Hi Brian; Excellent site. One request?! Is it possible to add your opass efficiency stat to your advanced box scores?Seeing how you have established that it's the most predictive? I believe you now are using adj. y/a ? Just curious why you are using it? thanks Bill
I agree with Anon. It'd be nice to have net passing yards per attempt in the advanced box score. Currently I go to ESPN to find those numbers, despite the fact that ANFLS persuaded me to value that statistic.
Brian; Could you clarify where you are at with 'fumbles' in your model?
1. I believe you established that fumbles are more predictive than fumbles lost so for your OFUM number is it: total offensive fumbles /total offensive plays?
2.Why have you taken out Defensive forced fumbles which you used to have in and established a high correlation ? thanks Dan
Regarding fumbles. I include offensive fumbles (either lost or not) in the prediction/ranking model. It is relatively heavily regressed to account for its high variability. My hunch is that it's probably totally random except for one component--QB strip sacks.
Defensive forced fumbles do not correlate within season from week to week, so it is excluded.
Hi Brian;
ReplyDeleteExcellent site. One request?!
Is it possible to add your opass efficiency stat
to your advanced box scores?Seeing how you have established that it's the most predictive? I believe you now are using adj. y/a ? Just curious why you are using it? thanks Bill
I agree with Anon. It'd be nice to have net passing yards per attempt in the advanced box score. Currently I go to ESPN to find those numbers, despite the fact that ANFLS persuaded me to value that statistic.
ReplyDeleteBrian;
ReplyDeleteCould you clarify where you are at with 'fumbles' in your model?
1. I believe you established that fumbles are more predictive than fumbles lost so for your OFUM number is it:
total offensive fumbles /total offensive plays?
2.Why have you taken out Defensive forced fumbles which you used to have in and established a high correlation ?
thanks Dan
Brian;
ReplyDeleteWhat is your position on defenses forced fumbles?
Isn't this somewhat a sign of a strong defense?And somewhat repeatable?
Bill
Suggestions taken. I'll put them on the list.
ReplyDeleteRegarding fumbles. I include offensive fumbles (either lost or not) in the prediction/ranking model. It is relatively heavily regressed to account for its high variability. My hunch is that it's probably totally random except for one component--QB strip sacks.
Defensive forced fumbles do not correlate within season from week to week, so it is excluded.
thanks fro our responses Brian!
ReplyDelete"My hunch is that it's probably totally random except for one component--QB strip sacks."
Or if your name is Michael Vick argh!
So much talent except HE CANT HANG ONTO THE BALL!!
hard on the nerves for my fav. team!
dan