Just following up on a question I had during the last gamethread. It's pretty common knowledge that bunting decreases the run expectancy of the inning, but in situations such as last night, the amount of runs you score is less important than the chance of scoring at least 1 run (to tie the game).
Plugging various numbers into http://www.tangotiger.net/markov_wes.html gives the following run frequencies (chance of scoring):
2010 Redsox: .465 for runner on first 0 out vs. .434 for runner on second 1 out.
2009 AL: 4.34 -> 4.18
And just for curiosity, the combined lines of Scutaro-Pedroia-Ortiz (the batters who were up): .456->.423.
So it seems that bunting is indeed almost always a bad decision, even in these situations. Good call Ben.
EDIT: Typo, accidentally wrote 2009 Sox when I used 2010 stats.