What is the best XV of the group stages, according to statistics?

[RugbynExcel]

Hello everyone

I'm making this post to share with you the best xv of the group stages of the World Cup, based exclusively on statistical data.
The data comes from the official World Cup website and the World Cup Fantasy game.

As I'm a fan of rugby and statistics, I thought I'd take this opportunity to produce a statistical analysis of the performances of the players taking part in this World Cup.

The ranking is based on a multitude of criteria, divided into different categories:

Offensive activity : kicks, passing (unfortunately, I wasn't able to add the runs, since the World Cup website doesn't display this data for all the players...)
Offensive efficiency: Linebreaks, defenders beaten, metres run
Key passes and offloads: Offloads, Try assists, Linebreak assists
Tackling: Successful tackles, percentage of successful tackles
Turnover: ground turnovers, interceptions
Lineout: Lineout won, Lineout steals
Scrum: Scrums won
Discipline and errors : Penalties given, yellow and red cards, handling errors
Scoring: Tries and points scored

For the first 5 categories, the statistics achieved by the players are averaged over 80 minutes of play.

This means, for example, that the players with the highest tackle score are not necessarily those who have made the most tackles,
but rather those who made the most tackles in proportion to their playing time.

All these categories give each player a score out of 10, and are added together to give an overall score for the player.
Finally, a bonus is awarded to each player depending on his team's final ranking in the pool.

I've tried to make my analysis as complete and balanced as possible, but there are bound to be big differences in points depending on the positions played. It is therefore not very relevant to compare the scores of two players occupying two different positions. That's why I'm presenting my results as a 'best-xv' rather than a top 10 of the best players.

All players who have played a minimum of 160 minutes (the equivalent of 2 full matches) during the group stages are ranked. This represents a list of 280 players.
I can detail the scores of the players not present in the best-xv presented if that may interest some of you.

Here is my best xv for the group stages, based on the statistical performances of the players:

Props
B. Gigashvili (Georgia) - 33.9 pts
B. Tameifuna (Tonga) - 33.7 pts

Hooker
P. Mauvaka (France) - 48.0 pts

Lock
S. Cummings (Scotland) - 42.1 pts
T. Beirne (Ireland) - 41.9 pts

Loose Forward
J. Morgan (Wales) - 47.2 pts
M. Leitch (Japan) - 46.7 pts
N. Martins (Portugal) - 46.2 pts

Scrum Half
F. Lomani (Fiji) - 39.4 pts

Fly Half
D. Mckenzie (New Zealand) - 46.9 pts

Centre
J. Tuisova (Fiji) - 40.0 pts
B. Aki (Ireland) - 39.9 pts

Outside back
D. Penaud (France) - 49.3 pts
T. Ramos (France) - 47.8 pts
D. Graham (Scotland) - 44.9 pts


This XV is therefore made up of the players with the best overall statistics.
Of course, this remains an analysis based only on numbers, which are not always representative of the reality of what happens on a rugby pitch.There are many parameters that determine a player's level or impact on the game, but these cannot be quantified. Because what defines a good rugby player cannot be measured or calculated. As World Rugby says, a good rugby player is first and foremost someone who shows integrity, passion, solidarity, discipline and respect, and all these fine values cannot be quantified

 

[retired9]

Interesting post, good mix between top teams and some further down. It would be interesting to see if in second or third place there is a lot of Irish and All Blacks for example - showing where the individual "brilliance" or collective team cohesion is.

 

[RugbynExcel]

Interesting post, good mix between top teams and some further down. It would be interesting to see if in second or third place there is a lot of Irish and All Blacks for example - showing where the individual "brilliance" or collective team cohesion is.

Thank you for your reply

There are indeed a lot of All Blacks in the second places of the ranking.
Here's what the substitutes' bench looks like for this team:

C. Taylor - 42.6 pts
T. Francis - 33.1 pts
G. Thomas - 32.7 pts
A. Fakatava - 40.8 pts
A. Savea - 46.1 pts
C. Ollivon - 43.9 pts
W. Jordan - 43.5 pts
B. Barrett - 42.0 pts

Defining a team's profile (in terms of individual and collective performance) on the basis of this data alone is complex and the results must be interpreted very carefully.
Here is a table showing the average score of the players per team as well as the standard deviation of the scores



It is quite clear which teams came out on top in the group stages and which struggled to compete with the others.
However, several results seem surprising at first glance:
-Portugal's relatively high ranking compared with their position in the group
-The low ranking of Argentina and South Africa, despite being second in their group.

How do you explain that? In Portugal's case, there were some players who really stood out (I'm thinking of Martins, Simoes, Marques, Storti and Tadjer, for example). But in view of their match, it would be a mistake to explan their fine performances with the play of just a few of them, given how impressive their cohesion was.

For Argentina and South Africa, their results can be explained more by the absence of players who clearly shone through with their individual performances, unlike teams like New Zealand or France, but also by the absence of players who played poorly. These are teams where the performances were homogeneous, as shown by the standard deviation of the scores. In short, everyone played well, but no one was incredible. The same goes for Ireland
These teams have also played matches where the game has remained very tight (I'm thinking of England-Argentina and Ireland-South Africa), which have therefore not been very productive in terms of statistics. I haven't found any data on the average effective playing time of these teams during the World Cup, but if it was low, that could also be a possible explanation.

Keep in mind that this only concerns matches played in the group stages. I'll be doing a similar analysis once the World Cup is over, and the results will probably change.