Home » Tactics » How much has the Premier League changed?

How much has the Premier League changed?

Last week I contributed some charts and data to an article by The Economist’s James Tozer on the current season of the English Premier League. I thought I’d expand on a few of the things he covered.

On the general point of the league being a bit wacky, I think several forces are at work. There’s a wide variety of coaching philosophies on show, and clubs are at different stages in their recruitment and performance cycles. For example, managers of teams with a lot of young talent, like Mauricio Pochettino at Tottenham and Louis van Gaal at Manchester United, probably don’t expect to win the title this year – but they can certainly think about making a run next year. It makes sense for them to build a strong tactical foundation and suss out the most powerful combinations on the pitch, which may mean experimenting with different line-ups and formations.

A few other teams are also doing some slightly atypical things. Alan Pardew has been platooning his quick forwards – something he started doing at Newcastle – in a fascinating way at Crystal Palace. Chelsea was already Hazard-reliant last season and is even more monotonous this season. Throw in creative overperformers like Sam Allardyce and Tony Pulis, and you’ve got a very interesting league indeed.

But these are just observations – let’s focus on what the data can tell us. One of the big points in the article was about strategies for attacking. In the past several seasons, two relationships seemed to hold: (1) teams that used a slow build-up rather than fast, direct attacks tended to penetrate deeper into opposing defenses, and (2) the combination of possession speed and distance from goal pretty much explained the quality of a team’s non-header shots from open play. So far this season, both of those relationships have broken down:


Clearly, teams are finding more ways to affect shot quality besides possession speed and distance from goal, or besides the other variables for which those two are proxies. What’s interesting here is that in 2010-11 – where incidentally I trust the data the least, based on my experience cleaning them – possession speed and distance had the strongest relationship with shot quality. Yet that same season showed no relationship between the two explanatory variables.

In fact, four teams in that season stood well apart in their use of speed to penetrate defenses: Blackburn, Bolton, Stoke, and Wolverhampton. None of them managed to create shots of similar quality to Leicester’s, but none were relegated, either. In the following season, Blackburn’s attack slowed significantly, having lost two of its focal points: Mame Biram Diouf went to Hannover, and Nikola Kalinic went to Dnipro. Bolton, Wolves, and Stoke kept up the same pace, but their penetration was nowhere near as deep. All but Stoke were relegated, and the league seemed to give up on the fast-attack strategy for a few years.

This is why Leicester stands out. As one of the charts I produced for The Economist showed, Leicester have managed to create attacks with fast attacks that become just as dangerous as the attacks Manchester City creates with slow build-ups. Burnley also managed to penetrate reasonably well with fast attacks last season (hence the lower correlation), but it wasn’t able to create good shots the way Leicester has (hence the tiny change in the R-squared). This season, other teams have also tried a Leicester-like approach with varying degrees of success.

Of course, if you slow Leicester down, they become much less potent. And as I suggested in the article, one of the best ways to do this is to stop passes from midfield, especially by N’Golo Kanté. Both Daniel Drinkwater and Kanté are prolific passers to Leicester’s speedy front three, but Kanté’s passes – despite the fact that he plays somewhat behind Drinkwater – are deadlier. Moves starting in the first two thirds of the field are 20% more likely to lead to shots when he’s involved, and the shots are 70% more likely to score, based on historical averages. Drinkwater’s passes don’t have such a distinguishable effect.

Leicester’s fullbacks, especially Danny Simpson and Christian Fuchs, also play an important role in the fast-attack strategy. They occupy defenses, spreading them wide, so that the midfielders can pass the ball straight ahead. But congesting the midfield and forcing the fullbacks to advance the ball costs Leicester time; typically, an attack that goes out to the flanks has to cover a greater distance before the eventual shot than one that goes up the middle.

I’m curious to see how long it takes other teams to figure out and stifle Leicester’s attack. If they do, then the Premier League – possibly with the exception of Manchester United – might once again focus on the patient build-up as the primary path to attacking success.