Arsene Wenger’s Title Predictions Pt 2

Further to my last post, I’ve now got fascinated by the possibility of predicting the eventual number of points the league champions will win the title with by looking at the table after nine games (much more exciting than xG). Arsene Wenger alerted me to this with his prediction that this season’s champions would win it with 82-86 points, based on the fact that the leaders had 20 points after nine matches.

Dividing 20 by nine to get the points per game (2.22r), and then multiplying by 38 gives 84.44r, which tallies with a prediction of 82-86 points. However, doing this calculation for most seasons doesn’t make for a very accurate prediction. But the unusual thing this season is that the top three all had 20 points after nine games. If you average the points of the top three and then do the calculation you get a remarkably good correlation with the points of the eventual champions. Well, you do for recent seasons anyway. I’ve calculated it for the last 20 years – the Arsene Era, as Arsenal fans refer to it – and although the first few weren’t always close, in recent seasons it’s been very close. Here’s the graph to prove it:

predictions-graph

It’s been within two points in nine of the last 11 seasons, and only one point away in seven of those – though oddly never exactly level. If I was a betting man I’d be off to the bookies with this information.

Trying to predict the champions’ tally just by looking at the leader’s total after nine games is a waste of time though. Here’s the graph with that line added:

predictions-graph-2

Very little correlation there. Even if you deducted a set number of points each season to get the prediction line as close as possible to the champions line, the two are different shapes. This makes sense when you think that different clubs will have different periods of form through the season – the early pacesetter will often tail off, while someone who started only averagely often picks up more points later in the season to win it. But rarely does anyone keep up the same pace throughout. Another factor is that once teams have won the title they often relax and drop points in their final games when there’s nothing at stake, meaning the champions’ tally would on average be a couple of points higher.

Here’s the table of data if you’re interested (I know you’re not, but that’s fine, really):

points-prediction-data-table

Twitter: @AngryOfN5

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s