The aim of this study was to analyse the variance of network centrality levels between different competitive levels and tactical positions during attacking moments. Twenty-four volleyball players (U12 – 12 players, 11.7 ± 0.4 years old, 1.9 ± 0.9 years of practice; Amateurs – 12 players, 24.9 ± 4.2 years old, 12.6 ± 2.8 years of practice) voluntarily participated in this study. The multivariate MANOVA revealed no statistical differences in the independent variable of competitive level (λ = 0.001; F(3,22) = 0.001; p-value = 1.000; η^2 = 0.001; Very Small Effect Size). Statistical differences were found in the independent variables of tactical position (λ = 1.015; F(15,72) = 2.454; p-value = 0.006; η^2 = 0.338; Moderate Effect Size). The post hoc values revealed that position 3 is the prominent player during attacking moments in both competitive levels and that position 5 reveals the smallest centrality levels in attack.

Keywords: Graph Theory; Adjacency Matrix; Match analysis; Volleyball; Attack; Network; Metrics.