League Equivalency (July 2, 2003)
Clay and Nate employ the standard method of comparing "adjacent" entities. You can use this for a whole bunch of things, like trying to figure out timeline adjustments, or minors to majors equivalencies, or Japan / MLB, age adjustments, etc, etc. It's all based on the principle that taking the same guy in two different environments has the same "input" (presumably the player hasn't changed), and a different "output". The difference in outputs is the difference in environments.
The one place you have to be careful (generally in these equivalency things) is "regression towards the mean", since the "input" being measured is itself a sample and not the true rate. This is probably the most important thing to control for.
--posted by TangoTiger at 10:21 PM EDT
Posted 12:24 a.m.,
July 3, 2003
(#1) -
Dave Studenmund
(homepage)
Tango, I understand that regression to the mean is a problem with studies like this. MGL mentioned this regarding the park effects study (that I'm still working on). My question is, what's the math here? How should someone like me (a statistics 101 type) try to account for regression to the mean in my analyses?
Posted 3:10 a.m.,
July 3, 2003
(#2) -
Michael
I think a bigger problem is the non-random differences between the players who change leagues and those that do not. Young good players do not tend to change leagues b/c they are not yet free agents. That means the group of players changing leagues is unlikely to be similar to the group of players not changing leagues.
Posted 11:50 a.m.,
July 3, 2003
(#3) -
bob mong
That means the group of players changing leagues is unlikely to be similar to the group of players not changing leagues.
Might it be better to do a matched-pair study? Match each league-switching player with a non-league-switching player of the same age - you might want to try to match for playing time and quality as well, I don't know.