top of page
  • Writer's pictureSimon Chi

WORLD RUGBY: TEAM OF THE DECADE PLAYER RATINGS – BACKS

In December 2020, World Rugby released the results of a fan poll which sought to identify the Team of the Decade (2010 – 2019). Predictably there was passionate discussion after the team was released around selections (both made and omitted), so we thought it would be a fun exercise to see which players advanced analytics would identify to be the top players for each position. For this exercise, we looked at data from the following competitions:

  1. Rugby World Cup (2015, 2019)

  2. Six Nations (2014 – 2019)

  3. The Rugby Championship (2014 – 2019)

Those of you who are keen observers may be quick to point out that the data doesn’t cover the entire decade. This is clearly a limitation of the study, but we can’t study data that doesn’t exist. Regardless, we will be well positioned to conduct this analysis for the next decade of players! From the available data, we decided to split the analysis into forwards and backs due to the fact that there are some game actions that were specific to forwards compared to backs (e.g. forwards scrum, backs coif their hair). For this study, we looked at the following actions for backs:

  • Carry metres

  • Kick metres

  • Pass metres

  • Lineout receive (quick lineouts)

  • Lineout throw (quick lineouts)

  • Jackals won

  • Tackles completed

  • Missed Tackle Impact

  • Penalties conceded

  • Turnovers conceded

It’s important to note that this study was limited to players who played a minimum of 800 minutes. While this threshold was arbitrary, the rationale was if you couldn’t play the equivalent of at least 10 full games for your country over 10 years, you likely weren’t a strong candidate for inclusion in the Team of the Decade. So, if your favourite player is not listed it is very likely that they didn’t meet the 800 minute threshold. The only players who played less than 800 minutes that were included were those players selected in the Team of the Decade poll. The EPA for each of these players was collected and normalized to 80 minutes. This is so we can compare apples to apples across all measures for all players. The top 10 candidates within each positional cohort were listed.


One confounding factor that was not initially anticipated was the fact that several players played multiple positions over the course of this study. Notable examples were Beauden Barrett (10/15), Owen Farrell (10/12), and Ruan Pienaar (9/10). Given the way that the data is structured, it would require a separate project to allocate game actions to the appropriate positions played for these players. For this reason, these individuals that played multiple positions were excluded from this study. While it might be disappointing to not see where these outstanding payers rank amongst their peers, we will look to remedy this problem for future analyses.


To be clear, the players identified in this study are those that make the greatest overall positive EPA contributions to their teams in selected areas of the game. This is by no means a definitive statement as to the best player in each position. This is an evolving process where we can look to make improvements over future iterations of this type of analysis. At the very minimum, we can use this type of analysis to act as an initial filter to identify players that should be considered, after which coaches, fans, and other experts can have their own discussions adding their own context as to who deserves to start on this team.



SCRUM HALF (Table 1)

This particular positional cohort might be a bit controversial. One thing that stands out is the relative position of Aaron Smith to TJ Perenara. Rather than immediately jumping to the conclusion that Perenara should be starting over Smith, perhaps it’s worth considering various explanations for the difference? Based on the minutes played, it can be implied that Smith was typically the starter and Perenara was the “finisher”. Thus, it is likely that Perenara made his impact in the second half when the game “loosened up” and it’s also likely that he came into the game with a lead. While this provides a possible explanation for the difference in scores, it is also important to note that the coaching and the performance analysis staff for the All Blacks would also be able to provide a lot more context around performance and selection that would provide further insight into both players. Regardless, these results could no doubt stimulate some lively discussion at the pub?

TABLE 1: Scrum Half player ratings (EPA_80) based on performances in the Rugby World Cup, Six Nations, and The Rugby Championship. World Rugby Team of the Decade selections are highlighted in yellow.



FLY HALF (Table 2)

What stood out to me was despite a small sample size from games taken at the end of his career, Daniel Carter still performed at an exceptional level in the most important games. In terms of carry/kick/pass metres he was at or near the top when compared to all of his peers, but where he separated himself from the pack was that he made relatively few errors (MTI, penalties conceded, turnovers conceded, etc.).

TABLE 2: Fly Half player ratings (EPA_80) based on performances in the Rugby World Cup, Six Nations, and The Rugby Championship. World Rugby Team of the Decade selections are highlighted in yellow.



WING (Table 3)

TABLE 3: Winger player ratings (EPA_80) based on performances in the Rugby World Cup, Six Nations, and The Rugby Championship. World Rugby Team of the Decade selections are highlighted in yellow.



CENTRE (Table 4)

This is a case where you have two legends of the game (O’Driscoll, Nonu) being captured towards the ends of their respective careers.

TABLE 4: Centre player ratings (EPA_80) based on performances in the Rugby World Cup, Six Nations, and The Rugby Championship. World Rugby Team of the Decade selections are highlighted in yellow.



FULLBACK (Table 5)

What stood out to me here was how similar Stuart Hogg’s numbers were to Ben Smith. The only areas where they differed were kick metres and MTI. Hogg was clearly the more prolific kicker. Since the fullback in most scenarios is the last line of defence, the MTI is an important measure of a player’s ability to tackle in open space. Smith produced the better number in terms of MTI. However, when aggregated into a total score, Hogg’s kicking was the point of difference between the two players.

TABLE 5: Full Back player ratings (EPA_80) based on performances in the Rugby World Cup, Six Nations, and The Rugby Championship. World Rugby Team of the Decade selections are highlighted in yellow.



THE BETWEENERS (Table 6)

Some of you may be wondering how the players who play multiple positions fared? What is impressive is that they were still ranked amongst the top players indicating they were capable of playing at an elite level at multiple positions.

TABLE 6: EPA_80 player ratings for notable “betweeners” based on performances in the Rugby World Cup, Six Nations, and The Rugby Championship. World Rugby Team of the Decade selections are highlighted in yellow.



FUTURE WORK

I’ve alluded to the fact that this is an evolving work in progress. Other aspects of the game that we would like to include in future models:

  1. Kick retention reception percentage (for contestable kicks and restarts)

  2. Goal kicking

  3. Breakdown order of arrivals (OOA)

What other aspects of the game for backs would you like to see incorporated into this model?

Comments


bottom of page