On Chelsea in November

Gabriele Marcotti’s article today (h/t to Colin Trainor for bringing the article to my attention) draws attention to Chelsea’s poor points return during the months of November in the past few Premiership seasons. To quote:

“Seasoned Chelsea-watchers will tell you that it’s a November thing, that in recent years the Blues have always faltered this time of year. Look at their point totals:

2010-11: 10 in 7 games
2011-12: 7 in 6
2012-13: 7 in 7
2013-14: 7 in 5

See? It’s over now. Nothing lasts forever, even cold November rain.

Well, I don’t believe in voodoo and I don’t buy these kinds of stats.

While it’s true that five different managers — Carlo Ancelotti, Andre Villas-Boas, Roberto Di Matteo, Rafa Benitez and Mourinho — took charge of Chelsea in those November games over the past four years, and they all struggled to varying degrees, there is no curse linked to the 11th month of the year. ”

We know weird things can happen in a calendar month – last year I showed that straight up weird things happen to United sh% and sv% in February, and two years ago I showed that Barcelona’s lack of points with Xavi off the pitch was because the teams sv% collapsed, something unrepeatable that he could have next to zero control over.

What I’m basically saying is that, as much as we want to read into these things, a bloated/collapsing sh% and sv% over a small sample size is a common cause. Thus, I suspect Marcotti is spot on when he says there’s no curse. And I’d go further and suggest that this has almost no bearing on how Chelsea will score points next November.

It would, however be interesting to see if we could dig out the root of Chelsea’s problem over the span of seasons Marcotti mentions (from the beginning of the ’10-11 season – I’m using only Premiership games). The table below gives a quick summary (and it’s not hard to spot the cause).

*For an explanation as to why % of shots for is important see here, and for an explanation of PDO and its importance see here and here*

Screen Shot 2013-11-27 at 12.30.37 PM

So in November Chelsea take essentially the same proportion of shots, but (as suspected) their sv% dips slightly and their sh% drops off a cliff. We know that sh% and sv% are very much luck driven processes, whereas the proportion of shots taken is much more skill driven – and so, in short, we’d expect that, come next November, these numbers will regress towards the mean, and that Chelsea will collect the number of points that their play deserves.

I guess that there is an alternative explanation – that Chelsea players forget how to shoot in November, regardless of who their manager is. There remains, however, a lot of ground ahead of anyone who believes that to be the case.


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 )

Google+ photo

You are commenting using your Google+ 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 )


Connecting to %s