Performance prediction of elite athletes: a systematic review of current methods

With the globalization and commercialization of athletic sport, accurately predicting the performance of elite athletes has become particularly crucial for training planning, athlete selection, and competition strategy formulation. This study systematically analyzes and summarizes the theoretical achievements in the field of predicting the athletic performance of elite athletes both in Mainland China and in European and American regions, revealing the current research status, existing problems, and future development directions in this field. This study is based on the CNKI and Web of Science databases, and employs bibliometric and knowledge graph analysis methods to conduct in-depth analysis of representative literature with high centrality. Research has found that research in European and American regions mainly focuses on high commercial value sports such as basketball and football, while in Mainland China research mainly focuses on Olympic events; In terms of data sources, research in European and American regions tends to use publicly available data, while research in Mainland China relies more on internal data obtained from long-term tracking of athletes or teams; In terms of prediction methods, machine learning algorithms based on computer technology, especially neural network algorithms, have shown significant advantages in prediction accuracy. This study emphasizes the core role of selecting key indicators and collecting and analyzing data in improving prediction accuracy. At the same time, it points out the limitations of existing performance prediction methods in terms of accuracy and applicability, and proposes the necessity of constructing more accurate and personalized prediction models by combining individual differences in athletes, training load, recovery status, and psychological factors. In addition, this article also explores the promoting effect of interdisciplinary integration on the innovation of performance prediction theory and methods, hoping to provide theoretical reference for research on performance prediction in the field of sports training in China.