And for your next installment of "What's on Duke's Mind?": Generally, the best course of action with BABIP is to assume everyone will regress back to the mean. It's usually right, because nearly everyone will. There will be guys on the fringes, your Darwin Barneys and Paul Goldschmidts, that won't. But, over enough of a sample, most guys will. It's just that it takes so long for this to stabilize that we won't really know that Paul Goldschmidt is Paul Goldschmidt for awhile. And by that time, he might have changed. What if he starts hitting more fly balls, or pulling the ball more, or loses some speed? Then the BABIP god we knew might not be a BABIP god anymore. But, how do we know for sure? I don't think this version of xBABIP alone will be able to tell us. But maybe it stabilizes faster. And maybe we could combine regular BABIP and xBABIP and something like raw exit velocity and create some sort of panacea to help us understand BABIP freaks in a more timely manner. We can also throw in speed, pull data, shifting data, etc. We can at least get a better understanding.