Brendan Sudol
September 2015

EPL standings scenarios

Football season is now underway and I couldn’t be happier. And of course I’m talking about the English Premier League. However, one thing that bugs me at the start of seasons is when pundits make seemingly crazy extrapolations about final league standings when there’s so much football still to be played. It seemed like people were already crowning Man City after week 5, and Chelsea has been completely dismissed after an (admittedly) awful start.

I wanted to see how fair these prognostications were, so I compiled some data and made a little scenario calculator. It’s very simple. You input how many games have been played so far and your team’s current position in the table (or their number of points or the number of points between them and the top of the table). Then you select where you hope for them to finish and it will tell you how often a team in that situation has gone on to achieve that final result over the last 20 years (1995-2015). Give it a whirl:

How often has a team ranked...
after games finished
How often has a team with...
points after games finished
How often has a team that's...
points off lead with games left finished

A couple observations:

  • Chelsea has a ton of work to do this season — only 3 teams with 7 points in their first 6 games have finished in the top 4 (and no one has finished higher than 3rd).
  • No team has ever won the league after falling behind the leader by 13 points or more, although 2 teams (in 20 years) have overcome a 10 point deficient halfway through the season to win it all (Man U in 1996 and Arsenal in 1998).

If you’re curious about the nitty-gritty data details, my IPython notebook is here and the final, cleaned dataset is here.