I remember reading an article on working with limited data in baseball but I can't seem to find it anymore. If anyone can remember or help find it, it'd be much appreciated.
The gist of it was something like "we have some limited data on the performance of these two players in a minor league, which one do we predict will be a better player in the future."
It's maybe around two years old. I would have guessed it was on r-bloggers.com but now that I checked for everything related to baseball on r-bloggers, I'll say it's almost certainly not on r-bloggers.