Why bunt?

Spurred on by DK’s open question in his morning links post at the PBC Blog today, let’s talk a little bit about the value of a bunt in a close game. The reason I want to do this isn’t to dwell on last night’s loss any more than is necessary, but because the bunt is a pretty basic baseball play (A runner on third is better than one on second no matter how many outs, right? Wrong.) and I think sometimes people don’t completely understand why I come so unhinged over plays like Tabata’s bunt last night. Instead of just ignoring it, I think this is a good chance to explain win expectancy and why bunting in most situations is stupid (understand that I’m not breaking any new ground here; you can do an easy Google search to find plenty of sabermetricians who have blazed the path on this front — I’m just trying to put it into digestible form for non-saber-oriented Pirate fans).

Win expectancy is exactly what it sounds like; it’s the chance at any given point in the game that a team will win based on score and situation. The number is based on the historical percentage of teams that have won games in the same situation and is context neutral (Author’s note: I screwed up this detail in the initial post, but Matt W set me straight in the comments — this is why starting posts at 1 AM when I’m exceptionally angry isn’t the best of ideas. Thanks to Matt for catching my mistake.), so each team starts the game with a 50% chance to win. When Andrew McCutchen grounded out to start the game, the Pirates WE dropped to 47.8%, because that’s how often road teams win after the first batter fails to get on. When Jose Tabata made the second out, it dropped to 46.3%. When Neil Walker walked and Garrett Jones singled with two outs, their WE rose to 50.1%. FanGraphs charts the WE changes for every batter in every game, so you can look at last night’s graph to get a better idea of what I’m talking about here.

So let’s zip ahead to the tenth inning. Using Tom Tango’s win expectancy list, when the road team doubles to lead off the ninth inning of a tie game (or any extra inning), their WE leaps to 67.2%. That is, the road team wins 67.2% of the time after a leadoff double in the the ninth inning or later of a tie game without even considering who’s up next, who the base runner is, etc. Bunting that runner over to third does raise the road team’s WE, but by a measley 0.7% to 67.9%. That is, the road team has about the same chance of winning with a runner on third and one out as they do with a runner on second and no outs. Successfully bunting in that situation is basically the same as saying, “I don’t think the guy at the plate can win the game for me, but I do think the guy up next can do it.” Great strategy with the pitcher up, not such a great strategy with your #2 hitter and one of the hottest guys on the team up.

On the flip side, if you screw the bunt up the way Tabata did and get that runner thrown out at third, the WE drops to 50.6%. Had Tabata gone up to the plate and struck out looking, the WE would’ve been 56.6%. That is, the risk (Tabata fails entirely and McCutchen is thrown out at third) of failing the bunt is greater than the risk of letting him swing away.

Now, tonight’s bad bunt notwithstanding, Tabata’s not a bad bunter. In fact, according to Baseball-Reference, tonight’s bunt was the first sac bunt he failed to get down all season. So let’s be friendly here and say he had a 75% chance of getting his bunt down properly (I think it’s less than this given the situation, but like I said I’ll be friendly). Most of the time, it would be a sac bunt (move McCutchen to third), but a small fraction of the time it might be a perfect bunt that Tabata legs out for a hit. The WE for first and second with no outs is about 70.3% and the WE for first and third with no outs is 80.6%. Of the 25% of the time he fails to get the bunt down properly, what happened last night probably only happens a small percentage of the time as well; more often it would likely result in McCutchen being left at second with one out, which would affect the WE% the same way a strikeout would.

And what happens if Tabata swings away? Well, he’s hitting about .310, so there’s a 31% chance he’ll get a hit. Let’s say there’s an 85% chance that the hit would score the speedy McCutchen from second base, which makes about 25% of the total outcomes. A run-scoring single raises the Pirates WE to 86.7%, a double to 88.8%, a triple to 91.3%, and Tango’s chart doesn’t even consider two-run leads in the ninth inning because the chances of winning there are pretty damn good. So there’s a 25% chance he knocks the run in, a 6% chance he reaches on an infield hit or outfield hit that doesn’t score ‘Cutch, and a 6% chance he walks. That equals his 37% OBP. He strikes out 15% of the time, so those PAs are a total wash. That means we’ve got 42% of the time that he’ll fly out or ground out. McCutchen will advance on a big chunk of those plays, too, though I have no idea how many.

If ‘Cutch goes to third on half of those in-play outs, that’s gives Tabata a 58% chance of either reaching safely or moving McCutchen up. If it’s more than half (and it makes sense that it would be since McCutchen is fast and Tabata hasn’t hit one infield fly ball this year), it starts to approach the 75% bunt success rate pretty quickly. I don’t have data for this, but it also seems to me that the chances that McCutchen is thrown out at third are significantly lower than the they are if Tabata is bunting. On a bunt, his job is to move from second to third and he’ll try to do so. If the hitter is swinging, he’ll generally be more careful.

What it amounts to is this: a bunt will never win you the game, while Jose Tabata is capable of driving the runner in by himself. If the bunt is successful in the situation the Pirates bunted in last night, it usually won’t move the needle on your chances to win the game significantly and the chance of complete meltdown (what happened last night or god-forbid a double play, which would’ve really pissed me off) is probably a bit higher than if Tabata’s just allowed to swing away. And if he is allowed to swing away, more often than not he’ll drive McCutchen in or at least move him to third. So what’s the benefit of the bunt in that situation? At best it takes the bat out of one of the Pirates’ best hitters and any thought process that validates that just isn’t acceptable to me.