Julio, is that the likelihood of scoring one run exactly, or one or more runs? "one or more" seems to be the appropriate percentage here, since the data is taken across all innings, and the innings where 2+ runs would score would also win the game. Or am I missing something?
Looking at the calculator, it shows one or more runs, runner on second, no outs as 62%, and runner on third, one out as 66%. Only a 4% increase, which would seem to make bunting not a great idea if there is any chance at all of messing up, which would drop the odds to 41%. So you are gambling 21% to gain 4%, so your bunting success rate would have to be better than 5/6 (83%) to be worth the gamble.
