Thursday, October 18, 2012

Gameweek 8 Preview

The below data is still based on the older goals-based model, as the underlying stats model is still being tweaked. I am hoping to get this up by the weekend, but either way it should be good to go for gameweek 9. Looking at the top ranked teams, I don't see too many reasons to massively downgrade anyone based on their underlying stats, though Swansea's defense at four is probably a touch high once you look at the volume of shots they've given up this season.

On the attacking side, Fulham aren't creating as many chances as we'd like so we could downgrade their ranking a touch, especially when we consider that Villa are merely a mid ranked defense according to their underlying data. You still like the matchup, but some caution should be exercised for anyone thinking of paying for a Fulham player this week.


3 comments:

Adam Turner said...
This comment has been removed by the author.
Adam Turner said...

Interesting to see that Chelsea have such a low attacking rating. With Chelsea scoring plenty of goals (I think they have scored in every game) and Tottenham (apart from the last game) conceding in every game I would be interested to understand the rationale. I hadn't considered benching Hazard but if we agree that both Hazard and Ben Arfa post a similar threat for their respective teams then maybe Ben Arfa should play ahead of Hazard?

John Doe, 2008 said...

@ Adam -

I would probably take a gander at this post: http://premierleaguefantasy.blogspot.com/2012/10/forecasting-player-performance-goals.html. That explains a lot of the rationale behind goal forecasting which can be applied to team attacking rankings as well.

I will say that Chelsea's 15 goals on 36 shots on target is well above league average and probably indicates they have been a bit luck offensively.