by Mean Dean » Sat Sep 09, 2006 2:21 pm
I don't think you could, no; not only would the data have a lot of "noise" in it (strikeouts are putouts, teams with more errors have more chances, etc.), it gets really screwed up by the simple fact that no matter how good your defense is, it can't record more than 27 outs in a standard game, and no matter how bad it is, it will eventually record 27 outs in a standard game.
The PC game can tell you what percentage of their X-chart rolls your players and team are converting, but unfortunately that was not brought over here. (You can see the leaders at each position in the league leaders, though, so that shows you what a good percentage is -- to grossly generalize, a 1 is usually around 95%, a 2 is usually around 80-85%, under 65% is bad, 50% or lower is a butcher.)
If you had absolutely nothing else going on in your life, you could guesstimate your middle infielders' double play turned rate. Take team GDP turned (under "team fielding"), and divide it by an approximation of the number of runners your team's pitchers put on first base: hits allowed, plus walks allowed, minus HR allowed, minus opposing SB, minus opposing CS. It still wouldn't be quite right, though, mostly because I don't see a way to find out doubles and triples allowed (and since not every steal attempt is an attempt to steal 2nd, and there are other tiny factors like balks, PB, HBP and SH.) I would bet, though, that if you did that for every team in the league, the ones with the highest number would tend to have the best middle infields. I'm not gonna bother to try it :P