Artificial intelligence in football talent identification: A systematic review

Vollständiger Abstract

Worum geht es in dieser Arbeit?

Objectives To synthesize current evidence on the use of artificial intelligence (AI) in football talent identification. Design Systematic review. Method This review followed PRISMA 2020 guidelines and was registered in PROSPERO. PubMed, Scopus, Web of Science, IEEE Xplore, and ACM Digital Library were searched for studies published between 2016 and 2025. Eligible studies applied AI-related methods to football talent identification or player evaluation. After duplicate removal and screening, 24 studies were included in the qualitative synthesis. Results Twenty-four studies were included. Supervised machine learning was used in 19 studies (79%), deep learning in 14 (58%), ensemble learning in 11 (46%), and other specialized approaches in 5 (21%). Technical indicators were most frequently assessed, followed by physiological, psychological, and tactical variables. Although several models showed promising performance, comparability was limited by differences in samples, features, outcomes, and validation methods. Conclusions AI-based approaches may support multidimensional football talent identification, but current evidence is limited by uneven data quality, limited longitudinal validation, interpretability concerns, and ethical issues related to privacy and bias.