Premier League Projections, Week 19: Return to shots on target

Paul Gilham

Tottenham Hotspur won, but so did basically everyone else around them, so Spurs still find themselves on the fringes of contention. I take a look, eight matches after the Newcastle debacle, at the club's shot on target conversion numbers. Turns out that we've been doing just fine since Newcastle. Well, except for all the other things.

Remember the innocence of November? Tottenham Hotspur were among the very best defensive sides in the Premier League, held back by only one little statistical indicator. If we could just convert our shots on target into goals at normal rates, why, we'd be unstoppable. True title contenders. Then the last six weeks happened, injuries abound, a manager sacked, a crazy new manager hired, and a once-great defense collapsing into a puddle of goo before our eyes. No one's mentioned shot on target conversion in part because it seems the least of the club's problems. But also because it actually isn't one of the club's problems.

The following table lists the data on shot on target conversion this season, broken down into two parts. The first eleven matches of the season (for Spurs, that's through Newcastle), and then the next eight. I list xG, expected goals scored based on shots on target from my Shot Matrix numbers, and actual goals scored, excluding own goals and penalties. The "Diff" column is the number of goals by which the club under- or over-performed expectations in their finishing.

Club xG 11 G 11 Diff xG 8 G 8 Diff
Arsenal 16.8 21 4.2 11.2 14 2.8
Aston Villa 9.3 10 0.7 6.8 7 0.2
Cardiff City 8.4 9 0.6 4.4 5 0.6
Chelsea 16.7 16 -0.7 11.9 15 3.1
Crystal Palace 6.8 5 -1.8 8.9 6 -2.9
Everton 14.1 14 -0.1 15.9 17 1.1
Fulham 9.1 10 0.9 8.3 6 -2.3
Hull City 6.6 6 -0.6 9.9 12 2.1
Liverpool 19.9 18 -1.9 14.0 18 4.0
Manchester City 20.8 26 5.2 15.5 23 7.5
Manchester United 13.6 17 3.4 10.4 11 0.6
Newcastle United 14.2 15 0.8 10.6 11 0.4
Norwich City 9.7 8 -1.7 6.9 7 0.1
Southampton 13.3 13 -0.3 8.8 11 2.2
Stoke City 9.4 9 -0.4 8.5 8 -0.5
Sunderland 6.9 7 0.1 10.3 6 -4.3
Swansea City 14.7 12 -2.7 5.9 6 0.1
Tottenham Hotspur 16.9 6 -10.9 7.7 10 2.3
West Bromwich Albion 10.5 11 0.5 9.3 9 -0.3
West Ham United 9.1 8 -1.1 8.0 7 -1.0

No one had conversion numbers so terrible in the first eleven matches, as we had missed out on nearly a goal scored per match because of bad finishing and bad fortune. Since them, it's all been basically fine. Spurs' attack has been much less impressive in terms of actually creating chances and putting shots on target, but those that we have created, we've converted at above-average rates.

Shot on target conversion simply hasn't been a problem for this club since the Newcastle match. It's just all the other problems (note again the large decrease in expected goals) that are the problem.

I do note that Manchester City and Arsenal have been pretty consistent in overperforming xG. This could be a fluke, but it could also be a combination of high-quality finishing and game state effects. While Arsenal's strike force are hardly the most clinical bunch, City's are pretty good I've heard. Both clubs have spent a lot of time in the lead, and clubs in more advantageous game stats tend to convert a higher number of their shots. This latter issue has been studied most by Ben Pugsley of Bitter and Blue and Statsbomb, though I did a little study of my own last summer. Right now it's just a hypothesis that the difference here might not be just random variation, and it's something I want to look into.

But to go back to Spurs, the club has scored 2.3 goals more than expected based on shots on target since Newcastle. The problems with this club lie elsewhere, I think.

Projections and Power Rankings

Because of rounding, not all the numbers necessarily add up quite right.

