Ross Barkley and other thoughts

One of the interesting aspects of a new manager taking over at a club is seeing how players’ roles and performances compare to under the previous regime. In the case of Ross Barkley, it seems to be a bit of a mixed bag this season. He started each of Everton’s first seven games but was dropped against Manchester City, with Ronald Koeman calling it a “wake-up call”. Two games later Barkley produced one of his best performances in recent memory against West Ham but again was found missing, along with the rest of the team, in the 5-0 loss to Chelsea.

A number 10 who doesn’t create

A few years ago Barkley came on to the scene as a flashy and physical specimen with clear power and technique. The problem was that for a supposed central attacking midfielder, he created little:

We got lots of dribbling and good pass completion sprinkled with a handful of shots, but non-existent output when it came to goals or, more importantly, through balls, key passes, and assists. It should go without saying but if you are trying to be a number 10, you should occasionally help your teammates score goals.

The following two years showed incremental gains in key passes, assists, and shooting accuracy, while maintaining a good pass accuracy and high dribbling output. This seems to suggest a slow progression in key parts of his game, and the eye test more or less told the same story; undoubtedly he still frustrates but most would agree that he’s progressed rather than regressed since he broke into Everton’s first team.

These are half decent numbers for a 20-22 year old, but not enough to put him in the top tier at his position.

Last season was by far Barkley’s busiest—he was used in about 90% of all of Everton’s league minutes. Despite the slow progression mentioned above though, he still wasn’t much of a creator. Key passes were up, yes, but still in the Jonjo Shelvey range. Furthemore, it’s not a terribly great stat on its own, as it only tells you that a player passed to someone who then took a shot. Anyone who watches football will tell you that this can mean many different things on a case-by-case basis. It’s helpful but far from perfect.

We can dig a little deeper using an expected assists (xA) model, which is sort of to key passes and assists what expected goals is to shots and goals. That is, it takes into account relevant information about a particular pass to determine how “key” it really was, i.e. how often we could expect that pass to actually become an assist if we simulated it thousands of times.

What I’m getting to is that Barkley’s xA per 90s last year were pretty paltry for someone in his position (hovering around the edge of the top 50 last year, according to Paul Riley’s model). Ted Knutson’s model seems to have him a little better but still, look at his other “creative” numbers (through balls, key passes, passes into the box):

To really hammer this home, check out his pass map and xA map from last year as well, courtesy again of Paul Riley:

Note (a) the amount of sideways passing and (b) how there’s a bit of a force field around the top of the box. The xA map shows a bit of promise but that’s an entire season of data. Here’s Christian Eriksen’s map from the same period:

The point of all of this is that according to the numbers and charts and fancy man stuff, Barkley kind of just looks like a guy who dribbles around a lot, then either makes an easy pass in a position of no danger or launches a hopeful shot. He occasionally seems aware that maybe he ought to find a teammate in the box, but for the most part vision is not his forte.

The Koeman effect

This brings me to the current season. A month ago, Ted Knutson posted this:

A few things are happening here. The first is that dribbling, the one statistical category in which Barkley has consistently been top tier for three years, has completely fallen off the map. Secondly, his key passes are way up BUT expected assists are the same and through balls are way down.

It’s hard to know what to make of this. Did Koeman tell Ross to stop dribbling? Did Lukaku tell Ross to stop dribbling? How is he managing to rack up so many key passes without playing any through balls or really upping his creative output?

With this in mind I dug into a bunch of clips of Barkley from this season and my somewhat vague and half-hearted conclusion is that Ross’s style of number 10 is almost like an advanced water-carrier rather than a through ball merchant or killer passer in the mold of a Mesut Özil or Iniesta. In other words, he lacks vision to some extent but he does understand how to get into good positions and he likes playing quick one-twos with teammates who are close to him:

He seems more comfortable than in recent years finding and hanging out in those pockets of space in and around the box, which is a good thing. Decision-making is still an issue—I’ve yet to see him not take a shot when presented with the opportunity—but at least his shots are better and more accurate than they used to be. His shot chart mightttttt even just hint a bit at a player who is starting to learn that some locations are better than others:

So why isn’t he dribbling? I’m not sure I’ve given an airtight answer here but I think Koeman is trying to teach him to keep his head up and be a provider. The problem is that he still isn’t great at being a provider. That force field around the box that I mentioned above is fully back in action this year:

Sometimes you could argue here that a player’s teammates in front of him aren’t making clever enough runs or getting open. I don’t think that’s the issue here—Lukaku in particular is an active and willing runner who loves getting in behind. In any case, at the moment Barkley isn’t cracking Riley’s xA top 25 (both Lukaku and Bolasie are though, for what it’s worth).

To end on a high note, it is worth remembering that even if Barkley isn’t a pure creator, he does have good forward instincts for a midfielder and he’s in with some good company when you consider expected goals alongside xA:

And he can still do this:

For now I think we just have to accept that Barkley is a work in progress, but that his teammates and manager are pushing him to be the creative force we all want him to be.

Defensive midfield problems

Everton were torn to shreds for a few reasons this weekend but one of them was having Gareth Barry without Idrissa Gueye next to him. I mentioned in my Chelsea preview Paul Riley’s work in this regard, but essentially the idea is that Barry and Gueye have been getting pulled out of position way too much and that eventually it will come back to haunt the team. Barry is 35 now and I cannot for the life of me figure out why Everton is relying on him more than ever (90% of league minutes so far this year compared to 82-83% since he was signed).

He fouls more, he gets dribbled past more, he intercepts less, and he’s even hitting fewer long balls, which I thought was one of the reasons for letting him hang around. Tackles are up, yes, but tackles are a highly contextual stat, maybe not even desirable, and there are good reasons for not liking the context here. I said at the beginning of the season that the decline was real, and I’m not sure what’s been done to address it in the last three years besides hoping Muhamed Bešić can stay healthy.

Interestingly, Idrissa Gueye’s tackles are also up but his interceptions have decreased by 65%, which is A LOT. All of this to me supports the conclusion that he spends a lot of time running around like a chicken without his head because he needs to cover for Barry’s lack of pace. Consequently he is often out of position and cannot clog the lanes like he needs to. Anyway, if you haven’t read Riley’s article by now you should. It’s quite short and thought-provoking.

I’m going to go out on a limb here and say that this is going to a be a problem all season, especially against better teams, unless they sign someone in January or Bešić comes back and plays well.