Quick Question Thread

[sbrylski]

Quick Question: Where does the 11 come from in the bolded part below?

 

Here is the log in question:

http://www.fangraphs.com/plays.aspx?date=2007-05-08&team=Diamondbacks&dh=0

We see that for this play in question, the LI was 3.13, and the chances of winning (with an average hitter) was .501. The impact of the out was .089 wins.

(You could also have done this in your head, as an out has a win value of around .027 wins, which times 3.13 is .085).

Eaton, on 314 PA, actually has good hitting stats for a pitcher: .190/.264/.271 (wOBA of .240-.250). Regressing that somewhat toward the pitcher mean of .165, let's say he's a true .200 wOBA hitter. Remembing that there's a PH penalty, then the average PF would be a bit over wOBA of .300. That puts him at around -.08 runs per PA worse than average (-.04 runs if no regression on Eaton's hitting). The runs per win is 11/3.13=3.5, meaning the -.08 runs divided by 3.5 is -.02 wins in impact.

(I didn't check for handedness of pitcher and hitter.)

He'd have to be +.02 wins better than his relievers to justify the move, just for the 8th inning. That +.02 wins per inning is the best pitcher in the league.

A ridiculous move from the outset, and more ridiculous if we look at it here.

Is he just using 11 runs per win instead of 10 there, or is that number something different altogether?

 

[rluzinski]

Yes, 11 is the runs/win in an average situation. I think fangraphs calculates run/win for each season. You probably can dig up their methodology in one of their WAR articles.