STATISTICAL MODELS FOR ANALYSING TRAINING FACTORS AND PREDICTING SPORTS RESULTS

Abstract

The article examines the problem of assessing the impact of training factors on sports results using statistical methods of data analysis. The main objective of the work is to develop and test a methodology that includes: conducting a correlation analysis to select the most significant factors that influence sports results; building a multiple linear regression model with parameter estimation using the least squares method; checking the statistical significance of regression coefficients and the adequacy of the constructed model using Student’s and Fisher-Snedecor criteria, respectively. To demonstrate the methodology, an experimental data set of 50 observations was used, which includes the following training factors: amount of water consumed, duration of sleep, number of sessions per week, duration of one session, training location, time index. The target variable Y is the time taken to cover a distance of 1000 m. The practical implementation of the proposed methodology is carried out in the form of a web application with client-server architecture. The user enters data on factors and results in the form of a table, after which the system automatically performs statistical analysis – selects significant factors, evaluates model parameters, checks their statistical reliability, and calculates a forecast for a specified future date. The proposed approach provides the opportunity to perform a personalized analysis of the training process.The use of statistical methods makes it possible to improve the validity of management decisions in sports training and contributes to achieving more stable and predictable results.