While I had an idle moment at work, I was thinking about Sparky's terrific play last night in the ninth inning.
For those of you who missed it, with no one on and one out in the ninth and the Jays down 5-4 Sparky bunted for a base hit, dropping a bunt on the third base side in front of Shane Halter, who was playing back at third. Sparky later scored the tying run and the Jays eventually prevailed. Clearly, bunting leaves you open to tremendous criticism later; if you fail, everyone will point out that an extra-base hit would have been really handy in that situation. The bunt took some guts on Johnson's part.
Now as Tom Cheek and Mike Wilner pointed out on the radio broadcast, Troy Percival doesn't field his position well, and Reed Johnson is a genius at deadening the ball. But I was left wondering whether it really was a smart play. On further analysis, I think it probably was.
If you look at the particular circumstances, Halter isn't a bad third baseman. He has 229 career games at third and his defensive numbers at the position are pretty decent. But he hasn't played well this year, especially with an .875 fielding percentage that leads one to believe that he has a good shot at throwing it away if you pressure him. Percival's well known as a lousy fielder, and the numbers bear that out... he makes less than half as many plays as the average pitcher based on his innings pitched. Furthermore, Halter had played in on the first pitch but had moved back, and Sparky had a 1-0 count which is a good bunter's count. A strike is likely, and with no strikes it's less harmful to foul one off.
Now for the real analysis... let's head to the Win Expectation Table. (Thanks to Tangotiger for making this available).
Down one in the bottom of the ninth with no one on, the Win Expectation is .108. In other words, the team at bat wins 10.8% of those games. If Sparky makes it to first, that WE nearly doubles, to .208. If he makes an out, it declines to .042. I'll ignore the effects of fouling off the pitch and facing a 1-1 count instead of a 1-0 count (this is somewhat significant, but very hard to measure). These presume the average hitter is at the plate... which Sparky very nearly is.
So if success increases your win expectation by .100 and failure decreases it by .066, then you should attempt the bunt single if your chances of success are 40% or more. At 40%, you are exactly breaking even with the strategy of swinging away. I'm betting that Johnson had a better than 40% chance of success on that bunt. Smart play, and one that came up roses for the Blue Jays.
https://www.battersbox.ca/article.php?story=20040527110937999