其他摘要：Currently, the development and competitive level reached by the rugby game has been demanding performance levels to reach a higher standard. In this sense, the acting of the team in training or competition situation should be analyzed to the detail. The analysis of the game has been little explored to characterize, to monitor and to evaluate the performance in rugby games. The aim of the present study was identifying the variables of actions of the game and result that best discriminate between winning and losing teams. Likewise we intended to identify these tendencies for the games that the end result was more balanced, unbalanced and very unbalanced. This type of analysis will be accomplished independently for an IRB and Super 12 group. The data used in the present study were obtained starting from two bases: 1) Rugby Stats Fair Play Sports Analysis Systems and 2) Rugby Match Analysis and Statistics. The sample was constituted for 342 rugby games. For the effect we formed two groups: (i) Group IRB constituted by 135 belonging games the: Rugby World Cup (RWC) of 2003; Rugby World Cup Sub 21 of 2005; Six Nations of 2004, 2005 and 2006 and tournament of the Tri Nations of 2004 and 2005 and (ii) Group Super 12 constituted by 207 games belonging to the tournaments of the Super 12 of 2003, 2004 and 2005. The games was contain according to their final score difference, trying to investigate that way tend to associate amongst themselves in terms of difference in end result. The variables that were analyzed were subdivided in variables of actions of the game (e.g. passes, penalties conceded, turnovers) and result variables (e.g. tries, conversions, penalty kicks). The statistical procedures i) initial explore analysis, ii) cluster analysis and iii) discriminante analysis, allowed to associate games to each other, having been defined by us as grouping criterion the difference in the game final score. We separated and classified groups of data with the aim of finding a linear function that it defines the maximum separation among groups that is, to evidence which the most powerful variables in the classification of the subjects in the 3 clusters groups. The results allowed finding that the balanced games happened in larger number and the largest contribution for the game final score occurred of the tries scored. The winners in the balanced games had more tackles made, they made fewer mistakes in the use of the ball and they varied more your game. In unbalanced games, the winners in the group IRB, they won more balls in the lineouts and they were more effective in the defence (less tackles missed). These variables discriminate the groups, IRB (SC=-0.43) and (SC=-0.42) and Super 12 (SC=-0.34) and (SC=-0.38). In the very unbalanced games the winners they have more penalties, more penalty conceded and they missed more tackles. The differences in groups (IRB and Super12) it occurred for the values of tries scored and conversions, IRB (SC=0.53) and (SC=0.46) and Super 12 (SC=0.72) and (SC=0.57). To understand the reasons that contribute to the success or failure of the teams in games balanced, unbalanced and very unbalanced, it allows turning the most specific training and it contributes decisively to the improvement of the sport performances.
关键词：Rugby; game analysis; categories of games; discriminante analysis
