Serie A attack & defence: 2014/15
While I’m really happy that the scatter graphics I’ve been churning out for the past four years continue to be so well-received, I’ve long wished for them to look a bit prettier. During last season, @bootifulgame and I produced an experimental interactive version which looked much nicer and used striped “contours” to better distinguish which teams were performing better than others. It’s taken me a while to get around to it, but I’ve adapted the code to produce a static version that I’m pretty happy with. There are three graphics here and you can click any of them to bring up a full-sized version in a new tab.
As usual, I’ll start by looking at how the number of shots taken by each club compares with those they face in return. In this graphic, the average number of shots taken per match by each club is on the horizontal and the average number of shots 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:
Before we look at the positions of the clubs, I just wanted to flag what’s changed. The biggest structural difference between this and the old version is that rather than only shading the outlying areas of the graphic, the entire thing is now coloured in using the standard “green = good, red = bad” approach. The diagonal lines are basically contours where, in this case, the ratio between shots taken and shots faced is the same.
The axes are still centred on the average and one of the diagonal lines passes right through it – this is the line where shots taken = shots faced, so everything below it contains teams who take more shots than they face, with the stripes getting greener as they get more dominant, and everything above it contains teams who face more than they take.
Some observations to give you the idea:
- Unlike the German and Spanish top divisions, where a few sides dominated far more than the rest, there are six teams in the bottom right who all took far more shots than they faced in return. As we’ll see further down, it was quality and not quantity which decided the order in which these sides finished.
- Relegated Cesena took a lot fewer shots than anyone else, and allowed plenty in return, while Cagliari – who shared their fate – took and faced a relatively average number.
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:
The contours now show sides who scored goals at the same rate, so clubs in a greener stripe have scored more goals per match and vice versa.
- At the top of the graphic, Chievo‘s profligate attack stands out. The average of over 16 shots they took for each goal they scored is six higher than the average and over double what some of the most clinical sides – chief amongst them Lazio – required.
- Inter‘s disappointing season wasn’t down to a lack of effort – only Napoli created more chances – but plenty of other clubs were more clinical in converting them.
Finally 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:
The stripes now pick out clubs who have conceded at the same rate.
- Juventus‘ defence stands apart here – while three other sides permitted slightly fewer shots at their goal, nobody came close to soaking up as many for each goal conceded.
- Interestingly the second most resilient defence by this measure was 14th-placed Chievo‘s, although few sides allowed opponents more attempts at breaching them.
- While Napoli allowed the second-fewest shots, the ones that did get through went in almost as frequently as against relegated Parma – something that will definitely need addressing over the summer.
A quick note on the stripes
I’m still experimenting with how wide to make each stripe and to tie them to something tangible. At the moment I’ve picked numbers that give a sensible number of stripes, but I’ll make this more explicit in future posts.