Sensible Stats round-up 30/04/16

This column aims to look at use of football statistics in the more mainstream areas of the online world, good or bad, and explain why they’re good or how they could be improved.

Sky Sports News HQ, Man City vs Real Madrid preview stats

This is an interesting mix of basic (raw goals) and slightly more advanced (shot conversion) stats. Using more advanced stats, looking for insight, over just a 10 game sample in a cup competition runs risks with sample size and quality so really I don’t mind that they’re using the basic stats. Their job here is to be able to tell a story rather than to make any significant claims with the numbers.

I think it would make a lot more sense if goals conceded and clean sheets were directly below goals scored, just because that would be easier to follow. Possession seems fine too, they all indicate that Madrid have been dominant so far in the competition and City more mediocre.

Average number of passes per match is a bit odd. Without any context it’s hard to know what it means, and with context it’s essentially the same as possession. You could replace it with something like average number of shots taken per match, which would also make the use of shot conversion a bit more relevant. For this stat, with no context, people are likely to be left just admiring the pretty combination of curves and lines., Liverpool should go all out to sign Mustafi

This is mainly a comment about defensive stats.

Mustafi stats

Strikers are judged by goals or the goals they are likely to score. There’s only one way to score goals, in the end, by shooting, and that’s why shot numbers and metrics work for strikers.

For defenders, their job is either to stop goals or stop chances, and there are various strategies to do this. You can go forward and try to tackle anyone in sight, you can sit back behind a deep midfield and head away a couple of crosses.

At the moment, the most one can probably do with the defensive stats is determine what a centre-back or a team’s defensive style is, rather than their actual quality (this may change in the future, but that is the case for now). Listing figures of total stats is at best a meh and unclear way to assess a centre-back, and at worst a bad and misleading one.

You could look at Mustafi’s stats, try and determine what style of player that makes him, and try and relate that into Liverpool’s system, but that is still open to fairly large risks of misinterpretation if you don’t know the stats or players well. Defensive errors themselves have a whole section devoted to them below.

Squawka, Most defensive errors

For a team to only make 29 defensive errors in 34 games is actually quite impressive. That’s less than one mistake, by the whole team, per game. Clearly, Opta’s defensive errors are things that are notable. This article found that numbers of defensive errors aren’t repeatable from year to year, and even that teams who make more defensive errors don’t necessarily concede more shots or goals, which makes the stat seem particularly suspect. There isn’t a definition for the stat on Opta’s blog post of their event definitions either.

This isn’t necessarily the end of the world. Stats are often used on Twitter as harmless trivia, which is fine. However, probably because of the lack of clear meaning around defensive statistics, the defensive errors stat tends to be presented in a more factual manner. “Liverpool’s centre-backs are bad; look, they make loads of defensive errors” type of thing. And this isn’t fine.

WhoScored, Why Roma are desperate to keep Nainggolan

This followed what is suspected to be WhoScored and Squawka’s (WhoSquawked, if you will) formula of getting someone to write an article and then inserting some stats and graphics afterwards. The result can be mixed.

This is a nice article, I thought, on Nainggolan. Even the stats graphic inserted was alright.

Nainggolan WS managers graphic.png

While the WhoScored rating at the bottom makes this look a little like the stats above it are measuring quality, really the graphic is most useful at comparing how Nainggolan has been used under each manager. Under Garcia his defensive stats are higher and he’s involved in fewer goals; under Spalletti it seems that he’s clearly been used further up the pitch.

Unfortunately the article did not mention this explicitly in relation to the graphic (though it did mention Nainggolan’s change of positioning under Spalletti), and a more general audience might just have scrolled past the numbers.

Then comes the patented “and here’s the stats” paragraph. This one is less blatant than most though.

Nainggolan WS stats paragraph

Whether edited or straight from Horncastle, the writer, the main article has been integrated with the stats, and the two actually relate which is more than can be said for some similar typed articles.

The stats themselves are lacking context though. I’ve no idea if recovering the ball 233 times this season is high up in the league or indicative of a quality player. It feels like a semi-relevant info-dump. Which is a shame, because I otherwise liked the article.


If you see any mainstream stats use, good or bad, over the coming week let me know at @SensibleStats or @ETNAR_uk on Twitter, or, and it might turn up in a future article.


2 thoughts on “Sensible Stats round-up 30/04/16

  1. Pingback: Sensible Stats round-up 07/05/16 | Every Team Needs A Ron

Leave a Reply

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

You are commenting using your 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