Centre-back statistics are famously difficult to make sense of. I’ve probably written that sentence too many times, but it continues to be true and people continue to use defensive stats in out of context ways.
Obviously, some defenders will do more defensive work than others by virtue of their team’s system, making tackle and interception figures difficult to deal with. Nicolas Otamendi makes 5.6 tackles+interceptions per 90 minutes, Ryan Shawcross makes 2.5. David Luiz makes 3.1. This is not the respective ordering and distances in quality of centre-backs going from good-to-bad.
Similarly though, and something which has maybe gotten less coverage, is that tackle success rates differ depending on a player’s style too. Some players take more risks than others, and while choosing your moment is definitely skill, it is also definitely a style, you’ve got to acknowledge that when these tackle percentages crop up on your timeline or TV coverage. (I don’t have concrete figures, but I’ve found before that the average centre-back in some leagues have better/worse tackle percentages than others, presumable for this very reason).
It also doesn’t necessarily make a lot of sense as a statistic. Opta definitions are sometimes difficult to penetrate, but from what can be gathered it seems that Unsuccessful Tackles are classed as ‘Dribbled Past’, and wouldn’t include fouls, which seems important. (This isn’t to be confused, confusingly, with Tackles Won and Lost, which are all successful tackles and depend on where the ball ends up after the successful tackle).
The other key defensive stat is interceptions, and unfortunately we don’t have a way of knowing how many interceptions a player misses, or how often their indecision on whether to intercept or not results in them being out of position (which also be classed as a missed, or failed, interception, depending on how much you want to argue about semantics).
But the problem still remains about how to interpret these things. Theoretically, using stats like tackles against others like blocks, which suggest that a player is more prone to backing off, one could create a sort of ‘front foot’ rating for centre-backs. We could then, in theory, plot this against a sort of overall Success Rating to get a bit better of an idea of how players fared.
In the manner of Blue Peter, here’s one we prepared earlier.
This has no claims to being perfect, but it’s a reasonable glimpse of things you can do with simple data that makes some theoretical logical sense.
The weird things which would require extensive game-watching to work out to what extent they were weird flukes first:
Curtis Davies has a good reputation but being that far away from the trend line is weird and suspicious. He has 3.8 interceptions per 90, a very high figure for a centre-back who are generally sheltered to some extent by their midfields, and that’s the main departure between him and the rest of Hull’s centre-backs at least.
I believe that he has made more than 4 interceptions in a game on 8 occasions this season, against: Leicester (H), Swansea (A), Manchester United (H), Burnley (A) [first four games of the season], Chelsea (H), Tottenham (A), West Brom (A), and Bournemouth (H) [Phelan’s last and Silva’s first games].
Without diving any deeper, it’s hard to see much of a pattern. There’s a spell of 8 games in the middle of the season where he only made around 1.8 interceptions per game. Perhaps the poor start boosted Davis’ stats, but Jake Livermore only averaged 2.5 interceptions per 90 in his starts at CB at the start of the season. Colour me baffled.
Ditto Bernardo and Ayala. Neither have played a huge amount of minutes at starting centre-back, but that they’re both so high above the line and both play for Middlesbrough is intriguing.
James Collins is less standout, but still on a bit of an island, and looking deeper at his stats it’s possible that he’s highlighting an issue with using ball recoveries as a ‘back foot’ stat. These require the team keeping possession for two passes following the recovery, and the rest of West Ham’s centre-backs have an interception:ball recovery ratio (not that the two are related, they’re just easy figures to compare) of 1:2, whereas Collins’ is nearer to 1:1. Maybe he does something weird when he recovers a ball, like clearing or passing badly a lot more than his teammates. Colour me confused and intrigued.
Alfie Mawson is the other side of the same coin. He makes an extraordinary amount of ball recoveries compared to the rest of his defensive actions, and with these don’t figure in the Success%.
Tottenham’s centre-backs are in a weird clump. That they’re in a clump tallies with the pretty structured nature of Spurs’ overall system, particularly their defence. They also make an extraordinary amount of ball recoveries, possibly relating either to Tottenham’s safe possession style or due to Spurs’ combination of a relatively deep defensive line (when they’re actually defending), and combative defensive midfielders, which might result in a lot more loose balls for the centre-backs to pick up than in other teams.
Like I said, it’s not perfect, but it’s an idea.
Eastbridge Soccer Betting Broker, bespoke sports betting for the discerning client.