The 3rd match on the 13th of February of the Premier League (or the closest date to the 13th that matches were played).
The above is the first thing I wrote of this article. For a while I’ve thought that the amount of clearances that teams make in matches, and places where they make them, can be used to tell the story of the match, in a similar way to shots.
Both, on the surface, have obvious flaws. You can shoot from anywhere and it’ll boost your shot numbers, and you don’t know how good these shots actually were. Similarly with clearances, many clearances might show good defending* or it might show opposition pressure or both, it might be hard to tell which is which.
I picked a random match to try and use clearances to tell the story because I figured this would help reduce a level of confirmation bias. If I’m picking the match myself, I’ll probably end up choosing one that shows how useful clearances are.
Anyway, the match in question turned out to be Chelsea vs Everton (1-0, from the 11th of February). I do not remember it (I see now that Matic was WhoScored’s man of the match, maybe this has something to do with it).
I go to look at the possession. Chelsea had 58%, Everton 42%, so Chelsea obviously dominated to a degree.
The shots back this up:
Chelsea had 21 shots to Everton’s 7, with a good handful of Mourinho’s side’s being in the ‘danger zone’, central in the box.
You get something similar with the clearances:
Both show Chelsea as dominant. Clearances increases both sides’ in-box presence, although this may partly be due to defending corners, where clearances are likely to happen in dangerous looking areas of the box (although the same goes for shots, if you get on the end of it).
I’ve made some observations about clearances during my year or so looking at defensive stats. These aren’t backed up by thousands of matches of data, but I think they *generally* hold true.
Clearances made in the dark green area have a fairly good chance of having been headed clearances by defenders (from long, maybe hopeful, balls forwards), and I think the more dominant your team performance, the more likely they are to be headed (if you’re under pressure yourself then clearances in this area are more likely to be ‘genuine’ on the ground clearances, and you’re also more likely to be clearing away hopeful hoofs if your own team is dominating – these are guesses).
The yellow area are clearances that are bordering on irrelevant, having a higher likelihood of being made by wingers or full backs from throw-ins, for example. It is, therefore, somewhat intuitively, the red zone that is most important.
So, if we’re to glance at the clearance map and tell the ‘story’ of the game, we can see that Everton had some presence in Chelsea’s box (1), but given the relative lack of clearances elsewhere (2) we can be suspicious of this and tentatively guess that corners had something to do with it (I checked, it did). Everton were also clearly bombarded (3) and the fact that a lot of their clearances were so deep (as shown by a rough, vertical line of best fit of their clearances, 4) shows just how pressed back they were.
Is there anything there that we couldn’t glean from looking at the shot map? Maybe not. We can see Everton had some presence in Chelsea’s box, but that most of their shots were from outside. We can also see that Chelsea dominated, and that Everton must have been pressed deeply back into their own territory.
So what does this mean for clearances? To be honest, I don’t know. The random match didn’t show anything great in support of taking more notice of clearances, but it also didn’t tear down the idea completely. It also gives a sense of where the team’s defence is in one handy map, almost like a defensive counterpart of the shots map, the other side of the coin.
(You can, of course, get maps of all of the team’s defensive actions, but they are generally more confusing, with tackles and interceptions across the pitch. I think clearance maps, like shot maps, are less messy and that comes with its pros and cons).
The use of the clearance maps in practical terms, probably, isn’t much. They might help you get an idea of how a game went if you missed it, but you can generally watch highlights, read a match report, or your Twitter timeline. It might give you a sense of whether a team was hanging on for a result, or just couldn’t break through to get one, but, as I’ve said, shot maps can give you this sense too.
Basically, this is an idea that’s been knocking around in my head and I thought I’d have a small bit of a closer look and report my findings, whatever they turned out to be. Of course, neither shot nor clearance maps are a substitute for watching the match analytically yourself, nor do I think clearance maps will really change how people think about the game. This is more of a tossing of the clearance hat into the ring and pointing to it, saying “Hey, what about this?”
*(For the record though, in my previous work with defensive stats, I’ve come to the conclusion that clearances aren’t a measure of good defending, at least on an individual level, even when possession adjusted, you’ve got to juggle defensive stats a lot more than that to get something worthwhile. Obviously match to match everything that stops an attack is good, but I believe clearances tell the story of opposition pressure more than individual or team defensive strength).