Norwich City v Liverpool - Barclays Premier League

Liverpool Football Club has worked diligently all season toward their first league title since 1990. Now, at the end of April with the team on 80 points, five ahead of Chelsea, it is realistically within their grasp. Should they win the prize that many couldn’t over the past two and a half decades, they will have earned every last inch of the trophy.

Even so, Liverpool’s triumph will have been aided heavily by some fairly unsustainable numbers, numbers which should spur Brendan Rodgers to action in the club’s off-season.

On the surface, Liverpool’s dominance is something a statistical outlier (“That’s cuz those numbers don’t explain passion and romance! etc. etc.”).

Why? Well their shots ratio, which has held steady most of the season, is ranked 5th in the Premier League (for an explanation of all these concepts, read this piece).

Meanwhile Liverpool’s PDO (sh% plus sv%, a measure of good luck when well above the median of 1000 and bad when well below) is high but not screamingly high at 109.47 (thanks Ben).

At first glance, it’s difficult to see where Brendan Rodgers is getting his secret sauce. Taken together, based on how well TSR correlates with points (and how quickly PDO regresses to the mean), these numbers shouldn’t really put Liverpool top of the table on 80 points with three games to go. This is a low probability event.

I’ve posited before that Game States might hold some of the answer. LFC spend a lot of time winning, which means a lot of time at at least a +1 game state. What does that mean? Well teams at +1 tend to take fewer (but generally more accurate) shots and concede more, which would in theory lower their overall TSR. Yet Man City have spent even more time winning than LFC and their TSR is sky high, commensurate with a first or near first finish.

Maybe that’s because City build up leads, which tends to diminish the GS effect (at +2, things start to revert back to normal). Or maybe City’s TSR is skewed by a small number of games in which they wildly outshot inferior opposition.

Maybe though something else is going on, and this is where we arrive at this pair of tweets from James Grayson which I nearly missed yesterday, but that carry a difficult truth for Liverpool FC.

Shots on target percentage is exactly what it says it is—the percentage of total shots which are on target, for and against. You can measure this percentage against your average opposition SoT%, and come up with a differential. Liverpool’s is evidently quite high.

Why does this matter? Well as Grayson worked out last year, SoT% is roughly split between luck and skill. As Grayson notes in the second tweet, In Liverpool’s case it would explain why a team with a mediocre TSR and a not-that-high PDO is rocking it in first place.

The temptation in Liverpool’s case would be to look at LFC’s very high SoTF% and their very low SoTA% and conclude, “It must be something Brendan’s doing on the training pitch.” This was my line of thinking yesterday when I speculated on why LFC’s differential is so high.

Yet this would only make sense if SoT% for and against was consistent at the team level year over year. So then, here’s the bad news, at least as far as next season is concerned:

In other words, their SoT%s are going to regress. And if LFC post a similar TSR and PDO next season but with a much lower SoTF% and SoTA% differential, they likely won’t finish near or at first place.

Does this mean Liverpool aren’t worthy of their Premier League trophy, should they win it? Of course not. You win it on points, not on abstract probabilities.

What it does mean is that the team should consider seriously strengthening their squad ahead of next season.

Regression is never a popular subject in sports, let alone football. No one wants to believe their team is winning on the back of random variation. We want everything on the pitch to be explainable, even if we know, like all football-loving folk, that luck plays a big role from game to game, hence the mantra about championship teams winning matches “they should have lost.”

Liverpool will deserve their PL trophy, should they win it. But they shouldn’t fall prey to complacency. Look what it did to Moyes’ United.