Relationships between performance parameters and final ranking in professional volleyball

Statistics in volleyball are very important at professional level, enabling an analysis of team’s and players’ performances in different skills. Many authors already investigated volleyball performance through statistical data but previous studies considered few parameters or included few games in their analysis. Aims is to detect the importance of eight selected parameters on the determination of the ranking at the end of the regular season in the Italian professional leagues. We tested especially the hypothesis that teams’ performance in passing, traditionally considered very important by many coaches, is not so different between successful and unsuccessful teams. All the matches of the Italian Professional Leagues are recorded through the software Data Volley® (Dataproject s.r.l., Bologna, Italy) by scoutmen registering each touch of the ball during the game. These data are available online at the websites www.legavolley.it and www.legavolleyfemminile.it. We downloaded and analyzed the data relative to all games (M:182 and W:132) of the entire male (M) and female (W) A1 regular season 2004-2005. For each team (M:14; W:12) we considered the mean game value of the following parameters per set: Aces (B#), Attack Points (S#), Attack Errors (S=), Attack Stuffed (S/), Blocks (M#), Serve mistakes (B=), Perfect Passes (R#), Pass Errors (R=). A Pearson r correlation coefficient was computed between these variables and team’s points in the final ranking (FP). A hierarchical and non-hierarchical cluster analysis has been carried out to detect the relationships between parameters in the determination of the success of volleyball teams in the entire championship. Final ranking (FP) are very correlated to S# for both W (r=0,93) and M (r=0,82) and also to M# strongly in M (r=0,74), but weakly in W (0,30). FP and B# are more correlated in W than in M (0,79 vs 0,51) and also FP has a negative high correlation with S/, similar in both leagues (W=-0,82; M:-0,78). FP and R= have inverse correlation (r=-0,79 in case of W and r=-0,41 in M), while for R# it is –0,53 in M and –0,45 in W. FP and B= correlations are weak for both the leagues. In hierarchical cluster analysis M shows two initial clusters: S/ and M# is the first one, FP and S# is the second. At a weaker correlation-level, anyway very high, these two clusters are reciprocally linked. B# and R# make a cluster with medium correlation. In W: FP, B# and S# are strongly linked in the same cluster. R= and S/ are linked with the first cluster at a weaker level. The parameter with the highest correlation with final ranking’s points is, for both the leagues, the number of attack points per set. So attack may be considered the most important skill of volleyball. Passing performance does not appear fundamental, but it is important avoiding passing errors. Teams have to train pass receptions to save the ball and to take attacking chances. This numerical analysis can help coaches to better understand the game and to plan their program of training.