Clean Sheet Correlation
Clean Sheet Correlation
If you read one of my articles earlier in the week, you will know that I provided a list of Premier League clean sheet data and in doing so, it would allow you to make data driven betting predictions over the course of the midweek action.
However, in a slight one-eighty, I’m not necessartily going to look at the data from a betting point of view but more from an analytical one and focus on whether there is a correlation between clean sheets and league position this season.
While the way to do that is by building a table that combines both the league positions and the amount of clean sheets each of the 20 Premier League outfits have managed to achieve this season, something that looks as such:
THE LAUNDRY TABLE
|Team||Position||Result||CLEAN SHEET||CS %||CS Rank|
As you can see there is one very obvious correlation and that is, not only have Liverpool romped away at the top of the table, they have also kept the most clean sheets and although the run of shut outs was not extended against Wolves, there is no doubting the influence of Allisson since his return.
Therefore at it’s most basic level we can say that there is a correlation between keeping the most clean sheets and league position. However, for this to be a perfect hypothesis, there would be a mirror image of clubs 1 to 20 having the most to fewest clean sheets.
Something that quite obviously doesn’t ring true, although in the case of Norwich they are also doing their part to support the theory, as the Canaries are bottom of the table and have kept the joint least amount of clean sheets.
Therefore we can say at the two polar extremes of the data sample, there is a correlation – but if that proves to be the rather obvious, then why is the rest of table so jumbled when it comes to clean sheets vs league positions.
PRAGMATISM MAKES PERFECT
Let’s take Watford and Burnley for example, the latter are built on a bedrock of defence and therefore it is probably no surprise that they rank so high when it comes to recording clean sheets, even if they are just 13th in the table.
Although looking at Watford being so high in the clean sheet rankings, especially when second bottom does come as a huge surprise and when you consider that one third of their games end without conceding, there is some form of anomaly.
Then again this one is perhaps a little easier to explain and with Nigel Pearson coming in as the new man in charge, the Hornets have managed to keep clean sheets in three of their last five Premier League outings.
Which makes you think maybe it is teams at the bottom of the table that will keep more clean sheets than their counterparts that are higher up. With the threat of relegation hanging over their heads and with more to lose, maybe the aim is to shut up shop?
That idea certainly has an elemet of truth to it, of the 10 teams in the above table, 6 of those are in the bottom half of the Premier League table. Which suggests that less caution is thrown to the wind the lower down you go.
So there’s two things we now know:
There is a direct correlation at the absolute extremes of the table
Teams in the bottom half just about keep more clean sheets than those at the top
So let’s try and explain this better in graph form:
The blue line represents the league table and how the ranks from 1 through 20 follow each other in order, as they would when viewing the table at the end of the Match Of The Day. While if there was to be an absolute match with clean sheets, the orange line would follow the same course.
However, as you can see the shape of the orange line is much more erratic and this is because the amount of clean sheets does not follow the same trajectory – going back to the bottom half of the table achieving more defensive perfection.
Of course, one should not forget that as long as you are scoring goals and plenty of them you can absorb the lack of clean sheets by outscoring your opponents and that is going to be the next step of this analysis as I then look at the correlation between league placing and goals.
Happy punting and thanks for reading. Dan
If this has grabbed your interest and you would like to discuss/feedback then please feel free to drop me a message at firstname.lastname@example.org. While I am always looking for new football/data projects to work on and if you feel that my skills would be of use, I can be contacted at the same address.