View Single Post
11-14-2012, 03:17 AM   #48
dougdirt
The Boss

Join Date: Jan 2006
Posts: 33,503
Re: Why we need computers calling balls and strikes

Quote:
 Originally Posted by cincinnati chili I do agree with the comment that I don't see how Molina's contribution adds up to 50 runs prevented in a year.
This is from the comments at the original article, but it explains it rather easily:

Quote:
 Molina caught 709.2 IP in 2012. Converting one walk into a strikeout has a value of ~-0.6 runs. As a very rough approximation, if Jose Molina can convert one walk into a strikeout every 10 IP or so, he's right around 50 runs/year. You could do a more detailed analysis and figure out the linear weight for converting one ball into one strike on a frequency-averaged count, but the super-rough figure of "stealing" one strikeout per 10 IP doesn't seem intuitively out of line. As another example, if you plot all regular American League home-plate umpires by K/9, they ranged from 8.9 (Dan Iassogna) to 6.2 (Sam Holbrook) in 2012. K/BB ranged from 1.88-3.58, almost a factor of 2. There's already huge variation in umpire tendencies, and it's not unreasonable that a Crafty Molina could sway a malleable umpire by ~0.8 K/9, when umpires vary from each other by 3x that.
Now, his math is slightly off, as that is only 43 runs, but still.... that is 4 wins worth of value. That is like adding Brandon Phillips to your team by fooling the umpire.
__________________
www.redsminorleagues.com