Originally posted by merk
Originally posted by Ahoda
Originally posted by po
Why do you love Josh so much and hate me?
He is fucking 22-9 in one run games and +3 pyth and I am -4 pyth.
Wang was a great pick up eh?
Fuck me. Just gotta make the playoffs though.
you do realize that no one but me has any idea what you just posted?
I understand... pyth = runs scored / ( (runs scored)^2 + (runs allowed)^2) )^-1/2...it is supposed to predict the fraction of games you win....so if you multiply pyth by the number of games played, you can see if your team has been lucky or not...At some point a couple years ago, I wrote my own baseball sim (much of the effort going into a smart cpu) that kept track pyth and a number of other relevant stats...
Also, winning or losing in 1 run games is supposed to be a matter of luck...
Nice. Very nice.
I have never seen Pyth explained like that but very nice. Its been calculated thought that using 1.83 instead of 2 for the exponent gives a little more accurate of a result. So it looks like this:
(Runs Scored)^1.83
---------------------------------------------------------
(Runs Scored)^1.83 + (Runs Allowed)^1.83
But what Ahoda was refering to is we were talking about our OOTP league in which he runs, and is currently trying to run my team out of the playoffs.
I do love me some sabermetricians though.
Originally posted by Ahoda
Originally posted by po
Why do you love Josh so much and hate me?
He is fucking 22-9 in one run games and +3 pyth and I am -4 pyth.
Wang was a great pick up eh?
Fuck me. Just gotta make the playoffs though.
you do realize that no one but me has any idea what you just posted?
I understand... pyth = runs scored / ( (runs scored)^2 + (runs allowed)^2) )^-1/2...it is supposed to predict the fraction of games you win....so if you multiply pyth by the number of games played, you can see if your team has been lucky or not...At some point a couple years ago, I wrote my own baseball sim (much of the effort going into a smart cpu) that kept track pyth and a number of other relevant stats...
Also, winning or losing in 1 run games is supposed to be a matter of luck...
Nice. Very nice.
I have never seen Pyth explained like that but very nice. Its been calculated thought that using 1.83 instead of 2 for the exponent gives a little more accurate of a result. So it looks like this:
(Runs Scored)^1.83
---------------------------------------------------------
(Runs Scored)^1.83 + (Runs Allowed)^1.83
But what Ahoda was refering to is we were talking about our OOTP league in which he runs, and is currently trying to run my team out of the playoffs.
I do love me some sabermetricians though.
Last edited Sep 16, 2008 12:16:51