ROLE OF STATISTICS AND MACHINE LEARNING IN SPORTS PERFORMANCE AND PREDICTIVE ANALYTICS

Abstract

The integration of statistical analysis and machine learning has transformed modern sports analytics by enabling accurate evaluation and prediction of athlete performance. This study investigates the role of statistics and machine learning in sports performance and predictive analytics using a dataset of 300 athletes generated from key performance-related variables, including training hours, fitness score, sleep duration, diet quality, experience years, injury index, and overall performance score. Descriptive statistics, correlation analysis, and linear regression techniques were employed to identify the relationships between independent variables and athletic performance outcomes. In addition, machine learning algorithms, including Linear Regression and Random Forest models, were implemented to enhance predictive accuracy and evaluate sports performance forecasting capabilities. The findings revealed that training intensity, physical fitness, sleep quality, nutrition, and professional experience positively influenced athlete performance, whereas injury levels negatively affected sports outcomes. Among the predictive models, the Random Forest algorithm demonstrated superior predictive performance and lower prediction error. The study concludes that the integration of statistics and machine learning provides an effective framework for performance optimization, injury prevention, athlete evaluation, and evidence-based decision-making in modern sports science and analytics.