Premier League Projections, Week 20: The New Big Four

Michael Regan

Simply put, my stats think there are four truly top teams in the Premier League and three second-tier contenders following them. I want to talk a little about how it comes about.

Tottenham Hotspur got a big win over Manchester United at Old Trafford on New Year's Day. The joke running around the Arsenal twitters was that it didn't really matter what happened in that game. My numbers suggest that joke was basically correct. Spurs and United aren't in any way ruled out of contention for top four, but I have the two with a combined top four chance of roughly 20%, and a combined title chance of under 1%.

In fact, it's not just Spurs and United. My table shows a huge gap right now between 4th and 5th place. I think the best way to present this data is to go against my usual methods and present the projected table first, then explain how it's come about.

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 82 +55 159 98% +1 0% 0 52% +0 2%
Arsenal 24 7 7 79 +35 142 92% +1 0% 0 24% -1 6%
Chelsea 23 8 7 77 +32 136 84% +9 0% 0 13% +3 10%
Liverpool 23 7 8 76 +41 150 82% +4 0% 0 10% -1 12%
Everton 18 13 7 68 +21 125 22% -10 0% 0 0.5% -0.5 30%
Tottenham Hotspur 19 9 10 67 +6 121 14% +3 0% 0 0.5% +0.5 23%
Manchester United 18 10 10 64 +17 124 7% -9 0% 0 0% -0.5 15%
Newcastle United 16 8 14 57 +2 98 0.5% -0.5 0% 0 0% 0 2%
Southampton 14 12 12 54 +7 111 0% -0.5 0% 0 0% 0 1%
Swansea City 12 11 15 47 +1 104 0% 0 0.5% +0 0% 0 0%
West Bromwich Albion 10 14 14 44 -6 91 0% 0 2% -4 0% 0 0%
Aston Villa 11 10 17 43 -12 79 0% 0 3% -5 0% 0 0%
Hull City 11 11 16 43 -8 79 0% 0 3% +0 0% 0 0%
Stoke City 9 13 16 40 -19 73 0% 0 9% -2 0% 0 0%
Norwich City 9 10 19 37 -27 70 0% 0 24% -1 0% 0 0%
Fulham 10 5 23 35 -36 60 0% 0 42% -21 0% 0 0%
Crystal Palace 9 7 22 34 -26 70 0% 0 46% +1 0% 0 0%
Cardiff City 7 11 20 33 -31 59 0% 0 54% +3 0% 0 0%
Sunderland 7 11 20 33 -27 77 0% 0 56% +14 0% 0 0%
West Ham United 7 11 20 32 -23 72 0% 0 60% +15 0% 0 0%

I have four clubs above 80% for top four, and then a chasm before we get Everton at a little over 20%, and the clubs below Everton, summed together, have about an equal chance of qualifying for the Champions League as Everton. That is certainly a finding. I checked the bookies' lines here, and they seem to roughly agree with me. They have Chelsea, City and Arsenal as basically nailed on for top four, and Liverpool a big step down, Manchester United close behind Liverpool, and Everton and Spurs another step back. The big disagreement between my stats and the bookies' is that I have Liverpool as way better than even odds to finish top four and United as worse than even odds.

So what's going on here? The actual table is much tighter. Liverpool are only one point ahead of Everton and two ahead of Spurs, while Chelsea, Arsenal and City are only 4-6 points ahead of the Reds. That doesn't look like a top four race that's anywhere near settled.

But if you look at the underlying stats, the gap begins to increase. The following table shows the key stats in shots on target, shot location, and "big chances" for all clubs in the EPL. I list attacking numbers first and defensive numbers second.

