Expected goals league tables, 30 Aug – 1 Sep 2019

These alternative league tables are calculated using expected goals (as seen in my match timelines). There’s a full explanation in the first post here.

Premier League

It looks like Man Utd could easily be vying at the summit with City if they’d fared better in the three games in which they out-created their opponents. Wolves’ position in the actual league table could well be misleading given that one of their draws could easily have been a win, while Crystal Palace’s win over Man Utd (one of the three in which United created the better chances) has left them higher in the table than perhaps they should be.


Swansea’s shock win over Leeds today extended their complicated record with expected goals: they sit top of the actual table but if the model settled every match using performance data then the Swans would have ten fewer points on the board. Leeds look to be the division’s stand-out performers, having out-created all six of their opponents so far, while Brentford have been the biggest underachievers.

League 1

Unfortunately we’re down to a 23-club table now and the heavy defeats being experienced by Bolton are a relatively fair reflection of their performances. There aren’t any huge differences from the real league table, although both Oxford and Southend look as though they could easily be four points better off. Rochdale meanwhile have only lost one league match so far but they’ve been significantly out-created in four games.

League 2

Exeter may be top of the real table but the data suggests that Grimsby and Swindon could easily be up there. Leyton Orient and Mansfield in particular look to be underachieving so far, while Macclesfield and Forest Green are sitting higher than the data suggests they should be.