And don't take my "lack of work" personally, you are looking for very specific information to prove myself. I would need to provide two identical players or at least virtually similar players except for one attribute, AND have them play in the same ballpark effects AND have them face similar opponents, cmonnn you know I can't do that.
It's my job to come up with the hypotheses and YOU disprove me with the post from the devs. And even then, "You have to search for your own explanations even when you're told to accept the facts." The information I would be looking for to prove my theory would be a team like Mike proposed that was entirely made up of guys with crappy skills but 99 makeup, and we would see how they did. We are assuming they would be terrible, even me, but maybe they'd over-perform compared to what we would expect them to be. Maybe they would win 60 games instead of 50. That would be statistically significant, wouldn't it?