AI-Enabled System to Predict Match Results and Potential Subsequent Injuries for Players in the QSL (Qatar Stars League)

As football (soccer) is one of the most popular sports worldwide, winning football matches is becoming an essential aspect of football clubs. We focused in this dissertation on analyzing players’ performance in Qatar Stars League (QSL) using Machine Learning (ML) and Deep Learning (DL), and analyzed the subsequent injury of these players using Markov Chains. In the ML approach, the collective performance of the players in key playing positions was analyzed to understand their effectiveness in winning games and logistic regression-based model was considered the best performing model, with more than 80% accuracy. The pro- posed ML model identified shots on target by forwarders, distance covered by forwarders and midfielders at very high speed, and successful passes, that can be considered as the most effective performance features for winning football matches in QSL. We also showed that players’ performance features from the last three seasons would provide sufficient discriminative power to the proposed ML model to predict the match-winner. In the DL model, we proposed SoccerNet, a Gated Recurrent Unit (GRU)-based deep learning-based model to predict match winners with over 80% accuracy based on 15 mins interval match segment. We analyzed players’ performance at different positions on the field at different stages of the match. Our analysis suggests that the last 15–30 minutes of match segments of the matches from QSL have a more significant impact on the match result than other match segments.

Finally, we analyzed the association between injuries and their subsequent injuries for foot- ball players. Our analysis showed that the most common injuries are reinjuries from same body part and same nature. The most common nature is Muscle and the top three body parts are Thigh, Lower Leg, and Hip and Groin. The most common categorization in subsequent injuries is Hamstring muscle injury where it’s subsequent injury is of same categorization. We believe the present dissertation will ultimately support the players and coaching staff to improve players’ performance, increase the chance of winning a match and potentially consider preventing subsequent injuries.