Thursday, 23 January 2014

Player Spotlight: James Tomkins

#5
James Tomkins
DOB: 29/03/1989
Position: Defender

West Ham United Career
Appearances: 168
Goals: 7

Debut: 22/03/2008 vs Everton


Debut Goal: 04/04/2009 vs Sunderland

Analysis
I’ve heard and read a lot about Phil Jones in the past couple of weeks, and how his versatility is hurting him. The consensus among journalists, supports and ex-players is that the England international needs to settle into one position for club and country before we see the best of him.

This got me thinking.

The general opinion on James Tomkins amongst the opinions I read is that he has the ability to be a top class (world class?) defender one day, but he suffers from inconsistency and concentration errors. This kind of fits with what I have always felt about the academy graduate and his performances.

But after reading about Jones recently, I wondered to myself how much Tomkins’ versatility has hindered his progress. In our Championship season, Tomkins was a regular at centre back, playing to the right of either Winston Reid or Abdoulaye Faye. However, since we regained our Premier League status in 2012, Tomkins has been a bit more of a utility player.

In the past season and a half, Tomkins has made 41 league appearances. 30 of these have been at centre back, 6 at rightback, and 5 (inexplicably) in midfield. On paper, it seems that he has made the vast majority (73%) of his appearances in his preferred position. 

However, we all know that centre backs prefer one side or the other. Reid always plays on the left, James Collins always plays on the right. This isn’t just confined to West Ham though. John Terry and Nemanja Vidic have been formidable defenders on the left hand side of the central pairing, while Rio Ferdinand and Per Mertesacker always appear on the right hand side.

Tomkins preference appears to be on the right. When he plays with Reid, he plays there. However, when James Collins (or more recently Roger Johnson) is playing alongside the former England under-21 defender, Tomkins moves across to the left. The logic behind this is seemingly that Tomkins is recognised as a better footballer than the other two (technically speaking), and is seen as being more adaptable to an unfamiliar position.

In the past, Tomkins has been in the side alongside Reid and performing admirably, until Reid is sidelined for a period of time. This tends to result in Collins stepping back into the side alongside Tomkins, and Tomkins moving to the left hand side of the two. Tomkins performance levels then appear to tail off, and when Reid returns from absence, Collins keeps his place in the side.

My belief is that the regular switch in position, however minor on paper, lead to Tomkins struggling to maintain his form.

According to Squawka’s performance matrix, Tomkins best performances do come when in his favoured RCB slot. On average, his performances in this position have comfortably the best performance score of his ratings in each position (19.32), as displayed by the graph below. It is also the only position in which Tomkins averages a higher rating than his overall average for his 41 appearances (15.18).


However, despite performing better on average over the course of time on the right hand side, there do seem to be exceptions to the rule. According to performance score, Tomkins best five performances since the start of last season have come in a variety of positions:

1. vs Fulham (30.11.2013); LCB; 65.86
2. vs Wigan Athletich (27.10.2012); RB; 60.88
3. vs Swansea City (02.02.2013); RCB; 50.37
4. vs Tottenham Hotspur (25.11.2012); RCB; 43.43
5. vs Sunderland (14.12.2013); LCB; 36.77

Similarly, his worst five performances have also come in a mixture of positions:

1. vs West Bromwich Albion (16.12.2012); RB; -30.12
2. vs Manchester United (21.12.2013); RCB;  -14.00
3. vs Cardiff City (11.01.2014); LCB; -11.50
4. vs Arsenal (26.12.2013); RCB; -9.00
5. vs Manchester City (19.10.2013); RCB; -5.63

What this suggests, is that Tomkins has the ability to perform (or not perform) in any position, highlighting a possible lack of consistency. However, his average performance level is greater in his favoured RCB position than any other.

This is highlighted by the lower range of performance scores in that position than others – although central midfield provides the smallest range of performance scores, this can be largely discounted by the small number of minutes played in this position (216), compared to others (RCB = 1447, LCB = 903, RB = 388).

The chart below demonstrates the differences in his performance levels by position. What you can see in the box is that Tomkins’ average performances at RCB are consistently higher than any other position.

On average, his best, median and worst performances are better at RCB than in any other position, demonstrating an increased level of consistency. The top whisker, though, tells us that we are less likely to see exceptional performances from Tomkins when he plays in this position. Similarly, the lower whisker is shorter than most of the other positions, demonstrating that you are less likely to see an appalling performance at RCB than elsewhere.

