In 2012, Michael T Hughes, Michal M Hughes, Jason Williams, Nic James, Goran Vuckovic and Duncan Locke wrote an insightful academic journal discussing the performance indicators in rugby union during the 2011 World Cup. They gathered various materials from professional analysts working for coaches and player at the World Cup event, and verified the reliability and accuracy of their data against video footage from different matches.
This research study analyses the influence of the following key performance indicators in the final outcome of a game:
Scoring Indicators:
Points scored
Total points scored in WWC 2011
Points scored per game
Points scored agains Tier A teams
Points scored per game against Tier A teams
Tries scored
Total tries scored in WWC 2011
Tries scored per game
Tries scored from set pieces
Percentage of tries scored from set piece
Tries scored from set pieces per game
Tries scored from broken play
Percentage of tries scored from broken play
Tries scored from broken play per game
Quality Indicators:
Total Possession - Times and Productivity
Minutes that ball is in play in the match
Rest minutes in the match
Minutes with possession in the match
Percentage of time with possession
Number of possessions in the match
Minutes per possession
Minutes of possession per point scored
Number of possessions per point scored
Minutes of possession per try scored
Number of possessions per try scored
Total number of line breaks
Total number of line breaks per game
Minutes of possession per line break
Number of possessions per line break
Total number of set piece line breaks
Total number of set piece line breaks per game
Percentage of set piece line breaks
Total number of broken play line breaks
Total number of broken play line breaks per game
Percentage of broken play line breaks
Number of phases in the match
Percentage of phases per possession
Attacking penalties won
Attacking Possession
Number of possessions in opposition's 22 line
Number of converted possessions in opposition's 22 line
Percentage of converted possessions in opposition's 22 line
Number of points from opposition's 22 line
Number of points from opposition's 22 line per game
Number of points per possession in opposition's 22 line
Kicking game
Total number of kicks at goal
Total number of kicks converted
Percentage of kicks converted
Penalties conceded
While these key performance indicators of a rugby union game or tournament can be useful to summarize the some elements of a team's performance, what M. Hughes et al (2012) found was the there was little correlation between each individual metric, or set of metrics, with the final outcome of the World Cup 2011 tournament. For example, France was identified as one of the worst teams in most of these metrics, though they were the runners-up of the tournament.
The paper also touches on the challenges individual player performance analysis in rugby union. Due to the nature of the sport, a specific position on the field will require its own set of performance indicators. The study suggests to analyse an individuals performance against common key performance indicators and use that individual's performance profile to run intra-position comparisons (Hughes et al, 2012). This also leads to the creation of position profiles, where strengths and weaknesses of players playing in each position can be identified. It is also suggested that the individual player profiles should be based in the context of the team's profile as well as the opposition team's strengths and weaknesses, as these elements will impact a player's performance profiling.
Similarly to most team sports, randomness and luck can play a big part in the final outcome of a rugby union match. Therefore, predicting the performance of a team based on a few data points might not be enough to correlate it to the final performance achieved by that team. There are many complex interactions that occur during a rugby union game between teammates and oppositions which are difficult to account for through today's available statistics. However, studies like the one carried out by Hughes et al (2012) are another step towards narrowing down the best procedures to follow to successfully apply analytics to rugby performance predictions and team sports in general.