A systematic review of sport-related packages within the R CRAN repository. Issue 2 (April 2023)

• Record Type: Journal Article
• Authors: Casals, Martí; Fernández, José; Martínez, Victor; Lopez, Michael; Langohr, Klaus; Cortés, Jordi
• Is Part Of: International journal of sports science & coaching. Volume 18:Issue 2(2023)
• Page Start: 621
• Page End: 629
• Publication Date: 2023-04
• Subjects: Data analysis -- sport analytics -- statistical software; Coaching (Athletics) -- Periodicals; Sports sciences -- Periodicals; Coaching (Athletics); Sports sciences; Periodicals; 796.077
• DOI: 10.1177/17479541221136238
• Languages: English
• ISSNs: 1747-9541

Abstract

The sports statistics community is rapidly evolving. R is an open-source software in constant development that has gained a lot of popularity within statistical communities and sports analytics. This work presents a systematic review of the available R CRAN sport packages following PRISMA guidelines. We consider all packages active as of 18 February 2021. A total of 81 sport R packages created since 2006 were detected. Of these, 35.9% were authored by an American national, 69.1% include a dataset, 43.2% provide vignettes, and 65.4% have been updated at least once. The sport with the highest representation is basketball (n = 14, 17.3%), followed by soccer (n = 12, 14.8%) and packages related to physical activity (n = 11, 13.6%). There are no sports packages directed solely for the female gender, while 59.3% of packages are focused on professional athletes. Fifty packages (61.7%) are related to the sports performance analysis category, and web scraping (n = 43, 53.1%) was the main functionality in the collected packages. The creation of new R packages in the area of sports could help solve questions and tasks that still remain a challenge in this field, while continuing to help to improve the level of statistical education and computational thinking skills.