Tuesday, July 21, 2009

Goalkeeper strategy

With less than a month to go until the big kick off, its time to get serious with team selections and I will start the first in a series of position focused posts at the logical place - between the sticks. In previous seasons I have often gone with the strategy of getting a consistent 'keeper on day one and sticking with him throughout the season, to give myself one less headache on selection day. Last year I rode van der Sar all year, which was great for that epic 9 game clean sheet run, but was it really the best strategy?

The answer is almost certainly no, especially when you consider that I, along with around a quarter of the league, held Schwarzer from day one - who only went on to lead all keepers in scoring despite only costing 4m. Most would think that this was an aberration, that Schwarzer was a one off steal and no other can give that kind of value. However, the top four keepers in 2008-09 in terms of points per million spent (PPMS) all cost less than 5m and Sorensen at 4m had an incredible 1.021, second in the league among all players to Schwarzer. Yes the big names (Cech, Reina and van der Sar) outscored these bargains but at what price? Van der Sar had just 12 points more than Sorensen and 15 more than Jaaskelainen but cost 2m and 1.5m more respectively.

The new strategy I am proposing is therefore to utilise both players, picking a pair of keepers whose fixtures complement each others. Obviously, in order to be cost effective both players need to be cheap, so I have set the limit at 4.5m. This means we will spend between 8m and 9m on our keepers, at least 1m less than anyone who selects one of the big four.

Using last years goals for and against (adjusted to account for the promoted teams and for teams additions and disposals) we can generate an approximate goals per game value for each team, eg. Chelsea average 1.79 goals per game on the road so even a home fixture looks like a bad play, where as Fulham scored just 0.68 and therefore appear to be a good play. We can then take two keepers and make an estimation as to which one we would play each week based on the two matchups (ie the one with the goals per game value!). Whichever pair gives the lowest average score will therefore be the best pair of keepers.

The four keepers I have chosen for consideration are Sorensen, Jaaskelainen, Kirkland and Robinson. This is based on value, historical performance and, in Robinson's case, a hunch that Blackburn will be tighter under Big Sam this year. Without going too deep in the analysis, the pair with the most complimentary schedules is Sorensen and Kirkland (and by some distance). In fact, the combined schedule is an absolute dream, with zero 'red' games (defined as an opponent averaging over 1.5 goals per game) and no less than thirteen 'green' games (defined as an opponent who scores less than a goal a game, over the first half of the season.

Using the combo approach allows for 6 games against the promoted sides (4 at home in the first 19), 15 home games (including Sunderland and Fulham twice) and only one game against a side that finished in the top 6 (we are forced to play Stoke at home to United in week 7) last season. I can see this strategy generating 150+ points from the 'keeper position, which should be close to anything the big four keepers can generate on their own, all the while saving at least 4.5m over those who select a big name GK.

What are your plans for your keeper this season? Do you prefer to lock up a top four player and watch the points roll in or are you tempted to roll the dice each week in search of cheaper points? As always, I look forward to hearing your thoughts either here, or on twitter @plfantasy.


marco said...

Good to see you back. For me this is not just a question about goalkeepers but all of defense. When you look at the top 3 defenders and goalkeepers, sometimes the goalkeeper is lower price and certain to start. Let's take Chelsea. To get Chelsea clean sheets, at this stage I rather pick Cech at 6.5 vs Bosingwa, Cole, Zhirkov, Terry, Alex who are more expensive and/or not certain to start. When you put this against let's say Shawcross vs Sorensen from Stoke, that's the real question to ask rather than GK only

marco said...

So for me the question is do I pick Cech / Shawcross or Sorensen / Alex, etc.

Chris Glover said...

yeh good point, that is definately an attraction of the big names, as like you say, Man U clean sheets for 6m is a good price.
My only reason for liking the budget guys is that sometimes they can score decent points from saves and bonuses even when the team concedes. Defenders however tend to get 2 points or a clean sheet (unless you get the odd goal).

NCNB made a great point about not wanting to pick the wrong guy each week and I think that is why we shy away from the combo strategy - fear of picking the wrong guy and looking stupid. If we accept we will sometimes make the wrong, but well informed decision, I think the strategy can work.

TD Ocho said...

I'm struggling with this myself. At the moment I have Hart and Sorensen, but I've considered picking up Cech in place of one. It's interesting that you found the Kirkland/Sorensen combo best since at first glance Kirkland's fixtures are horrible, but I guess it's all about how they fall in relation to Sorensen.

Chris Glover said...

Based on the rough numbers I would only play Kirkland three times in the first nine games, but over the first half of the season they would actually split 10-9.

I also like Hart but i didn't include him in my stats as I know close to nothing about Birmingham or the Championship last year. They had a good defense by the stats but whether that will translate is a tough one to call.

Thommas said...

I really can't make up my mind about the keeper issue! Currently I'm going with Schwarzer and Sorensen, but I'm beginning to like the Sorensen/Kirkland combo. But I'm not sure yet tho.

nehrmt01 said...

This season is going to be out of control for the first five weeks whilst everyone is clamoring from this insane transfer season. I think the most volitile choices for keeper lay in teams like chelsea and MU where so much learning one another's style on the pitch is about to take place. The only! choice for your starting keeper has to be schwarzer. Not only is he on a stable team set to play steady-unimpressive, yet reliable football, once again but his value frees up points to spread out on the rest of the field where a safe but more expensice player might be just whats necessary to score solid points in the first few weeks.

Anonymous said...

I recently came across your blog and have been reading along. I thought I would leave my first comment. I don't know what to say except that I have enjoyed reading. Nice blog. I will keep visiting this blog very often.



NotoriousDre said...

i luv ya blog..keep it up!