Reporting of clustering techniques in sports sciences: a scoping review

Multivariate statistical methods are among the most used ones in sportssciences with clustering methods emerging as prominent unsupervised learningtechniques. This study presents a scoping review of original articlesutilizing clustering techniques in sports sciences, following the PRISMASCRguidelines. A comprehensive search across various databases using theboolean "AND" combination of "clustering" and "sport" yielded 278 articles.Notably, 86.7% of these articles were published within the last 14 years,with a predominant focus (66.2%) on sports performance analysis. The majorityof studies included professional athletes (56.4%), with football/soccer,basketball, and tennis being the most commonly studied sports, representing12.2%, 7.5%, and 2.2% of the selected articles, respectively. Hierarchical clusteringwas the most frequently used method (31.6%), followed by the k-meansalgorithm for partitional clustering. However, the clustering method was notreported in 26.6% of the articles, and 55.0% did not specify the criterion usedfor determining the optimal number of clusters. Moreover, more than 85%of the articles lacked computational details related to data reproducibility.These findings underscore the urgent need for substantial improvement in reportingpractices regarding the methodology, algorithms, criteria for clusteridentification, and software usage in sports science literature.‎