Can the performance of the Celtic central defenders be assessed in terms of their consistency?
Thanks to @seanmcneal3 on Twitter for asking whether Boyata’s performances were consistent.
The key aggregated metric for assessing defensive actions is the Defensive Action Success Rating (DASR). I’ve explained it before and the Glossary is the big lad at the back who holds it all together in these situations.
Oh, alright then here is the definition:
Defensive Action Success Rate % = The proportion of Successful Challenges as a % of all challenges attempted (Successful Challenges plus Unsuccessful Challenges).
The DASR% average for all Celtic central defenders (at least 900 minutes) this season is:
Ajer has slipped from 81% (also known as “The Van Dijk Line”) in the last couple of matches since I wrote Head Bhoy. The fame obviously went to his Head.
Most Celtic central defenders of the last four years have been around the 75% mark. This doesn’t tell us about consistency.
So, let’s look at each players’ DASR% per game, and then calculate the r-squared measure of consistency for their ratings.
Ajer
To the eye that doesn’t look too consistent. Shaky start against Rosenborg. A little dip of late as well.
Bitton
Now that looks a more reliable collection of scores around his average. The nightmare in Astana stands out!
Boyata
Looks like very few performances on the average! Five matches see the score drop below 60%.
Simunovic
The meltdown at Hearts away stands out a bit. Two other performances approach 50% whilst one performance hits 100%. Again, not many performances are on the average.
Consistency
Using r-squared to test the degree of linearity of the game by game DASR% scores, reveals as our eyes told us that it is Bitton who displays the most consistent defensive performances. In essence the closer the value to 1, the more linearity in the data set.
The other three are all around the same value. Indeed, the current 1st choices, Ajer and Boyata, display the least consistency but there isn’t much in it.
Bitton’s error rate is Ambrose-esque in central defence which is not factored into this analysis. But if you add in his passing out from the back skills, you can see why Rodgers was defensive about his Israeli defender!
I’m grateful for all your analyses and your use of statistical tests to bring a scientific approach to this subject. These are interesting results and it is useful to look at the opposition when comparing results above and below the individual players’ averages.