Contributo da Análise de Redes Sociais, a partir da Centralidade de Autovetor, no estudo da performance desportiva em Voleibol de Alto Rendimento.

Abstract

The optimization of performance in Sports Games is a complex subject and with multiple methodological ways to reach it. Among them, Social Network Analysis contributes to Game Analysis using mathematical pathways derived from Graph Theory, analysing the interaction between agents in a network. The Interaction Networks, most used in the sports field, analyse the players relationship (units analysis or nodes) through the passes (edges). The present thesis expanded the look given to the analysis units in the Sports Games and considered the direct and indirect relation between the game actions. Thus, the main objective was to verify the behaviour of each game variable in volleyball, thus revealing the real role of each game action in the six Game Complexes; as well as the influence of game standards on the effectivity of high-level women's and men's volleyball. The results suggest that considering game actions as analysis units provides important insights into game ecology. In this sense, important game patterns were found in the six game complexes, where actions performed in non-ideal conditions (such as the Conditions of Distribution, Zone and Attack Tempo) played a central role in most of the complexes analysed. Game actions revealed the same behaviour regardless of the effectivity level considered, indicating that individual ability to resolve game problems shows a greater importance role than the play standard in each complex. In short, the use of non-agency units, through the weighting of direct and indirect connections, provided refined information and the weight of the behaviour of each variable in all Game Complexes, contributing to the understanding of the ecology and dynamics of volleyball. It is suggested that future investigations consider tactical systems, players characteristics, or factors that may impact individual actions, such as situational restrictions of the game (momentum, game status, previous and post-error actions). KEY-WORDS: Analysis of Social Networks; Eigenvector Centrality; Performance Analysis, Game Actions; Volleyball.