The Application of Machine Learning Techniques for Predicting Match Results in Team Sport: A Review

Abstract


 
 
 Predicting the results of matches in sport is a challenging and interesting task. In this paper, we review a selection of studies from 1996 to 2019 that used machine learning for predicting match results in team sport. Considering both invasion sports and striking/fielding sports, we discuss commonly applied machine learning algorithms, as well as common approaches related to data and evaluation. Our study considers accuracies that have been achieved across different sports, and explores whether evidence exists to support the notion that outcomes of some sports may be inherently more difficult to predict. We also uncover common themes of future research directions and propose recommendations for future researchers. Although there remains a lack of benchmark datasets (apart from in soccer), and the differences between sports, datasets and features makes between-study comparisons difficult, as we discuss, it is possible to evaluate accuracy performance in other ways. Artificial Neural Networks were commonly applied in early studies, however, our findings suggest that a range of models should instead be compared. Selecting and engineering an appropriate feature set appears to be more important than having a large number of instances. For feature selection, we see potential for greater inter-disciplinary collaboration between sport performance analysis, a sub-discipline of sport science, and machine learning.
 
 

Sports analytics in the NFL: classifying the winner of the superbowl

“…Moreover, decision tree algorithms are more appealing because they are simpler and easier to implement. Furthermore, using other machine learning algorithms is among the top recommendations for future research in the sports analytics field (Bunker & Susnjak, 2022 ). Therefore, we decided to compare the classification performance of C4.5, ANNs and RF.…”

“…Therefore, there is a need for more accurate models (Haghighat et al, 2013 ; Bunker & Thabtah, 2019 ). Moreover, even though Artificial Neural Networks (ANNs) are the most commonly used machine learning algorithm in sports analytics, they do not consistently perform better than other techniques (Bunker & Susnjak, 2022 ). In fact, decision tree algorithms are the second most commonly used technique, but they are more appealing than ANNs because they do not generate black-box models which makes them more likely to be implemented (Bunker & Susnjak, 2022 ).…”

“…Moreover, even though Artificial Neural Networks (ANNs) are the most commonly used machine learning algorithm in sports analytics, they do not consistently perform better than other techniques (Bunker & Susnjak, 2022 ). In fact, decision tree algorithms are the second most commonly used technique, but they are more appealing than ANNs because they do not generate black-box models which makes them more likely to be implemented (Bunker & Susnjak, 2022 ). In addition, according to Delen et al ( 2012 ), classification and regression tree (CART) models are preferred over other machine learning techniques because they are easier to understand and can be easily integrated into a decision support system.…”

“…In addition, according to Delen et al ( 2012 ), classification and regression tree (CART) models are preferred over other machine learning techniques because they are easier to understand and can be easily integrated into a decision support system. Furthermore, the highest prediction accuracy in American Football was achieved using CART (Bunker & Susnjak, 2022 ).…”

Using Machine Learning Algorithms for Jumping Distance Prediction of Male Long Jumpers

“…As the studies indicated, it was clear that computers and especially machine learning (artificial intelligence) techniques could be used at many points that require experience from the choice of the athletes to the training load and the estimation of their degrees. Today, it is seen that ML techniques are used in many areas of sports, from predicting results in team sports (Bunker and Susnjak, 2022), to athlete health and injury prevention (Eetvelde et al, 2021). Despite the popularity of ML techniques in sports sciences recently, any study has been found in which ML techniques are used for prediction and modeling in the field of the long jump.…”