Club W D L Pts GD Team+ Top4% ΔT4 Rel% ΔRel Title% ΔTitle 5th
Manchester City 25 6 7 81 +55 159 97% +0 0% 0 52% -4 2%
Arsenal 23 8 7 78 +34 142 91% +4 0% 0 25% +7 6%
Chelsea 22 9 7 76 +28 136 75% +5 0% 0 10% +6 14%
Liverpool 22 8 8 75 +44 150 78% -4 0% 0 11% -3 12%
Everton 19 12 7 69 +21 125 32% +1 0% 0 1% +0 26%
Manchester United 19 9 10 66 +18 124 16% -1 0% 0 0.5% -0 19%
Tottenham Hotspur 18 10 10 65 +5 121 11% +2 0% 0 0% +0 14%
Newcastle United 17 8 13 58 +4 98 1% -2 0% 0 0% 0 3%
Southampton 15 11 12 56 +10 111 0.5% -1.5 0% 0 0% 0 2%
Swansea City 12 12 14 48 +1 104 0% 0 0.5% -0 0% 0 0%
Hull City 11 11 16 43 -8 79 0% 0 3% -9 0% 0 0%
West Bromwich Albion 9 15 14 42 -8 91 0% 0 6% -1 0% 0 0%
Aston Villa 10 11 17 41 -15 79 0% 0 8% -1 0% 0 0%
Stoke City 9 12 17 40 -20 73 0% 0 11% -2 0% 0 0%
Norwich City 9 10 19 37 -27 70 0% 0 25% +0 0% 0 0%
Sunderland 8 11 19 35 -25 77 0% 0 42% -7 0% 0 0%
Crystal Palace 9 7 22 34 -26 70 0% 0 45% -6 0% 0 0%
West Ham United 8 11 19 34 -20 72 0% 0 45% +1 0% 0 0%
Cardiff City 8 10 20 34 -31 59 0% 0 51% +4 0% 0 0%
Fulham 9 5 24 32 -38 60 0% 0 63% +20 0% 0 0%

  • Arsenal took another big win away to a good mid-table club. In this match, they did not allow a single SiBoT or a single shot of any sort from the danger zone. They were projected to take maybe 1.5 points, they took the full three instead. That bumps Arsenal into closer title contention as Manchester City were uninspiring in victory over Crystal Palace. I wouldn't read too much into the the drop in City's title projection—I think 60% is just about the highest possible title% at this point in the season for a club without a huge lead on the rest of the table. So anything that isn't great news for City will lead to them losing a couple percentage points.
  • Liverpool had a chance in this period to make a real run at the title, but I think people giving up on them as title contenders are pre-mature. They project to the same average number of points as Chelsea (the Reds are fourth by fractional points), and they have a slight advantage in title and top four percentage because of their superior goal difference.
  • With Spurs traveling to Manchester United on Wednesday, I've been tracking United's numbers, and they have not been good recently. In victory over Norwich City ,United did not attempt a single shot from the danger zone. I have both of United's last two victories as projected losses based on the underlying stats. Hull City dominated United with five shots on target from the danger zone to United's one, and Norwich had four shots from the danger zone against United, two on target. While obviously the Red Devils did not log an attempt from the close/central area. My numbers will not project a Spurs win, given the importance of home field advantage, but I don't think we should read too much into the "run of form" that United may seem to be on.
  • Speaking of Hull, I already mentioned the stats on their most recent victory in my piece on managers. Seven danger zone SiBoT plus seven SoBoT for Hull, with Fulham registering only one shot on target from inside the box, from a wide position, plus two SoBoT. Hull City have sprung to life over the holiday period and look very nearly safe from relegation. Someone should mention that it looks like Steve Bruce is doing a really good job.
  • Down in the relegation battle, Sunderland picked up another solitary point in a match they should have won. Sunderland created three big chances to Cardiff City's one, as well as six SiBoT to two for Cardiff. My numbers are growing more and more impressed with Sunderland even as they remain 20th in the table, and I have them now projected fractionally outside the drop zone.
  • Should I do it? I think I'm gonna do it. Here goes, see how this feels...
  • Fulham are terrible.

X
Log In Sign Up

forgot?
Log In Sign Up

Forgot password?

We'll email you a reset link.

If you signed up using a 3rd party account like Facebook or Twitter, please login with it instead.

Forgot password?

Try another email?

Almost done,

Join Cartilage Free Captain

You must be a member of Cartilage Free Captain to participate.

We have our own Community Guidelines at Cartilage Free Captain. You should read them.

Join Cartilage Free Captain

You must be a member of Cartilage Free Captain to participate.

We have our own Community Guidelines at Cartilage Free Captain. You should read them.

Spinner

Authenticating

Great!

Choose an available username to complete sign up.

In order to provide our users with a better overall experience, we ask for more information from Facebook when using it to login so that we can learn more about our audience and provide you with the best possible experience. We do not store specific user data and the sharing of it is not required to login with Facebook.

tracking_pixel_9353_tracker