Sunday, August 5, 2012

Do defences really strengthen as the season progresses?

As we continue to look back at last season, we turn our attention to the defensive side of the game. I wouldn’t go so far as to say it’s held as a self-evident fact, but it is widely stated that teams tend to ‘tighten up’ at the back over the winter period, with goals being harder to come by and, by definition, clean sheets more prevalent. Such information will obviously be useful in ascertaining when to spend big on your defense and when to use your funds elsewhere, so let’s take a look at the numbers, starting with the 2011-12 season:








The high home totals stand out in chart one (as in line with the hypothesis) but the graphs aren’t the easiest to quickly deduce information from, so we can add some lines of best fit to get a more telling result:



The data tells a totally different story to the standard narrative, with goals per game conceded increasing during the early-middle portion of the season, and clean sheets kept down accordingly. We can also observe a slight uptick in goals in the final weeks of the year, though some caution should be exercised here as much of that may be due to players being rested or injured.

How about if we expand our sample to include the past two seasons?



Not much changes with the increased sample, and while I'm not willing to suggest this is by any means conclusive (I would expand the sample for more years if I had more time), I think the data is at least persuasive in a negative sense: we shouldn't be expecting a glut of clean sheets in the middle of the season as team's 'tighten up. Before taking conclusions too far though, what if we look at which teams are keeping clean sheets and when.

The below graph shows the percentage of clean sheets earned each week by the six best defensive teams for that season (2010-11 : Ars, Che, Ful, Liv MnC, MnU; 2011-12: Che, Eve, Liv, MnC, MnU, Tot):


Over the past two seasons we can see that the strongest defensive teams have started out well without being spectacular (~35% of clean sheets kept) but as the season progresses they have gotten stronger and stronger, peaking around the mid season point. This ties into the narrative of good teams coming into their own, and perhaps suggests that it’s around that period we should be aiming shift our defensive funds towards the stronger teams.

Now, the problem of course is that these two findings appear to give somewhat contradictory advice. On one hand, we are saying that clean sheets decrease during the middle of the season, but on the other we’re saying the good teams keep more clean sheets, so what should we do? Well, it seems like an optimal strategy may be to open the season with 5 defenders from teams with upside, whose fixtures gel and who all come at an easily replaceable price (getting stuck with too many dud 4.5m defenders can be a major problem). As the cream rises during the season, we could consider shifting funds from having 5 good players to maybe just 3 very good ones (hopefully one will emerge at a bargain price) and then ride those studs to virtually the end of the season (which becomes a crap shoot depending on who has trophies to play for, and which ones).

For the immediate future then, the lesson is that big name defenders aren’t perhaps essential to open the season and you might be able to free up a couple of million here and there to use elsewhere if you so desire. If you decide to follow that strategy then flexibility will be key as you will almost certainly want to divert funds towards the better teams during the middle portion of the season.

As always, these are just musings and aren't really structured in a way to be overly convincing either way. We could probably find some data to suggest that the 'big' teams provide good value too, though I do find the above fairly persuasive. I am avoiding getting into names for this season for now, with those posts to come when I can get stuck into this year's price list when I get back from Ecuador later in the week. As always thanks for reading, and I will answer comments soon, so post away below or @plfantasy.

6 comments:

Mitchell Stirling said...

Interesting stuff, is making me rethink my current line-up of

Green / Krul
Rafael / Lescott / Kaboul / Cole / Wilson
Nolan / Silva / Marin / Sigurdsson / Osman
Rooney / Lambert / Progrebnyak

x00x said...

another good post chris - one thing that gives me pause over staying away from top tier more expensive defenders from the start is that the clean sheets are only one aspect of their points. A player like Kompany provides bps and attacking points as well, so to really dig into if/when one should have someone of his ilk we need to consider value rather than just clean sheets (perhaps isolating defenders % of points from clean sheets would help define a list?).

I get the 'dont pay for defender attacking points' narrative to some degree, but ultimately we need to maximize total points by combining best value and top point-getters at the appropriate ratio (whatever that may be) throughout our lineups as a whole.

At any rate, have a good trip home :)

Unknown said...

This is a great post. I think I have consistently overspent on defenders based on how important they were at the end of the season, but I think taking a more long term view might serve me batter, knowing that I can shift over to the value towards then end, and free up cash in the begining.

Nishant Agarwal said...

Great article. was having the same discussion with my dad who is an avid player too. It seems we're both right. Thanks. Really nice articles.

iluvtehlulz said...

2 seasons is still a pretty small sample size. It would be interesting to revisit this using 10 years worth, as you did in the analysis of the promoted teams, or even 20 years if you have access to all the data.

city said...

thanks for sharing.