And a conclusion…
November 11, 2005 4 Comments
Here are the correlations between regular old F% and Zone Rating:
2B = .81
3B = .86
SS = .81
So this shocks me more, because I thought that putouts would provide a lot of noise. Let’s see what happens if we remove all players with under 725 innings played (about half a season).
mF%
2B = .63
3B = .21
SS = .66
F%
2B = .62
3B = .19
SS = .63
Wow! I knew that restricting the list to only high-innings players would reduce the correlations but I did not expect that it would impact F% just the same as mF%. I would have expected mF% to hold up much better. Either way, it seems that this would only work at the middle infield positions, probably because errors there are more indicative of range that at third base, where they are generally an indication of a strong throwing arm that is perhaps a bit erratic. Either way, it doesn’t seem to offer much more information the regular old F%.
you might want to consider using something other than pearson’s r to compare the two stats. perhaps try a bland-altman plot to get a better sense of what is going on.
see this:
http://www.tufts.edu/~gdallal/compare.htm
I’m not a professional statistician so my comments come with a grain (or shaker) of salt. But I believe that the use of correlation here is fine. The bias of a method is generally unimportant if it correlates well, at least in this case. What we want to know is: does F% or mF% do a good job of measuring defense? Not, does it accurately tell us how many runs a player saved, but simply, does it generally distinguish between good and bad defenders. It’s not on the same scale as ZR, and yes, if we were to convert mF% of F% into runs, the impact would be muted. But the question is does it identify who is good and who is bad. In this case, the Pearson r, I think, is okay to use.
Nevertheless, that paper is an interesting read.
You see you got mentioned on ESPN? Congrats!
Thanks, Evan.