Comment on: Machine Learning for Understanding and Predicting Injuries in Football

The growing intensity and frequency of matches in professional football leagues are related to the increasing physical player load. An incorrect training model results in over- or undertraining, which is related to a raised probability of an injury. This research focuses on predicting non-contact lower body injuries coming from over- or undertraining. The purpose of this analysis was to create decision-making models based on data collected during both training and match, which will enable the preparation of a tool to model the load and report the increased risk of injury for a given player in the upcoming microcycle. For this purpose, three decision-making methods were implemented. Rule-based and fuzzy rule-based methods were prepared based on expert understanding. As a machine learning baseline XGBoost algorithm was considered. Taking into account the dataset used containing parameters related to the external load of the player, it is possible to predict the risk of injury with a certain precision, depending on the method used. The most promising results were achieved by the machine learning method XGBoost algorithm (Precision 92.4%, Recall 96.5%, and F1-score 94.4%).

Digitalization in the field of sport has already been a reality for a number of years. The growing increase in the volume of data that can be acquired on athletes today makes its use possible mainly for performance enhancement and also for injury prevention. We propose in this paper to evaluate the possibility of including Artificial Intelligence (A.I.) through Machine Learning (ML) as a mean for estimating injuries in professional football, by 1) discussing the addition of ML information in the interaction between stakeholders through graph network representations, and 2) presenting the injury risk estimation through two ML techniques adapted to the characteristics of data from players. We first constructed an elementary representation for an athlete and his/her environment, and we then created a complex network of 23 professional football players. We discussed the implication of ML methods for stakeholders such as coaches, players or medical staff. Regarding injury risk estimation, we focused on methods allowing 1) to work with few data and 2) to have a certain level of explainability to avoid the well-known “black box” effect. In particular, we used decision tree and logistic regression methods to predict the occurrence of hamstring injuries in 284 professional footballers for whom baseline data, as well as sprint acceleration mechanical output measurements taken from one football season were available. The results show that the estimation of injury risk is possible to a certain extent, and that the centrality of the technical team is crucial when incorporating such methods in team sports.KeywordsInjury risk estimationMachine learningProfessional footballSports scienceSports medicineSocial networks