RE: Fading a bad offense from the AL

I have the reason why it does not perform in the NL. The reason is the DH. When you consider an AL team that averages less than 4.5 runs a game, that would be the NL equivalent of a team that averages 4.2 runs a game. (The DH is worth 0.3 runs per game to an AL team). So when you query for a conference specific result at 4.2 runs per game instead of 4.5 runs per game, it verifies the DH value in the AL:
H and Average(runs@team and season) < 4.2 and p:margin <= -8 and …

What you see is at 4.2 runs per game it negates the DH factor in the AL, while it compensates at the same rate of the value of the DH in the AL, and thus the results for the NL achieve a statistically acceptable similarity at 4.2rpg.