| Applications of Artificial Intelligence in Sports Talent Identification: Opportunities and Challenges |
| Paper ID : 1746-SPORTCONGRESS |
| Authors |
|
Mehdi Nazemi *1, Leila Jedi2 1Assistant Professor of Sport Management, Department of Applied Humanities, Kashmar Higher Education Institute, Kashmar, Iran, 2M.Sc in Sport Management, Tarbiat Modares University Education Department, Ardabil Province |
| Abstract |
| Introduction: Sports talent identification is a fundamental component in the development of professional athletes. Early detection of potential talent enables more effective training, resource allocation, and long-term performance optimization. Traditional talent identification methods largely depend on coaches’ subjective assessments, which are often limited by human bias, inconsistency, and restricted data analysis capacity (Rein & Memmert, 2016). In recent years, Artificial Intelligence (AI) —particularly machine learning (ML) and computer vision— has emerged as a transformative tool capable of analyzing vast volumes of performance, physiological, and psychological data (Claudino et al., 2019). This paper aims to systematically review the current literature on AI-driven approaches in sports talent identification and to discuss their challenges and opportunities. Methods: This study employed a systematic review approach. Scientific databases were searched for the period 2015–2025 using the keywords "Artificial Intelligence", "Talent Identification", "Sports Analytics", and "Machine Learning". Out of initially retrieved studies, articles that explicitly applied AI or ML algorithms to the process of sports talent identification were included for analysis. Results: Results indicate that algorithms such as Random Forest, Support Vector Machine (SVM), and Convolutional Neural Network (CNN) achieved accuracy rates between 85% and 95% in predicting athletic potential. In sports such as football, basketball, and track and field, wearable sensors and video-based motion analysis were the most frequently used data sources (Bunker & Thabtah, 2019). Hybrid models that integrated physiological, cognitive, and psychological variables demonstrated higher predictive performance (Claudino et al., 2019). However, issues such as limited standardized datasets, age-related variability, and ethical concerns about data privacy remain significant obstacles. Conclusion: Overall, AI has the potential to make the sports talent identification process faster, more accurate, and fairer. To fully harness this potential, the development of comprehensive data infrastructures, transparent ethical frameworks, and educational programs for coaches and analysts is essential. |
| Keywords |
| Artificial Intelligence, Sports Talent Identification, Opportunities, Challenges |
| Status: Abstract Accepted (Poster Presentation) |
