Gary Cahill: good, or just English?

Gary Cahill, gazing into the future. (Creative Commons/Jacoplane)

I hate to feed the transfer frenzy stirred up by the general media, but this was something I couldn't resist taking a swing at. For those opposed to transfer mania, I'm sorry. But I'm doing it anyway.

The situation

The news concerning Arsenal's defense is not good. Thomas Vermaelen, out since August with an Achilles injury and setback after setback, has had another, and finally is set to have surgery that will put him out at least an additional four-to-six weeks. Sebastien Squillaci, who's slipped into the third-choice center-half role already, has also suffered an injury - he picked up a low-grade hamstring strain against Leeds, and while the injury is not too severe it will put him out fora few weeks, according to Arsene Wenger.

This reduces an already-thin central defense corps to two men - Laurent Koscielny and Johan Djourou - with Alex Song as an emergency back-up.

By the way, that would have to be an extreme emergency in my book. While Song is competent at the center of the defense, he's far too valuable in midfield to be moved. He's finally reached the balance that fans begged for and Wenger envisioned, where he functions effectively on the attack but doesn't neglect defensive duty. He's played very well as a defensive midfielder and I would not want to break his run of form or risk weakening the crucial midfield triangle.

This puts Wenger in an unenviable position - he has had to admit that he needs to buy another central defender. This will weaken his bargaining position, but there isn't much to be done about that.

The linked article from the Guardian lists several possible targets - Per Mertesacker from Werder Bremen, Emir Spahic from Montpelier, and Mamadou Sakho from Paris St. Germain - but I'm going to concentrate on Bolton's Gary Cahill, because I feel that he's the most likely get at this point. The question is this: Is Gary Cahill good as advertised and good enough for Arsenal, or is he a case of a player being overrated simply on the power of his English passport?

More stuff
The subject

I feel that I should explain more fully why Cahill is the one I'm concentrating on. Arsenal have been linked to him over and over again ever since the "we need a center-back" meme originated. Here's a report that he was going to replace Gallas (rather than Koscielny and Squillaci), and here's a story from May of 2009 that actually predates Coyle's tenure. Point is: this has been going on for a while.

A big part of the reason I'm on board this time (in terms of the rumor being valid, at least) is that it actually makes some sense logistically. Wenger and Coyle have a bit of a relationship, stemming from the Jack Wilshere loan last year. Wilshere was loaned to Bolton and has credited his time there as having a massive effect on his development as a player. It was undoubtedly a successful loan, and when you have success in a transaction that can surely help in fostering further business. Now Bolton are looking to loan Carlos Vela from Arsenal, and after the success that Wilshere had there I would say that's a splendid idea.

Here's where Gary Cahill comes into play. He's been put on the market by Bolton, due to a need stated by the club's chairman to reduce debts. They want a deal in the neighborhood of £15 million, and will likely be hearing from several clubs about him. Arsenal have the need, and can sweeten the pot with the prospect of Carlos Vela's loan - in fact, that may even bring down the asking price a bit. Coupled with the fact that both managers know they'll get a fair deal from the other and Cahill makes the most economic and logistic sense for Arsenal. In short, as Arseblog said,

Bolton are trying to sign David Wheater from Boro, Bolton want Carlos Vela on loan, Bolton need money, Gary Cahill plays for Bolton. There’s certainly scope there to do a deal but I suppose it all depends on price.

Cahill is particularly attractive for a few ancillary reasons as well. Among others, he's English. I don't tend to care about this too much - I'm not the Don Cherry of football or anything like that, and all of my favorite players aside from Theo Walcott are foreign (well, technically they're all foreign to me, but that's beside the point). His nationality is a positive mainly because of the roster rule changes that give advantages to English-born and home-grown players. Cahill would add an English passport to a squad that doesn't have many in the first-team at the moment. Additionally Cahill has Premier League experience already. I think a bit source of the issues that Arsenal's defense has had to this point has been the fact that two of our three first-choice center-halves are Premier League rookies. Both Koscielny and Squillaci have been candid about the fact that moving to the English style of play takes adjustment. Cahill is already adapted, unlike Mertesacker and the others. Finally, he has a quality that Wenger values highly - he's still young. Cahill is only 25 (and only just, his birthday was in December), should he move to Arsenal it's likely that he'd have several years of top-flight play left in him (barring injury, of course).

