Some rockin' links
January 26, 2009 1 Comment
From the “stuff we’re reading” file: Derek Carty takes a look at predicting BABIP for batters. He finds that a regression-based formula put together by Peter Bendix and Chris Dutton rates the best of them all. This is a good next step. There are multiple techniques out there for this (and for estimating a bunch of other things), so now we need to start looking at which is the best. If there’s one place for a follow up, it would be making sure that the errors aren’t systematically distributed (does the formula over/under-shoot in a consistent manner for certain types of players), but one thing at a time I suppose.