Scatter graphics: Premier League, 18 Sep 2016

With five rounds of matches having been played in the Premier League I can just about justify chucking the data into my scatter graphic templates for a quick early look at how the top flight is shaping up. It’s worth flagging that at this early stage the relative difficulty of each club’s fixtures will definitely be influencing their performances, so please bear this in mind. I’ll probably revisit these again at around the 10 game mark, when I expect things to have settled down a lot more.

Each of the four graphics is explained briefly below but there’s a longer explanation here for the curious.

Shot dominance

First of all, here is how the number of shots taken by each club compares with those they face in return. The average number of shots taken per match is on the horizontal and the average number faced is on the vertical, so bottom right (take plenty, allow few in return) is good while top left (take few, allow plenty) is bad. The stripes are like contours: the greener the stripe, the better the performance (and vice versa for red).


What’s immediately interesting to me (and probably no-one else) is that the clubs are strung out in a diagonal line far more so than they would be in an EFL division i.e. the relationship between the number of shots you take and the number you allow seems far stronger than it is lower down.

Unsurprisingly we have the likes of Chelsea, Man City and Liverpool in the dominant lower right, although it’s interesting to see Southampton among them despite some high-profile summer departures and a poor start. The next chart will shed some light on this.

Tottenham have taken the most shots of all – more than double what Burnley (the quietest attack) have created – but their defence is more permissive than the teams mentioned above.

However Hull‘s defence is the busiest of all, having allowed around eight more shots per match than the average top flight team so far and almost three more than the next busiest back line.

Attacking effectiveness

Now let’s look at attacking alone. The horizontal axis stays the same as in the graphic above, but now the vertical shows the average number of shots needed to score each league goal. Therefore bottom right is good (taking lots of shots and needing fewer efforts to convert) and top left is bad:


So here at the top we can see Southampton‘s problem: they may take plenty of shots but have been around twice as wasteful as the average side at converting them so far. The silver lining for the Saints is that such an extreme level of profligacy is very unlikely to be sustainable, so they should drift towards normality in due course.

In the bottom left we can see that Watford have been the most clinical finishers, as three goals from ten shots against Man Utd demonstrated earlier today. So far they’ve needed fewer than five attempts to find the net on average, which is just as well given that only four clubs are shooting less frequently.

In the bottom left we unsurprisingly have leaders Man City enjoying the best of both worlds: while they neither take the most shots nor convert them the most efficiently, they’re the only side in the top three for both.

Defensive effectiveness

Next let’s look at the defensive situation – basically take the above chart and replace the word “taken” for “faced” on both axes. Now top left is good – facing fewer shots and able to soak up more per goal conceded – and bottom right is bad:


This time it’s Tottenham who are stretching the graphic at the top: their defence has soaked up 28 shots for each goal conceded so far. This is unlikely to persist – even Everton‘s also-impressive rate of almost 17 efforts faced per goal conceded would be difficult to sustain over a full season.

In the bottom left corner we can see that while Chelsea are allowing the fewest shots, only three clubs have soaked up fewer for each goal conceded. Over on the right hand side, Hull have faced almost three times as many shots but have dealt with them far better and hence have only conceded one more goal than Antonio Conte’s side so far.

West Ham‘s defence is – perhaps unsurprisingly on Saturday’s evidence – the leakiest of all so far, although Stoke – who also conceded four times this weekend – aren’t faring much better.

Expected goals

Finally here’s an attempt at correcting the first graphic for the quality of chances created and allowed, using the same “expected goals” values that power my shot timelines (explained here). The reason for doing this is that the results tend to correlate more strongly with performance than when we treat all shots equally. FYI there may be some differences with expected goals values published by other bloggers, who have access to far more detailed data than I do:


Again it’s not a surprise to find Man City in the bottom right corner: creating over two goals’ worth of chances per match and allowing fewer than one in return.

However while the league leaders have the most productive attack so far, Everton‘s defence has been marginally more effective at containing opponents so far. It’ll be interesting to see how the Toffees’ defence fares against City when they meet in mid-October.

On the left hand side it’s worrying how little Burnley are creating so far – only around half a goal’s worth of chances per match – and Middlesbrough aren’t faring much better up front. In the Claret’s defence, they have had Chelsea, Liverpool and Leicester to deal with already, so perhaps things will improve.