Wednesday, 5 September 2012

Hitting the Target

I have done a brief analysis of my FPL points scored so far versus the F.SCORE my players have been getting each gameweek.  The straight correlation is 0.976, and 0.999 if I subtract captain points from my FPL score.  This is good news.  Very good.  



This basically means, to my mind anyway, that the F.SCORE ranking so far is an accurate indication of FPL points.  An important thing this also highlights is the contribution of all your eleven players to your score.  I've had Bale in my team each week so far and he has made 7 key passes and had 4 shots on target, gets a high F.SCORE but scored no attacking points in FPL.  On the other hand Michu has had 4 SoT and bagged 4 goals, the jammy sod, but his F.SCORE is less than Bale's.  Michu's very good fortune balances out Bale's poor returns. 

Overall my selected team has had 31 shots on target in the 3 game weeks scoring 15 goals.   This ratio of 2 shots on target to 1 goal scored (2:1) is higher than the Premier League seasonal average of about 3:1.  I first thought this meant just like Michu I have been lucky but it may actually indicate that I am picking the right players in the right fixtures to get above average returns.  Needs monitoring. 

This is all reminding me a little of Moneyball and trying to build a team to achieve an on base percentage good enough to score XXX runs which will win YYYY games, which has actually been at the back of my mind whilst building my model.  Overall I am pretty pleased with the results and it bodes well for the points forecasting which I am just getting started with and will have ready before GW4.





No comments:

Post a Comment