This is a reprint from my blog. I thought that it applied here as well, so, here we are. I apologize for the poor linkage, I don't know how to do them correctly on here.
Discalimer: DFP does not endorse bunting in almost any situation. The following case is a very specific situation, and should not be applied outside of this case.
In the A's come from behind victory yesterday, there was a play that made Steve from Fire Jim Tracy freak out. (See comment 196.) With the score tied 7-7 and runners on first and second, Kem Macha had Scott Hatteberg lay down a bunt.
Now, I'm no fan of the bunt, but I could at least see the logic here. Bunting in this situation does increase your likely hood of scoring the winning run, and you stay out of the double play. However, other factors can contribute to this, including the fact that getting a hit will win you the game, the odds of the bunt failing, and the chances of working a walk. The question that I wish to answer is is Scott Hatteberg such a double play risk that taking the bat out of his hands is the correct play.
Staying out of the double play comes down to three factors: likely hood to get a hit, likely hood to not make an out, and the odds of hitting into the double play (found here).
Using Nichol's Expected Runs Table (http://www.nssl.noaa.gov/users/brooks/public_html/feda/datasets/expectedruns.html) to find the percentage chance of scoring in certain situations, it is not a good idea to bunt if the following equation holds true:
([Batting Average] * 1) + ([Isolated Patience] * .882) + ([Double Play Rate] * .275) + ([1-Batting Average+IsoP+Double Play Rate] * .425) > .686 * .593 + .257* .425 + .882 * 15.0
([Batting Average] * 1) + ([Isolated Patience] * .882) + ([Double Play Rate] * .275) + ([1-Batting Average+IsoP+Double Play Rate] * .425) > .648323
The equation is finding the expected percentage of winning based on various game situations. In this case, assumptions are made, some of which do hurt the accuracy of the equation. It is assumed that getting a hit will score the runner, and hitting into a double play will erase the runners on first and second while advancing the runner to third.
Since a stat like "sacrifice attempts" does not seem to be readily available (despite my best efforts, the only definitive source seems to be the Bill James Handbook, and I do not own it), we'll go with some numbers derived from this article on Baseball Prospectus (http://www.baseballprospectus.com/article.php?articleid=2869). I have no idea where they came from, but they look good.
Situation Success Failure Overachievement
Runners on first and second 59.3 25.7 15.0
In this case, a success is defined as moving the runners to second and third. A failure is getting the lead runner thrown out, and an overachievement is loading the bases.
So how does this apply to Hatteberg? Hattebergs 2005 stats in the relevant categories are as follows:
Isolated Patience: .071
Double Play Rate: .199
Inserting it into the equation we get:
.273 + .071 *.882 + .199 *.275 + .455 * .425
.273+ .062622 + .054725 + .193375
In this case not having Hatteberg bunt costs about a six percent chance in winning the game. Really it does make sense. The sabermetric community routinely calls out managers for taking ridiculous chances by stealing bases. In this case, Hatteberg's double play percentage is so massive that simply having him swing the bat in this situation is taking an unnecessary risk.
Even the strategy of not having him take the bat off his shoulder and hoping he draws a walk is worse than this one, since league averages state the other team is twice as likely to screw something up and not get Hatteberg out as they are to walking him. (Hatteberg is probably somewhere under this 15% mark, due to his lack of speed, but it's close enough.) The moral of the story: bunting, in this situation, is a good thing. It plays to all of the things we preach, it avoids the unnecessary risk, and keeps you from wasting an out. While for the most part bunting is completely pointless and detrimental to your team's success, it does have it's time and place.