Match analysis of the 2006 Super 14 Rugby Union tournament

The aims of this study were firstly to determine the frequency and duration of different match profile variables of the Super 14 rugby teams during the 2006 tournament; secondly, to determine which match profile variables discriminate between the successful (top seven) and less successful (bottom seven) Super 14 rugby teams and thirdly, to determine which match profile variables predict the ranking of the Super 14 rugby teams of 2006. All the teams of the 2006 Super 14 rugby tournament were used in this study. Hundred and eighty five games were recorded on video and analyzed by means of the Opta Sports Data software package (Opta Sportsdata Limited, Harrogate, United Kingdom). The cluster analysis reduced the amount of test variables (16) to only 9. The results of the forward stepwise discriminant analyses revealed that the following match profile variables were the most important discriminators between the successful and less successful rugby teams: meters gained (F = 11.08), kicks out of hand (F = 5.9), linebreaks (F = 4.97) and percentage tackles made (F = 2.85). The classification matrix showed that the prediction functions were on average 92.86% accurate in classifying the rugby teams into the two original groups. Lastly, the forward stepwise multiple regression analyses determined that, line breaks (8%), kicks out of hand (28%), rucks won (3%), percentage tackles made (6%) and meters gained (46%) contributed the most to the variance (91%) in team rankings. Keywords: rugby, game analysis, performance