Que fatores estão associados ao resultado de um jogo de voleibol? – Análise Estatística da Superliga Masculina 2011/12

Abstract

Objective: To identify which fundamentals associated to the match are able to discriminate the teams that have reached or not the Super-league quarter finals and the winners and losers of the matches. Method: This paper was based on a sampling of Men's Volleyball Super-league 2011/12 matches. To compare the means of the fundamentals performances (attack, serving, blocking, setting, reception and defense) according to the quarter finals (classified vs. unclassified) and the match result (win vs. loss), we used the t Student test. To test the association between the match location and match result, we used the chi-square test. It was applied a logistic regression to assess the significance of a studied fundamental, when controlled by the other variables, over the likelihood of the team to qualify for quarter finals or to win a match. We adopted a significance level of 5% and a confidence interval of 95%. Results: According to the t Student test, the fundamentals attack, serving, setting and defense are significantly associated with the quarter finals. After analyzing the logistic regression, we have that for each increase of one unit in the explanatory variable attack, the chance of ranking increases 13% and for each unit increment in the variable defense, the chances of ranking grow 8.5%. All studied fundamentals, besides match location, are significantly associated with the match result, according to the t Student and chi- square tests. Interpreting the logistic regression model, we have that for each unit increase in the variable attack, the chances of winning increase 30.3%. Regarding the defense, for each one-unit increment, the chance of winning is 20.9% higher. For each extra one-unit in the variable serve, the chance of winning increases 18.2%. For each increment of one unit of the variable block, the chances of victory increase 15.6%. Conclusion: the attack and defense performances showed as predictors of the quarterfinals classification and the combination of the variables attack, defense, serving and blocking was predictor of the match result. The logistic regression models for the quarter finals and match result had a correct overall percentage of 74.2% and 81.6%, respectively.