Squawka’s Eric Bailly comparison matrix, and what’s wrong with it

Firstly, this isn’t meant as a venom-tipped jab at the writer of Squawka’s ‘How Eric Bailly compares to Manchester United’s current centre-backs’ article, or, particularly at Squawka themselves. You gotta create content, I get it. Not everything that the collective power of stats twitter puts out is gold, and I’m sure a whole lot of dross would get put out if we had to produce large amounts of content consistently every day in order to get paid.

But.

Eric Bailly comparison squawka

  1. Interceptions have essentially no correlation with the amount of shots you concede – see here. Using them, without adjusting them in a meaningful way, says more about a centre-back’s style than a centre-back’s quality. I can say that with fair certainty, as they say absolutely nothing at all about a centre-back’s quality.
  2. Blocks. I’ll be honest, I haven’t looked at blocks in proper detail. However, if you play in a low defensive line or block (as opposed to a high line) you’re likely to get in the way of more shots. Also, if you’re on a worse team, you’re probably going to end up blocking more shots too. I think Villarreal’s back-line play deeper than United’s, which may explain why Bailly’s blocks are higher.
  3. Tackles won. A ‘tackle won’ and a ‘tackle lost’ are both successful tackles, it’s just that tackles lost happen to go to an opposition player. I haven’t looked at ‘tackles won’ specifically, but making more tackles only has a slight correlation with conceding fewer shots. Again, see here. Again, it’d probably be better to use them to assess style instead of quality.
  4. Defensive errors mean fuck all. It’s a subjective measure that no-one in stats twitter circles seems to fully understand yet is used by fans as proof that players are good or bad because it SEEMS LIKE the most easily understandable defensive statistic. Part of this comes from the name. Neither Chris Smalling nor Marcos Rojo made ANY defensive errors last season? Piss off. (Post-publishing edit: I have not yet done proper work on whether defensive errors, statistically speaking, mean ‘fuck all’, but I strongly strongly suspect that my reasoning is sound).

 

This article is supposed to be educational, and I don’t blame people for not knowing the above, because none of it is necessarily obvious or something that you can easily do for yourself.

The best statistical minds in football have trouble making any sense of defensive statistics for defenders (maybe the ones operating in the dark at clubs have less trouble than those blundering in public, but who knows). You can make statistical comparisons between strikers fairly easily without, say, being a regular Statsbomb reader and you probably won’t be too far off what regular Statsbomb readers would do with the same level of data.

With centre-backs you need to know what you’re doing, and even then you might not have a fucking clue. And that’s personal experience.

You can get some nice bits of trivia. There are around 1200 McDonald’s in the United Kingdom and only 5 in Bosnia, but you’d be a fool to use that to imply that Bosnia is 240 times more healthy than the UK. The point is that when you present a statistic you’ve got to know what it actually means.

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s