Creating Chances v. Taking Them: The Premier League in 2009/10

Ever since I started calculating shot to goal ratios as well as conversion and accuracy ratios of teams' offensive (and defensive) production, I've been concerned that these ratios do not tell the full story of what happens on the field. Conversion, you may remember, is the ratio of goals to shots on target - that is, how many of the high quality chances teams generated actually resulted in goals: how good teams are at finishing. Accuracy is the ratio of high quality shots to all shots teams took: how good teams are at generating high quality chances.

But I've always thought that there's a difference between Team A that finishes 1 of 5 chances and Team B that finishes 4 of 20. While both have a conversion ratio of .2, I'd much rather be playing on, coaching, and watching Team B, wouldn't you?! Turn it around: if you're coaching Team A, you know that your strikers are doing well converting their chances; but you are not generating sufficient opportunities for them. Hence, you may want to work on midfielders finding ways to get the ball in front of goal where we know the vast majority of high quality shots are taken.

So I was thrilled to see this clever post on the On Football blog, which took my concern one step further and actually did something with it, using MLS data. The author distinguishes between creating chances (shots) and converting them and plots these two dimensions of offensive production. What you see is that the best teams cluster in the upper right hand corner of such a graph: the best teams in MLS create lots of chances and they take them.

So can we generalize this pattern to other leagues? Since I had nothing better to do (well, I did, but I can't tell you about that), I grabbed the 2009/10 dataset for the Premier League and replicated On Football's graph for the EPL. Take a look at the average levels of creation and conversion by team for the 2009/10 season.

Four things stand out. First, just like in MLS, the top teams that year cluster in the upper right hand corner. Second, the struggling teams cluster in the lower left hand corner - they don't create many chances and they waste most of them. The rest of the league is somewhere in the lower right hand corner: these teams don't create as many chances as the top teams (in fact about as many as the bad teams), but achieving higher levels of finishing (with some, like Aston Villa or Stoke as high as Chelsea). Four, teams do not cluster in the upper left hand corner; there are basically no teams that create lots of chances and then don't take them.

Congrats to On Football's authors on moving things forward. Before too long, I'll also take a look at defensive production along these lines and perhaps across leagues. Stay tuned.

PS: One small note: conversion rates appear significantly higher in the English league than MLS.