Player profiles: Visualising the results

Introduction

Over the past three posts, player profiles have been discussed for passing, shooting and defending. In these posts, the following disclaimers have been repeatedly made:

• The profiles are not all I would like them to be - in part this is because they are limited by the data that is available in the MCFCAnalytics dataset and in part this is because it is difficult to decide which factors should form part of each profile
• Although I believe the profiles to be useful, the results of them are not easy to display in a meaningful way.

While I still believe the first point to be an issue, I hope that this post goes some way to solving the second point, based on graphs shown in the defending post.

 Results

Excel's radar plots have been used to allow easy comparison between players. Three plots are required to describe the three profiles generated by the calculations described in my three previous posts. An example is shown below, for 5 midfielders at Manchester United: Anderson, Carrick, Cleverley, Fletcher and Scholes - midfielders are chosen as they are likely to be quite involved in all three facets of the game (click the image to enlarge):

Some notes about the display of the graphs:

• The axes are labelled according to the following convention: Factor: Inside - Outside. For example, the top axis on the left hand graph shows the players' involvement in the passing game, where points closer to the centre represent peripheral players and those closer to the outside represent the focal players
• The colour of each line is the same for each chart

Some notes about the content of these graphs:

• The main difference in the shooting profile lies in the foot used: Anderson favours his left foot, Carrick uses both and the others favour the right
• In the passing profile, Anderson stands out. He is more involved than the other players, plays further up the pitch and plays longer passes. Cleverley's passes are shortest and they are more backwards than the others
• Defensively, as might be expected [full disclaimer, I'm a Carrick advocate and believe that most critics do not fully understand the role he is expected to take], Carrick favours interceptions and blocks to tackles and duels (more passive than aggressive), whereas Cleverley is the opposite. Cleverley's clearances are the most aerial of the group

Another example, England midfielders:

Notes:

•  Shooting - Again, the main difference lies in the foot used. Barry prefers his left, Carrick again uses both and the others prefer their right foot. Lampard is the most focal, Carrick the least involved in set pieces
• Passing - Gerrard, Lampard and Carrick are the most focal. Gerrard passes furthest up the pitch, followed by Lampard and Barry, with Carrick and then Parker operating in the deepest areas. Gerrard's passes are also the longest, with Parker's the shortest
• Defending - Barry and Parker are the most involved with Lampard and Gerrard the least. Carrick and Parker are the most passive, with Gerrard the most aggressive. Parker's clearances are most ground-based.

Conclusions

I think these graphs make it much easier to see the differences between the styles of each player than the raw numbers ever could. Players whose behaviour is much different to others stand out quite obviously.

It is also interesting to note that in each of the examples shown, the players play - at least nominally - in the same position: central midfield. However, the way that each player plays is noticeably different. On one level, this is obvious: we all know that Scott Parker and Frank Lampard are very different players but these profiles gives us a way to quantify this and describe the differences. Possibly more importantly, these profiles allow us to quantify the difference between players whose differences are perhaps less obvious. To the "eye test", Barry and Carrick appear to fill similar roles in a team, but these profiles - and most importantly these charts - show their differences.

As mentioned at the top, the profiles themselves are still probably best described as "incomplete", but this post demonstrates the power of a good visualisation of data.

Please let me know what you think, either in the comments below or on Twitter.