Thanks to Chris Anderson who retweeted this table yesterday, along with whomever it was who took the time to put it together.

The points per game part of this table doesn’t interest me in the slightest, but the amount of money teams are spending on wages certainly does. In total the teams are spending £1,623M on wages this season. Moreover Premiership teams have had a net transfer spend of £352M this season (many thanks to Ravi Ramineni for directing me to this link). As such, the Premiership teams have spent £1,975M on transfers and wages in the last year.

Now one way to determine the value of a point in the Premiership you’d take the £1,975M and divide by 1,042 (the average number of points scored in a season (link)), giving you a total of ~£1.9M. However, I’d posit that this is too simplistic a model, and were it to hold true then we’d expect the top teams to be scoring more points than is possible. In reality, teams can achieve a baseline performance with minimal spending, and paying for the points beyond this baseline is where the real cost lies.

Instead I’m going to take a different route. We know, for example, that the teams promoted from the Championship score an average of 38 points in their first Premiership season (link), and that they do so in a pretty unpredictable manner. It also turns out that the three teams promoted last time round, (So’ton, Reading, West Ham) are very representative of that group, given that they’ve currently combined to score 80 points in 78 games. So lets see how much they’ve spent to hit 38 points this season (all numbers in terms of £M):

As a group they’ve spent an average of ~£64M to collect an average of 38 points each, at a rate of ~£1.7M per point. Extrapolate that across 20 teams and the first £1,280M is being spent to gain 760 points (38 points per team). Subtract those numbers from our original tally and we’re left with £695M to spend on only 282 points (1042 minus 760), an average of $2.46M per point.

I also happened to come across the data for Premiership wages in ’10-11 thanks to the excellent Swiss Ramble (link). Now if we repeat the exercise by looking at the promoted teams that season (Newcastle, West Brom, and Blackpool), once again factoring in their transfer spending, then we find that their average spend was a mere £32M. By following the same steps as above we find that the 282 excess points are being won at a price of ~£4.66M each.

Now I’d suggest that these numbers provide a range and that the truth is somewhere in between. For one a sample size of six isn’t enormous, and is open to can be impacted by a club record sale (Andy Carroll), or a club record signing (Gastón Ramírez), but it gives us a starting point, which is fitting given the current state of football analytics right now.

Finally, to apply this to the Premiership this year then this is how we’d expect the league to finish based solely on wages and transfer fees if we estimate it costs £48M to score 38 points (the average spend of these six teams), and the price of an excess point is £3.5M.

And you know what, I don’t think there’s all that many surprises in there. It highlights just how well Swansea and Everton are doing though, and, on the flip side, how poorly Villa and QPR have fared this season.

If anyone knows of more Premiership wages data I’d love to expand the data set. And I’m always willing to listen to suggestions as to how this can be improved.

Interesting read! I’d suggest that spend and wages for the previous seasons should also be factored in, in a weighted fashion (dependant on how long ago). After all, no team comes into a season with a net value of £0!

Intuitively it feels wrong to include player sales in your measurement of transfer spending. Since we’re not measuring a change from year to year, it seems irrelevant to factor in the money earned for selling someone. For example, Arsenal’s sale of RvP and Song puts their net transfer spending at a -3, even though they have no bearing on the points earned by Arsenal this season. Presumably you’re not offsetting the amount of wages they’re spending by subtracting their wages from the wage part of this equation, right?

It seems like you’d get a better picture of what teams are actually spending if you measure the transfer fees like you do the wages – add up the total transfer cost of everybody on the team, rather than the net transfer spending for the year.