clock menu more-arrow no yes mobile

Filed under:

Hot stat takes: On Manchester United, Chelsea, City and Liverpool

I take a look at the underlying shooting statistics from this week's big matches in the English Premier League. In a few cases I think the initial narrative is not quite right.

Michael Regan

This article was inspired by two matches and what I feel are somewhat misleading narratives that arose from them. In both cases, I watched the whole match and I found later that the numbers generally supported my subjective observation. With a couple of these matches, I didn't get a chance to watch, so I want to be a bit more circumspect in making any strong claims. It's always possible for the shot chart to mislead.

But I think that in the case of both Manchester United vs. Fulham and Tottenham Hotspur vs. Everton, the shot chart gives us the pertinent information.

Manchester United vs. Fulham

The story of the weekend is United's failure to take all three points against bottom-dwelling Fulham. They attempted a league record 81 crosses, which led to a fun time for all the nerds on twitter counting them in real time.

But here's the thing. Manchester United created 10 shots with those 81 crosses. Four of the ten were from within the six-yard box. This season in the English Premier League, there have been 11217 crosses attempted and 1281 shots assisted by crosses. United created shots with 12% of their crosses, compared to a league average of 11%. A bunch of people noted how Fulham's defensive scheme was based on allowing United space in wide areas to work crosses into the box. This is a good strategy, if you then complete the job and prevent those crosses from producing good chances. What I think has been missed is that Fulham did a pretty poor job of preventing shots off the crosses into the box they allowed. Four of the crosses were even completed within the six-yard box, which is the highest total of shots off crosses from the six-yard box all season.

Ultimately, I think the Red Devils were mostly unfortunate not to take all three points. You can point to some defensive frailty in allowing two big chances, both of which Fulham converted. But the attack looked dangerous. They crossed a lot, yes, but they crossed effectively. Only a blinder from Marcus Stekelenberg plus a few less-than-perfectly dispatched chances by the usually dead-eyed Robin van Persie and Juan Mata kept the match from being a laugher.

These are the basic shot totals for the match. "DZ" is the danger zone, the central area of the 18-yard box. "Wide" is the wide areas of the 18-yard box. "SoB" is outside the box. "Cr" are shots assisted by crosses, while "TB" are shots assisted by through-balls. For more on these terms, and what they mean in terms of expected goals, see my first three shot matrix articles (Shot Matrix IShot Matrix IIShot Matrix III).

To give a general sense of the values, shots from the danger zone are converted at a rate of about 1-in-6 (16.4%). Shots from wide areas in the box at a rate of about 1-in-16 (6.1%) and shots from outside the box at a rate of about 1-in-30 (3.3%). Shots assisted by crosses are slightly more than half as likely to be goals as regular shots from the same areas, while shots assisted by through-balls are between 50% and twice as likely to be converted for goals as regular shots.

Club S DZ S Wide SoB S Cr S TB
Manchester United 17 3 11 10 0
Fulham 4 0 2 0 0

That is not a close match. 17 DZ shots actually slightly underrates United's dominance. 7 of those 17 shots were taken from within the six-yard box. That's the highest total of the season, no other club has managed more than four shots from the six-yard box in a single match.

Tottenham Hotspur vs. Everton

The story of this match in the media seems to be about unlucky Everton and the one moment of magic from Emmanuel Adebayor that undid all the good work of the Toffees. My take watching the match was that the first 15-20 minutes were all Everton, but it was a messy, kind of violent stalemate from that point on. Spurs defenders (especially Jan Vertonghen) prevented Everton from creating anything too dangerous in the opening flurry, and Everton were toothless afterward. Spurs didn't create much, but had a couple good opportunities and Adebayor converted one.

This looks like a 0-0 draw (maybe 1-1) in the stat sheet and shows the value of having an excellent striker to convert that one chance.

Club S DZ S Wide SoB S Cr S TB
Tottenham Hotspur 3 1 4 2 0
Everton 4 2 5 2 0

Chelsea vs. Newcastle United

The story out of this match seems to be Chelsea's easy dominance and Newcastle's continuing incompetence since Yohan Cabaye's exit. I did not watch this match, and so I'm less confident bringing my hot stat takes to bear. But in the stat sheet, it looks pretty even. An edge to Chelsea, but when you consider both home field advantage and Newcastle's generally lower quality overall, I don't see either a particularly impressive performance by the Blues or that much of a howler from the Magpies.

Club S DZ S Wide SoB S Cr S TB
Chelsea 5 4 8 1 0
Newcastle United 3 2 7 1 2

Chelsea converted two of their five shots from the danger zone, and added a penalty. Newcastle converted none of their three DZ shots and neither of their two shots off through-balls. It's the latter stat that makes me think the stat sheet has some utility. Newcastle split Chelsea's defense for a shot twice and just didn't convert. If it's just a scoreline effect, you wouldn't expect to see Newcastle getting runners in behind Chelsea's back line.

Liverpool vs. Arsenal

I have no hot take here. Liverpool were rampant. I have xG around four goals, depending on the method. Arsenal got a few chances, but none of particularly high quality. In my projections, this match has taken Liverpool nearly up to even with Arsenal in expected points.

Club S DZ S Wide SoB S Cr S TB
Liverpool 12 3 7 4 2
Arsenal 3 0 7 2 0

Norwich City - Manchester City

This one is worrying for anyone who, like me, has been talking up Manchester City as title favorites. They should have won easily at Carrow Road, and instead this isn't too far from a pretty fair draw. Probably 0-1 fits the stat sheet better, but even 1-0 is not much of a performance for a title-winner. City had a few more chances, but there isn't a lot here. I'd expect that the chances City created chasing the game were somewhat lower in average quality. Three of the five danger zone shots were assisted by crosses, for instance, and none came from within the six-yard box. City low rate of shots on target (1 SiBoT) suggests as much, too.

Club S DZ S Wide SoB S Cr S TB
Norwich City 1 3 3 1 0
Manchester City 5 3 7 3 0

Projections to follow, probably.