Owen Coyle has said that Cahill won't move in the season. If I'm to take all of the transfer rumors with a grain of salt, though, I see no reason not to do the same here. It is surely more reliable - it's straight from the manager, and he is a respectable man - but during the transfer window, I feel like most of what is said comes from the point of view of negotiating rather than pure truth. It's possible that he means this, and likely that if Coyle had his druthers he'd never sell Gary Cahill. But it's also possible that he's simply saying this to try to raise Arsenal's bid - that's probably what I'd do in his place. This is different from the Cesc-to-Barcelona debacle, by the way, because the only ones talking about Cahill are Coyle and the press. If Jack Wilshere and Theo Walcott were saying that it was inevitable that Cahill would sign, that would be one thing. But they're not; if there's anything to this, so far it's been conducted with honor. So I'm not abandoning this thought experiment.

The process

So that's why I'm concentrating on Gary Cahill above all others. The better question is this: should Arsenal? There are very few Premier League statistics openly available, at least as far as I can find. I can look at the top goalscorers and assist leaders all I want, but I can't seem to find passing percentage or tackling stats anywhere, much less anything more advanced than that. So I'm going to do some math myself.

What I want to do is compare Cahill to what Arsenal already have, and see two things: one, whether he could fit into the Arsenal system; and two, whether he's better than or even as good as what we have. I'll use Guardian Chalkboards to do this - I'd rather have a database of rate statistics, but I'll make do with what's available.

To make this manageable I've had to limit the number of games that I'm looking at. I've made a command decision and the matches that I'll look at will be those played against big clubs. I feel these to be the most crucial matches (obviously since they have the most bearing on the title race) and the best approximation of the onslaught that Arsenal usually faces. For Bolton this means one match against Arsenal, two against Liverpool, one against Spurs, one against Manchester City, one against Chelsea, and one against Sunderland.

I have included Spurs because of the weight of the north London derby, Liverpool because though they are in poor form this year they still must be considered a "big club," and Sunderland because they are immediately ahead of Bolton in the table at the moment (in 6th place). Bolton are 1-0-6 (beating only Spurs) in these matches, scoring 6 and allowing 12. They also drew Manchester United 2-2 but Cahill did not figure in the match.

For Arsenal the relevant matches include Liverpool, Bolton (included because of the faceoff against Cahill's own team), two against Chelsea, two against Manchester City, one against Spurs, and one against Manchester United.

You may notice that there is one match more on the Arsenal side of the ledger than the Bolton side. This is not intentional, it just happened that way. If it were likely to influence the results it would be an issue; however, as you're soon to see, in terms of counting statistics a one-match difference won't make a big difference.

Since it's assumed by most Arsenal pundits that when healthy (if that's ever applicable again) Vermaelen is the ace center-back, the Arsenal players that I will compare with Cahill are those who would likely be in most direct competition with him for a starting spot: Djourou and Koscielny. I am using the stats from only one player from each match; thus, when both play, I've gone with Djourou, as he's the fan favorite at the moment and I personally rate him more highly than Koscielny. The statistics for the two players appear relatively similar, so I don't believe this will influence the findings too much.

I've chosen three events to focus on: passes, tackles, and clearances. The reasoning is Arsenal's playing style - first of all, these three are good indicators of pitch positioning; that is, I want to see where Cahill spends most of his time compared with the current Gunners. Since we play such a free-flowing game, positioning is particularly interesting, because if he already plays like an Arsenal player, he'll obviously fit into the squad more easily. Beside that, these three events are, to me, crucial to our style. Passing should be obvious to anyone who's seen even one Arsenal match. Tackles are crucial because Arsenal's game is based on possession - maintaining it is crucial, allowing opponents to maintain it is costly. Clearances are relevant because if the pace that our current team can display - a well-placed clearance to (for example) Walcott can lead to a breakaway chance on goal. This may seem obvious, but it's important to make sure it's known as it applies more directly to our style of play than that of other sides (for example, the Route 1 game that was employed against Arsenal so effectively by Ipswich).

Granted, clearances may seem like an odd choice. I went with that because 1. I hate sustained periods of opponent attacking coupled with poor clearances, which piss me off, and 2. I wanted a third event to increase the sample size, and of those available on the Guardian chalkboards clearances are the next most exclusive to defenders.

So now, at long last, we can look at...

The results

In the relevant games, Gary Cahill was actually fairly comparable to Koscielny and Djourou in both prolificity and efficiency, with some qualifications. In terms of passing he was significantly inferior. In the seven-game sample, Cahill completed only 120 passes out of 183 that were attempted, a success rate of approximately 66 percent (I have rounded all decimals). To compare, Koscielny/Djourou successfully completed their passes at a rate of 86 percent - a whopping 280 out of 324 in an eight-game span. Here is where the irrelevance of the one-game difference becomes evident - Cahill completed fewer than half as many passes as the Arsenal duo, which I doubt one game would affect.

