The numbers I use for MLE's will give you results that work for leagues. For example, the average hitter in the California league will hit more homers than the average hitter in the Florida State league. Yet, the MLE homerun total for these average players is the same. Does this work for all players, though? Could it be that a power hitter, like Brandon Wood, will take greater advantage of the CAL league than the average hitter? Do we need to let more air out of their MLE's?
I looked at, for 2002 to 2004, players who hit 15 or more homers in a CAL league season, and then looked at the same number of top power hitters from the FSL (they had to hit 12 to qualify). This is my smaple of hitters with some power.
My MLE formula predicts that the CAL guys will hit 79% as many homeruns when they go to AA, and the FSL guys 94%. The results? The CAL guys hit 77%. Pretty close.
The FSL guys? 107%. Turns out they hit more in AA, showing how hard it is to hit homeruns in the Florida State league. This is something I'll have to adjust next time I run MLE's, as the players will generally produce a lessor batting line (it is a level tougher), yet still hit more homers. One reason to beware of deriving my component league factors from the overall run scoring factor.