By Tony Attwood
Well actually I don’t know what the headline proves, other than the fact that even with my devotion to statistics, I too can find stats that don’t actually tell me much.
The same set of figures (from the Guardian) tell us that it is the centre backs who do the least running around in most teams. I think we probably knew that. You want them guarding the goal not lumbering around all over the place.
These figures also show that there is little relationship between crossing the ball and anything else. Huddersfield have done it most. It doesn’t even relate to the number of headed balls towards the goal.
Here’s another one: Watford have missed 25 big chances. (Big chances mean when the panel of experts putting this together thought that the player should score). Only one club has missed more big chances, and no it is not Arsenal. It is Manchester City with 26.
Ah but that is because City have more chances to miss. Yes, if you instantly said that you are getting the hang of this. Statistics are not nonsense, if you look at the parameters. But sometimes they can tell you nothing because other events are dominating the picture.
Here’s a good one: “Burnley have missed 20 fewer big chances than Watford while scoring one more.” Which says…? Well you can fill that one in.
Now if you are an avid reader you might recall that I have had a few bashes at expected goals, not liking the parameters that the whole notion is based on. In fact so annoyed have I got at some of this that I have decided to set up my own web site on this prediction lark. More details shortly when the site gets going.
But while we are on the subject of expected goals, Arsenal have scored 26 goals against an expected goals figure of 16.51. So we are on top of the league for converting expected goals into goals with a ratio of 157.48%. Burnley in second are mere amateurs at the game with 125.39%.
Which suggests to me that with that level of error, expected goals is not that meaningful a measure.
The problem with some of these stats is not that stats tell us nothing, but rather that we need to be careful what we are trying to show with one set of figures. Try the figures for touches of the ball.
| Rank | Club | Touches |
|---|---|---|
| 1. |
Chelsea
|
10,871 |
| 2. |
Manchester City
|
10,459 |
| 3. |
Liverpool
|
9,493 |
| 4. |
Arsenal
|
9,092 |
| 5. |
Tottenham Hotspur
|
8,857 |
| 6. |
Manchester United
|
8,313 |
| 7. |
Fulham
|
8,103 |
So how did Fulham get up there with the big boys whereas they are at the other end of the table? Ah, well, you see touches can be for all sorts of reasons – building up attacks on goal, or fannying around at the back when you shouldn’t be.
Try this one: the foot of the yellow card table
| Rank |
Club
|
Cards |
| 15. |
Huddersfield Town
|
17 |
| 16. |
Tottenham Hotspur
|
17 |
| 17. |
Manchester City
|
14 |
| 18. |
Newcastle United
|
14 |
| 19. |
Chelsea
|
13 |
| 20. |
Liverpool
|
12 |
Maybe Chelsea and Liverpool are by far the cleanest teams in the league. If so that is fine. If not, then…
Likewise the two bottom teams in clearances are Huddersfield Town and Manchester City, probably because with one of these teams not too many clubs burst through, needing the ball to be cleared, while with the other they just don’t get near the ball and so instead of clearing it, they let the ball get past them.
Manchester City also do the least tackles. West Ham do the most.
The point is, everyone can gather statistics, but for them to be of any use they have to indicate something. And not all statistics do that.
number of goals per tea ladies have always been a source of fascination for me.
I’m not sure that having a wide dispersion from the average negates the meaningfulness of a statistic. The normal distribution curve of men’s height would show that most men are close to the average but that there are significant minorities that are much shorter or much taller than the average. That does not undermine the fact that height is a useful measure of something and that it’s interesting to track it over time.
Maybe Arsenal are now (and maybe have been for a long time) better at scoring ‘unexpected’ goals? Did the two ‘goal of the season’ contenders against Fulham and Leicester come from ‘expected’ situations? Even when you look just at the final pass Ozil gave Auba a tap in (perfect situation?) while Ramsey turned a more problematic situation into a goal with an ‘unexpected’ improvisation.
Long distance free kicks rarely lead to a goal and are presumably counted as unexpected when they do go in. Granit has scored with two such efforts this season. Perhaps we score more unexpected goals and are maybe more fun to watch as a result?
I like your conclusion. I’ve always believed that. I used to wonder whether some writers/commenters on untold didn’t understand that when you argue things like spurs make plenty fouls but sit low in the table for red cards or Arsenal has more possession yet more goals against than other top teams. And then come to a conclusion that the stats don’t match due to ref bias.