Blame Rob, who suggested we needed an article on squad analysis when I did my last review of “relegation watch”. Or Malcolm, whose idle speculation led to the idea of specific comparisons. Some of them weren’t particularly enticing but they planted a seed so I trooped off to hosted.stats.com and did a bit of digging. Then I did a bit of moving things around and playing with the numbers on a spreadsheet and here we are.
This is the first of what I think will be two posts comparing our players. It concentrates on midfield and doesn’t include the Hoff as Sunderland have him listed as a defender rather than a midfielder. I’ve also excluded Rees Greenwood and George Honeyman, who made their debuts against Watford.
First, let’s take a lift (more or less) from the excellent hosted stats website:
|Games||Minutes||Goals||Assists||Shots||on target||% on target||Yellows||fouls||Crosses|
We’ll ignore Jordi Gomez and Seb Larsson for the time being as Gomez only made 6 appearances, averaging two thirds of a game in each, which doesn’t give us a strong basis for comparison, and while Seb made 18 appearances a lot of them were as a sub late on in the game.
Item 1: Scoring goals
You can see that our goal scoring record from midfield is dire. Only four players scored and they only managed 7 goals between them. Catts has the worst record despite being keen to get into the box. He managed to get 11 shots away, one was even on target. Compare that with Jeremain Lens, who was on the pitch for a little more than half of the time. Lens managed to get in 26 shots, over a quarter were on target and three went in. It’s the best record of the six we’re looking at, although three of the other four are worth a mention. Wahbi Khazri scored twice and Yann M’Vila and Jack Rodwell once (and he should have had a second but the ref got it wrong) Rodwell also had the highest accuracy when shooting.
So when it comes to goals We have to say that Jeremain Lens was our best midfielder and Catts was our worst.
Item 2: Fouls and Yellows
Is Catts also our dirtiest midfielder? With 40 fouls and 8 yellow cards to his name you could argue that he is. M’Vila, with only 17 fouls and 1 yellow, appears to be much cleaner. But this is a simplistic argument and it only goes to show the limitations of statistics. Do Catts and M’Vila have the same role on the pitch? Would it be fairer to compare Catts with Rodwell? Rodwell’s 5 yellows came at the rate of one every four hours of play, whereas Cattermole’s 8 came at the rate of one every 4.5 hours. But Rodwell did manage twice as many shots in half the time – a testament to his desire to get forward and resume the box-to-box role, perhaps.
Lens, by the way, got midfield’s only red.
What else should we look at?
Item 3: Providing opportunities
With Kirchoff as the break-up man we want these midfielders to be providers to our forwards. How well do they do it? Well, Catts scores poorly again. No assists and only 23 crosses, compared to M’Vila’s 4 assists and 69 crosses. Not bad from M’Vila, you might think, but Lens again demands comparison. He not only scored more goals, he also managed 3 assists and 70 crosses in half the playing time. And those who were impressed by Whabi Khazri (who wasn’t?) might be surprised to learn that his crossing rate was lower than Lens’s. In fact Lens put over a cross on average every 18 minutes whereas Khazri put one over every 22 minutes.
So you could say that when it comes to providing opportunities for forwards Lens was our best midfielder and Catts was our worst.
But, of course, such statements are based on simplistic calculations and they don’t always stand up to scrutiny.
|M’Vila, for instance, put over a cross only every 46 minutes on average. But how many times did he play the ball forward to feet? How many chances did he set up? We don’t expect to see him burst down the wing and put the ball over. That’s not his role.
Think back to Fabio Borini’s equaliser against Chelsea. Who put the cross in? It was Patrick van Aarnholt, not a midfielder, and getting down the wing and crossing is one of the things he does.
But who fed Van Aahnholt? It was Yann M’Vila, with a recovery, a quick look up and an inch-perfect pass. No mention in the stats for Mr M’Vila, but what a crucial role in that goal, and in others.
And if tangible things like killer passes to Patrick don’t appear in stats what chance is there for intangibles?
Which brings me back to Catts. If you had to choose one player who epitomised a will to win who would it be? John O’Shea or Wes Brown, perhaps. They both have form when it comes to winning and they both defend with heart and wouldn’t be bad choices. Or would you choose Catts? Over his time he’s had a lot of stick for impetuous tackles and rash decisions but no-one has ever questioned his commitment, his willingness to fight for the cause and his ability to rally the troops. Someone on this site said the team play better when he’s in it and there’s a lot of truth in that.
So I’ll finish with some of the calculations I’ve done. Be aware, however that even these need to be taken judiciously. M’Vila, for example, has been present since virtually the start of the season whereas Khazri, who appears to have the best points per game, only played in a settled team during the final third.
|pts per game||1.4||1.2||1.1||0.9||0.6||0.9||0.3||1.4|
|pts per 90 minutes||1.6||1.5||1.1||1.2||1.1||1.5||0.5||4.1|
Percentage wins: Khazri 29%; Catts 29%; Toivonen 25%; M’Vila 24%; Lens 20%; Rodwell 10%.
Percentage of games played in which points were earned:
Khazri 79%; Catts 61%; M’Vila 57%; Lens 45%; Rodwell 45%; Toivonen 42%
Catts and M’Villa played in all of our wins. That’s not surprising for M’Vila as he played in all games but one. But Catts played only 31 and not only did he play in every win he also played in 10 of our 12 draws. That’s some contribution. Another way of looking at it is that he missed 7 games, and when he didn’t play we didn’t win. We drew two of those games and lost the other five.
Points per game doesn’t necessarily take account of substitutions so I’ve included a total for points earned per 90 minutes on the pitch. You’ll have to judge for yourself whether or not it improves things. It does show Lens and Rodwell in a better light but makes little difference to M’Vila, who was an almost ever-present. Whatever measure you use, you can see that Whabi Khazri was the best midfield points earner, reflecting the improvement in the whole team in the second half of the season.
And that, for me, illustrates one of the weaknesses of such stats. I’m not comparing like with like. Khazri didn’t play the whole season. Catts came and went and came back. With one exception (Southampton) he played in every game that Khazri played in and M’Vila played in them all. If we only included the last 14 games of the season would Khazri, Catts and M’Vila have the same points per game? Probably, which takes us back to the top chart with its shots, goals, fouls, etc to enable comparison, and the weakness it has in ignoring different roles. Sometimes you need more, and you don’t always get it.
In the end, yer pays yer money and yer makes yer choice. Statistics help but they don’t make the decisions for you.