Anybody else notice that Freddy Galvis is an A- fielder in the NL and a B fielder in the AL? While I understand why everything is normalized by league, this makes me seriously wonder if defense should be normalized by the full available "world" of players in any given season. Or maybe even with a denominator that uses something like a 5-year moving average. The idea is supposed to be to normalize for differences between rules, equipment, and how official scorers tend to view plays. But right now, as far as fielding is concerned, the AL and NL are playing under the same rules, the umpires have been shared for 20 years now, and presumably the official scorers are no more consistent with others in their own league than the other league. When the numbers are off by that much - not 1 but 2 1/3 letter grades, meaning there is at minimum more than a full 1/3 grade of difference between the leagues - it raises a red flag for me. In this case it's almost certainly statistical variance from insufficiently large samples and inequitable distribution of talent resulting in the difference. Those are exactly the things we're trying to normalize out of play, but it looks like now we're normalizing them in.
EDIT: The difference is in range, not fielding. But this is generally irrelevant to the point.