Based on this information, we could suggest that Tomkins is more likely to produce his average performances at RCB than when he plays elsewhere, and that his average in this position is much higher than his average elsewhere.


Not only does analysis of the Squawka performance score suggest that Tomkins best performances come at the right side of a centre back pairing, but it also shows the impact that positional changes have on his performance.

The chart below shows the average performance score that Tomkins posts per position, compared with the average score for that position when he has just switch from another role (i.e. against Norwich this season was the first time he had switched from right to left).

As expected, his performance tends to be significantly worse in those games where he has just changed position. The anomaly here is, again, when he plays in midfield.


 Aside from the mathematical, qualitative data, we can also see this sort of conclusion when we look at Tomkins’ performance data.

One of the most interesting stats I’ve found whilst looking into Tomkins’ performances over the past two seasons, is that he gives away a foul on average once every 114 minutes in the Premier League.

What is particularly interesting about this is that there is a perception that Tomkins regularly gives away silly fouls, through a lack of concentration or ball watching. But when Tomkins is playing in his preferred position, he concedes fouls far less than on the other side of the central two.

When playing on the left hand side of the defence, Tomkins concedes a foul once every 82 minutes. This is the only position in which he concedes a foul more regularly than once every 90 minutes. When playing on the right, he concedes a foul once every 121 minutes. This means, that when Tomkins is playing on the right hand side of the defence, it takes him an average of 39 minutes longer to concede a foul than when he is playing on the left.

Once again, this seems to support the assertion that the versatility that we perceive Tomkins to have, may not actually be based in fact. His technique is required to change when playing on the other side of the field, his positioning would also need to adapt. The fact that he fouls more regularly on the left, suggests that his technique and positioning are significantly weaker in this position.


This inference is supported by Tomkins tackle success rate by position. When playing on the right centre of the defence, Tomkins is successful with 74% of his tackles. When moving to the left hand side, however, his success rate drops to just 50%.


However, there are some results that I didn’t expect to see. For example, 17% of Tomkins attempted passes are long balls. The position from which he plays the highest proportion of long balls is his favoured right centre back (20%).

On the left hand side, he plays just 16% of his passes as long balls. This seems to contradict the opinion that Tomkins is the ball player in our defence, particularly as the rate spikes when he moves into his ‘best’ role.

Another surprise was that the team seem to concede fewer goals when Tomkins is on the left hand side. Playing in this role, Tomkins sees his defence concede once every 65 minutes – the least regular of all positions played in the past two seasons. In the RCB role, a goal is conceded once every 58 minutes, still above his overall average. The rate worsens considerably with Tomkins completely out of position, one conceded every 55 minutes that he plays at right back, and once every 36 minutes when he steps forward into midfield.


Obviously Tomkins performance isn’t only influenced by his position on the pitch; his teammates can have a huge influence here. When playing in a central two, each half is reliant on the other to perform to the best of their ability.

Something that is noticeable in Tomkins’ performances is the change in emphasis dependant on his defensive partner. When playing with Winston Reid there seems to be a calming influence at the back, whilst many have remarked upon how hectic and rushed we can seem when Tomkins and Collins are paired together. This is, once again, borne out by the facts. When playing with Reid, Tomkins makes a clearance once every 8.73 minutes, compared to the more frequent rate of 6.98 minutes when paired with Collins.


What all of this suggests to me, is that James Tomkins isn’t as versatile as we think.

Despite being the only centre half in our ranks who is used in more than one position, his performances outside of his best right central defence suggest that he is far less effective when moved. Statistically, his performances in that position average at a higher level than in the other roles we have seen him dropped into over the past couple of seasons.

I suppose my opinion comes brings us full circle. Like Phil Jones, James Tomkins needs to be deployed in one role, in one position. We all saw in our Championship season how a consistent selection process worked wonders for both Tomkins and Reid. We need to see this again if we ever want the man from Basildon to achieve his full potential.

1 comment:

  1. Interesting analysis. Would be curious to know how Tomkins' foul rate compares to that of Reid and Collins. From watching the games, my sense is that Collins commits fouls at a higher rate than the other 2. I'd guess the rank is Reid, Tomka, then Collins. That's probably to be expected since Collins has less pace than the other 2 (altho with more experience). Of course, then there is Collins defensive "style" which likely produces a higher foul rate.

    ReplyDelete