Análise preditiva e fatores determinantes de sucesso para o voleibol - estudo de caso VNL Feminina

Abstract

The substantial increase in the volume of sports data has driven the use of Artificial Intelligence techniques, making Machine Learning an essential tool for performance analysis and decision-making in modern sports. In this context, this work aimed to develop and evaluate supervised Machine Learning models capable of predicting the results (win or loss) of women’s national volleyball team matches. The study used data from the Volleyball Nations League (VNL), covering the seasons from 2021 to 2025, extracted via web scraping from the official Volleyball World platform. The methodology involved the stages of data collection, cleaning, and feature engineering focusing on collective metrics and technical differentials, followed by the application and comparison of the Random Forest and XGBoost algorithms. For validation, metrics such as accuracy, precision, recall, and F1-score were used, with the models achieving accuracy between 77.17% and 79.35%. The results allowed not only predicting match outcomes consistently but also identifying that ranking variables and the blocking fundamental exert the greatest influence as determining factors for the teams’ success.