Width is underrated, or why some counter-attacks work better than others

This’ll be quick, and for the first few sentences I’m going to engage in a sin of journalism (not that this is strictly journalism) and use first person pronouns. Forgive me.

About the time that I started to hear about the world ‘halfspace’ – or perhaps it was about the time that I’d learned from stats Twitter how inefficient crosses can be – I noticed that Arsenal were avoiding the wings like they used to avoid 5th place. If anything, it was a little too much. But anyway.

Then plucky youngderdogs Monaco came along with their fiery brand of dagger-plunging counter-attacks. What made them overperform their Expected Goals tally by so much? Was it (as it seemed likely) linked to their efficacy on counter-attacks, and what made those counter-attacks so effective?

Counter-attacks like these:

This was originally intended to be a data-based project, like the work on crosses a while ago, but counter-attacks (where they develop enough to see them on TV) are far less frequent than crosses and time got in the way. But it became clear that Monaco kept pretty wide on their counters.

Watch the video again, and look at the spacing across the width of the pitch as the move develops:

Now check this one out:

Monaco’s counter-attacks were like this, looking like they had explicit instructions to occupy as many strips of the pitch as possible (by the way, if you’re unsure what halfspaces are, below is a picture from the great article on Spielverlagerung here). halfspaces Contrast that to Arsenal below. Though they have some width, Welbeck on the left never looks like a big part of the counter, Monreal is narrow enough that he’s pushed wide, and everyone else looks to be part of the central strip of the pitch.

The next example shows Sanchez trying to dive centrally at the earliest opportunity (and getting taken out, but the move continues), and Giroud occupying virtually the same central space as the man on the ball as it approaches the area.

The last example starts in a promising manner, but Oxlade-Chamberlain – stranded thanks to the lack of an Arsenal player in the right halfspace – cuts inside, and the whole move bunches centrally.

Take a look at this still of Barcelona’s 3rd goal against Real Madrid in *that* match where Real inexplicably decided to continue attacking in the final minute with a man sent off.

barca counter goal v madrid.png

If this were Arsenal, everything would be much narrower. Both wide forwards would, in all likelihood, be aiming their runs in a direction roughly towards the wider of Madrid’s 3 defenders. One of the other 2 wider players would probably be narrower too. Let’s mock up a doctored image

barca counter goal v madrid 2.png

While the Madrid backline is still hopelessly outnumbered, this scenario is far easier to defend. The wider defenders have their attention split less, and more importantly the angles of possible passes are much shallower. This limits the area of the pitch that the ball is able to be played into, and is more likely to push the receiver wide, slowing down the play. It’s better to start wider and arrive onto the ball moving inside than start closer to goal and adjust your run and speed by being moved off your line.

Anyway. Staying wider and occupying more zones of the pitch on the counter isn’t going to radically alter their efficacy, but there seems to be something to it. The pitch is how wide it is for a reason, Arsene.

EDIT: I’m going to add in some data even though it’s not as large a sample as I would have liked. I had a theory about counters in the centre being less effective than those in the halfspaces, but that was not borne out in the data (although may be in a larger sample).

Something which does suggest a link which I expected is the relationship between the amount of men per strip of the pitch and success – in other words, if your counter is clumped in the centre, will it be less successful.

Players per Zone
Outcome 1 1 – 1.5 1.5 – 2
Number 19 34 11
Shot 6 7 2
SoT 4 5 1
Goal 2 3 1
%Shot 31.58% 20.59% 18.18%
%Sot 21.05% 14.71% 9.09%
Conv.% 33.33% 42.86% 50.00%

While the conversion rates – number of shots turning into goals – rises, this is maybe/probably a symptom of a small sample of shots. The conclusion which is more likely to stick in larger samples is the chances, ie the shots, that these 64 counter-attacks create. There is, of course, the chance that this trend won’t hold up in a larger sample, but given that it intuitively makes sense I’m willing to place a small amount of faith in it.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s