Quote:
Originally Posted by LansdowneSt
You have the option to double-weight the 100 AB in year X to hit 200 AB.
|
That would do nothing in terms of distinguishing whether a the statistics put up in those at bats were a small sample from a perennial MLB All-Star or the greatest four weeks in the career of an organizational player. An example would be something like Garciaparra in 2001. In the midst of a run of All-Star seasons he missed most of that year with injuries, while performing at a level below that of the surrounding years but still pretty decent for a shortstop (.289 / .352 / .470). He is treated no differently than Roosevelt Brown (491 career plate appearances) or Mark Little (282 career PAs), who were similarly productive hitters as Garciaparra in virtually identical samples (NG 91, RB 92, ML 90). That is just the fact of how the one-year recalc works. Given that system, the only realistic way to avoid many of the issues cited is to begin with a curated database that manually has incorporated these distinctions.