Maybe its Pitching from the Stretch?
In a post earlier this month, I wondered Why does batting average go down by number of outs in an inning? Since the average plate appearances must go up with each out, the likelihood of having runners on base must go up as well. If pitchers are more effective pitching from the stretch, which they do with runners on base, that would explain the decline in batting average. I'll have to poke around the web and see I can get ERA, WHIP and opponents batting stats for pitcher in the stretch vs. the full wind up.
If they are better, I wonder if we'll ever pitchers pitching from the stretch exclusively. I know Lincecum recently pitched from the stretch w/o runners on base, but that had more to do with him finding his mechanics in a game where he felt out of rhythm.