibshirani, R., and Friedman, J. (2009). The Elements of Statistical Learning: Data Mining, Inference, and Prediction, Springer Science & Business Media. Number 2.
25. From local explanations to global understanding with explainable AI for trees10.1038/s42256-019-0138-9 ↗
26. Molnar, C. (2020, November 15). Interpretable Machine Learning: A Guide for Making Black Box Models Explainable. Available online: https://christophm.github.io/interpretable-ml-book/.
27. The Open International Soccer Database for machine learning10.1007/s10994-018-5726-0 ↗
28. Hassan, A., Akl, A.R., Hassan, I., and Sunderl, C. (2020). Predicting wins, losses and attributes’ sensitivities in the soccer world cup 2018 using neural network analysis. Sensors, 20.10.3390/s20113213 ↗
29. Chance involvement in goal scoring in football—An empirical approach10.1007/s12662-018-0518-z ↗
30. Deserving to be lucky: Reflections on the role of luck and desert in sports10.1080/00948705.2007.9714706 ↗
31. Interpreting physical performance in professional soccer match-play: Should we be more pragmatic in our approach?10.1007/s40279-013-0055-8 ↗
32. Changes in Acceleration and Deceleration Capacity Throughout Professional Soccer Match-Play10.1519/JSC.0000000000000805 ↗
33. Straight sprinting is the most frequent action in goal situations in professional football10.1080/02640414.2012.665940 ↗
34. Match outcome and running performance in different intensity ranges among elite soccer players10.5114/biolsport.2018.74196 ↗
35. Comparison of physical and technical performance in European soccer match-play: FA Premier League and La Liga10.1080/17461391.2010.481334 ↗
36. Factors affecting match running performance of elite soccer players: Shedding some light on the complexity10.1123/ijspp.2015-0029 ↗
37. The effects of situational variables on distance covered at various speeds in elite soccer10.1080/17461390903273994 ↗
38. Are Current Physical Match Performance Metrics in Elite Soccer Fit for Purpose or is the Adoption of an Integrated Approach Needed?10.1123/ijspp.2017-0433 ↗
39. Influence of contextual variables on styles of play in soccer10.1080/24748668.2018.1479925 ↗
40. Match statistics related to winning in the group stage of 2014 Brazil FIFA World Cup10.1080/02640414.2015.1022578 ↗
41. Vecer, J. (2021, March 05). Crossing in Soccer has a Strong Negative Impact on Scoring: Evidence from the English Premier League the German Bundesliga and the World Cup 2014. Technical Report. 30 September 2014. Available online: https://ssrn.com/abstract=2225728.
42. Dolores: A model that predicts football match outcomes from all over the world10.1007/s10994-018-5703-7 ↗
43. Hucaljuk, J., and Rakipović, A. (2011, January 23–27). Predicting football scores using machine learning techniques. Proceedings of the 2011-34th International Convention on Information and Communication Technology, Electronics and Microelectronics, Opatija, Croatia.
44. Stübinger, J., Mangold, B., and Knoll, J. (2020). Machine learning in football betting: Prediction of match results based on player characteristics. Appl. Sci., 10.10.3390/app10010046 ↗
45. Current Approaches to Tactical Performance Analyses in Soccer Using Position Data10.1007/s40279-016-0562-5 ↗
46. Modeling the Prediction of the Session Rating of Perceived Exertion in Soccer: Unraveling the Puzzle of Predictive Indicators
47. Relationships Between Training Load Indicators and Training Outcomes in Professional Soccer10.1007/s40279-016-0591-0 ↗
48. A Machine Learning Approach to Assess Injury Risk in Elite Youth Football Players10.1249/MSS.0000000000002305 ↗
49. Physiology of Soccer10.2165/00007256-200535060-00004 ↗
50. Physiological assessment of aerobic training in soccer10.1080/02640410400021278 ↗
51. Understanding injury mechanisms: A key component of preventing injuries in sport10.1136/bjsm.2005.018341 ↗
52. The influence of match location, quality of opposition, and match status on possession strategies in professional association football10.1080/02640410903131681 ↗