The difference shrinks greatly when we examine tackling. Cahill won 35 out of the 56 tackles he attempted in the sample, a success rate of 63 percent. Djourou and Koscielny both attempted and completed fewer tackles than Cahill - 27 out of 42 - but were incrementally more efficient, with a 64 percent success rate.

In clearances, Cahill emerges the victor on both fronts. He was prolific, with 40 successful clearances made and 61 attempted. He also was efficient - a 66 percent success rate. Compare this to Koscielny and Djourou - a joint 30 clearances out of 50 attempts, with a 60 percent success rate.

To examine positioning and effectiveness, let's look at chalkboards of both the most prolific (in terms of pass attempts) and most efficient (in terms of pass completion percentage) games for the players in question.

by Guardian Chalkboards

This is Cahill's most efficient passing match. Here we can see that Cahill plays a fairly conventional right-side centerback role, although he does move forward at times. In terms of passing, though, the majority of his play takes place on his side of the pitch in the defensive end. He roams within that general area, but does not appear to bomb forward in the way that, for example, William Gallas did.

by Guardian Chalkboards
Keeping in mind that Djourou was playing on the opposite side of the central defense, there are some relevant similarities and differences between the two performances. Looking at the heatmaps you can see that both players complete most of their passes in the same (or equivalent) sections of the pitch. Djourou is not Gallas either, in other words. However, in both the passing map and the heatmap there's an important difference - Djourou's passing is much more compact, and Cahill's more expansive. Both players pass from the same sections of the pitch, but Cahill finds himself in different parts of those sections. He also roams about at times near his own endline, where Djourou attempted only two passes in the sections closest to the endline.

To ensure this isn't an aberration, I want to look at the matches in which the players attempted the greatest number of passes, regardless of success rate. This way we can be sure we're properly seeing the players' typical positioning.

by Guardian Chalkboards

by Guardian Chalkboards
This set of maps is even more instructive because both Cahill and Djourou were playing on the same side of the defense. We see the same thing again - Djourou bunched up, for the most part, where Cahill is more spread out, even bombing forward on the right flank a bit. There's another difference that's more evident here than in the first game - Djourou's passing is (for the most part) much direct and short, where there's more variance in Cahill's, with some long launched and chipped passing.

A note: Koscielny actually had the best game in terms of pass attempts and efficiency in the same match; however, it was the season opener at Liverpool, so I didn't want to use it as it was his first with the club and I felt it might be an outlier due to his unfamiliarity with the Arsenal system.

The conclusions

I will say that I'm a bit surprised by what I've seen. I was expecting to see significantly less quality from Gary Cahill, considering the hype around him - if you couldn't tell from my title, I'd assumed that for the most part he was only rated as highly as he is because he's young and English, rather than actually being good. Based on these findings I think I've underrated him, but I still don't find him to be a worldbeater. In short, he's not just English, he's also actually good. But he's not Franz Beckenbauer or anything like that.

I think that it would take more adjustment than some think for Cahill to fit into Arsenal's system. Based on the data, he's not nearly as efficient in his passing as the current Arsenal central defense, and considering how important possession and accurate passing are to our system, that's a pretty significant issue. On the other hand, it's possible that the data are deceiving - consider the fact that the other players in Cahill's Bolton side are of vastly inferior quality to those playing for Arsenal, and some of the wayward passes may be explained. If Cahill had Cesc Fabregas or Samir Nasri on the other end of his passes, it's likely that his rates would rise. It's also likely that with an increase in the talent around him and an introduction to Wengerball, Cahill would attempt more passes.

On the other hand Cahill would fit in more than most other options. Aside from the reasons I discussed earlier, he clearly has quality (which might be amplified by joining a more talented side). He also plays a similar game in terms of positioning to what the current Arsenal defenders play, though not identical (not a bad thing, mind you - a little variance in the playing styles of players, as long as they fit the system, can be an added wrinkle for the opposition to deal with).

I don't think Cahill is quite as good as what we already have in terms of pure talent. However, he's better than much of what's available (for example, as awesome as he used to be, the occasional Sol Campbell re-return rumors leave a lot to be desired). He's played in nearly a hundred Premier League fixtures and knows the league, and like Koscielny and Djourou and even Vermaelen, he's still young enough to grow and improve.

We definitely need another central defender. If Wenger can get a relatively sane deal done, we could do a lot worse than Gary Cahill.