Split stats analysis.... help

[hot4tx]

Looking at split stats on Shea Hillenbrand and I found something very, very odd. From 2003-2006 Hillenbrand's LHP/RHP batting average shows a huge split... .320 vs LHP and .278 vs RHP. He batted:

LHP RHP
2006 .338 .256
2005 .325 .279
2004 .323 .304
2003 .298 .271

Being a righty, he hits lefties a lot better. However, look back further and the stat reverses itself:
LHP RHP
2002 .269 .299
2001 .227 .277

And that's with 1102 ABs in 2001 & 2002. So what does this mean? Did he use to bat left-handed or switch hit or something? Did he just learn how to hit LHP and forget how to hit RHP? I'm trying to come up with some reason behind the change in this trend. Or is this just statistical aberation?

Also as far as using this to predict the future is it safe to project .320 vs LHP in 2007 as he averaged over the last 4 years?



Thanks ahead of time for your help.... now go ahead and laugh at me for looking at Hillenbrand so indepth

[tomdoyle3]

uh you shouldnt take his stats from 4-5 years ago into account, the ones you want to be looking at are the last 2-3. stats from back then shouldnt really have much of an effect on your projections for this year..

[Koby Schellenger]

A lot of the variation comes from the small sample size. He probably only recorded around 150 at-bats against lefties while he had around 450 at-bats against righties in each season so random variation is going to play a pretty big role in the numbers.