The Underlying Stats

Club DZ SiBoT W SiBoT SoBoT BC DZ SiBoT W SiBoT SoBoT BC DZoT Diff SiBoT Diff SoT Diff BC Diff
Manchester City 69 20 34 39 23 14 33 26 46 52 53 13
Chelsea 45 18 46 32 33 14 13 17 12 16 49 15
Everton 43 27 46 24 26 14 28 27 17 30 48 -3
Liverpool 48 31 41 45 23 18 39 18 25 38 40 27
Arsenal 56 20 34 45 19 13 43 20 37 44 35 25
Tottenham Hotspur 34 14 57 35 37 21 24 27 -3 -10 23 8
Southampton 43 14 28 18 29 12 22 28 14 16 22 -10
Swansea City 28 22 38 17 33 11 29 22 -5 6 15 -5
Manchester United 35 24 34 23 38 13 29 26 -3 8 13 -3
Newcastle United 31 17 48 24 39 14 30 28 -8 -5 13 -4
Aston Villa 26 11 35 24 44 10 26 30 -18 -17 -8 -6
Hull City 26 9 31 17 35 15 27 20 -9 -15 -11 -3
Crystal Palace 26 13 27 17 30 17 31 27 -4 -8 -12 -10
West Bromwich Albion 37 12 18 31 22 21 43 18 15 6 -19 13
Norwich City 25 12 35 19 42 24 35 31 -17 -29 -29 -12
Sunderland 25 11 27 24 35 14 46 28 -10 -13 -32 -4
West Ham United 27 14 20 21 43 23 30 32 -16 -25 -35 -11
Stoke City 27 6 27 29 43 16 38 30 -16 -26 -37 -1
Fulham 31 14 29 27 64 17 39 42 -33 -36 -46 -15
Cardiff City 25 6 23 19 49 14 45 33 -24 -32 -54 -14

The stats have Arsenal, Manchester City and Liverpool as the clear class of the league. They have the best DZ SiBoT and SiBoT differences in the league. The inferiority of Spurs and United is equally emphatic. Both clubs have allowed more shots on target from the danger zone than they've produced. They have outshot their opponents from lower expectation regions wide in the box and outside of the box, but even if we weight those shots on target equally (and I don't), Spurs and United stand clearly behind the top three.

For more on United's poor statistics, see also Colin Trainor's recent article on United at StatsBomb.

Explaining Everton and Chelsea is a bit more complicated. Chelsea's very impressive shot on target difference is something of an illusion, because José Mourinho's pressing style has led to incredible SoBoT prevention while allowing a pretty good number of higher quality chances. (For more on managerial effects on team statistics, see my piece from yesterday on managers, stats and AVB.) But it appears that Mourinho's defensive style has helped in two related ways. As I laid out in the linked piece, Mourinho has Chelsea allowing shots off through-balls at a rate about 50% lower than league average, probably because of the relatively deep back line he prefers to play. A similar effect shows up in big chances allowed, where Chelsea have been the stingiest in the league.

There are three primary factors in Chelsea's placement in the top four over Everton. The first is the prevention of super high-quality chances. The second is that my projections still include some 2012-2013 data, and Chelsea outplayed Everton last year. The third, and most important, is that Chelsea have a six-point advantage over the Toffees in the table. Everton only rate about 10 points worse in my "Team+" statistic, but combined with the state of the table, that's enough to create a big gap in the projected table. Liverpool, indeed, rate as a significantly better side than Chelsea, but that only makes up 90% of the gap between the two clubs.

The above table also helps explain a bit about my bottom-of-the-table projections. West Bromwich Albion have big positive differentials in both big chances and danger zone SiBoT. They are not meaningfully at risk of relegation unless something goes terribly wrong, and their victory over Newcastle on New Year's was a nice feather in the cap for statistical projections. Despite the huge gap between the two sides in the table, the Baggies and Magpies rate as close to even in overall quality, and so given home field I had West Brom favored in that match. Which they won. Hey, I don't celebrate too many things here, let me take this silly one.

Finally, you can see how difficult it is to identify the worst sides in the EPL. Fulham had been the worst, but their nice little win last week picked them off the bottom. Cardiff City look the worst now, but they are switching managers and quite possibly going to make some purchases in the window. Crystal Palace and Sunderland have not been all that awful, especially since Tony Pulis and Gus Poyet imposed their improved systems following managerial chances. I'm holding off on calling anyone terrible until the numbers and rosters sort themselves out.

In This Article

Teams
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,

By becoming a registered user, you are also agreeing to our Terms and confirming that you have read our Privacy Policy.

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.vc97ec6e

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