Using multiple machine learning algorithms to classify elite and sub-elite goalkeepers in professional men\u2019s football

Mikael Jamil,Saumya Mehta,Marco Beato,Ashwin Phatak,Mark Connor,Daniel Memmert

doi:10.1038/s41598-021-01187-5

Mikael Jamil, Saumya Mehta + Show 4 more

Open Access

https://doi.org/10.1038/s41598-021-01187-5

Copy DOI

Abstract

This study applied multiple machine learning algorithms to classify the performance levels of professional goalkeepers (GK). Technical performances of GK’s competing in the elite divisions of England, Spain, Germany, and France were analysed in order to determine which factors distinguish elite GK’s from sub-elite GK’s. A total of (n = 14,671) player-match observations were analysed via multiple machine learning algorithms (MLA); Logistic Regressions (LR), Gradient Boosting Classifiers (GBC) and Random Forest Classifiers (RFC). The results revealed 15 common features across the three MLA’s pertaining to the actions of passing and distribution, distinguished goalkeepers performing at the elite level from those that do not. Specifically, short distribution, passing the ball successfully, receiving passes successfully, and keeping clean sheets were all revealed to be common traits of GK’s performing at the elite level. Moderate to high accuracy was reported across all the MLA’s for the training data, LR (0.7), RFC (0.82) and GBC (0.71) and testing data, LR (0.67), RFC (0.66) and GBC (0.66). Ultimately, the results discovered in this study suggest that a GK’s ability with their feet and not necessarily their hands are what distinguishes the elite GK’s from the sub-elite.

Highlights

This study applied multiple machine learning algorithms to classify the performance levels of professional goalkeepers (GK)
Significant differences were discovered for F1 when comparing Logistic Regressions (LR) with RF (p = 0.042) and when comparing LR with Gradient Boosting Classifiers (GBC) (p = 0.034)
Significant differences were discovered for accuracy when comparing LR with GBC (p = 0.032)

Summary

Introduction

This study applied multiple machine learning algorithms to classify the performance levels of professional goalkeepers (GK). Wearable technology[13] and semi-automatic and automatic tracking s ystems[14,15] are partly responsible for this surge in performance data available for analysis This increase in data availability has allowed practitioners to move away from the historical reliance on the subjective opinions and instincts of experienced former professionals (with generally high error rates), towards more accurate and reliable statistical analysis[16]. In terms of identifying informative performance indicators, the position of goalkeeper (GK) in football has been frequently overlooked in previous r esearch[21] This is somewhat surprising, considering the goalkeeper is the most specialised position in a football team[22] and their actions are considered to have a significant bearing on final match o utcomes[23]. In a recently published systematic review of 70 Talent Identification focussed studies on football, the authors stressed how goalkeepers were frequently overlooked in their reviewed studies[24]

Objectives

Methods

Results

Discussion

Conclusion