Every time this comes up I ask this, and I've never had anyone give an answer. Don't the run expectancies include the times when managers sacrifice bunt? Therefore, in order to prove the value/lack thereof with a "runner on 2nd, 1 out" and "runner on 1st, 0 out", wouldn't the second situation have to be modified to eliminate the sacrifice bunts, considering that as it is it includes all the sacrifice bunts that may or may not be useful? Your question is logical. But I would make two points. First of all, bunt attempts are only a fraction of these situations. Secondly, since post-attempt expectancies are the same, they must conform to the "clean" averages as well. I was reading and believing Bill James when most of this board was in diapers. I am no apologist for baseball's inherited "wisdom." But the bunt can be a weapon if used properly. The problem is that bunting is a rare skill. How often do you see a position player go down 0-2 by flailing away twice? If I were manager I would insist that all my guys work on their bunting.