Quantifying last night’s sacrifice bunt attempts

Many have criticized Clint Hurdle for the number of sacrifice bunts the Pirates attempted in last night’s 5-4 win over the Rockies. All in all, the Pirates laid down two sacrifice bunts on the night. In addition, Jose Tabata showed bunt on the first three pitches he saw in the seventh, before eventually singling to right field. Alex Presley tried to sacrifice in the eighth, but ended up reaching base on an error. Tabata then again attempted a bunt, but failed to get it down, popping up to the catcher. Let’s break down each of these bunt attempts from a strategy standpoint, ignoring the eventual results. For this purpose, I will look at Run Expectancy (RE).

One of the most common mistakes many people make when using RE to evaluate the use of a sacrifice bunt is that they try to compare the RE prior to the at-bat to the RE after the at-bat, while assuming the bunt is successful. This is essentially comparing apples to oranges. Consider the first inning of last night’s game. Presley led off the inning with a double, and Jose Tabata bunted him to third. In a basic sense, the bunt decreases the RE for the inning by making an out. But we also must consider that Tabata makes an out approximately 66% of the time when he swings away, going by his career on-base percentage of .341. In other words, there is a good chance the RE is going to decrease whether he swings away or bunts. The key is to compare the expected change in RE if he swings away to the expected change in RE if he bunts. There are many other variables that should also be considered. At what rate is the hitter successful in laying down the bunt? How frequently does he fail, making an out without moving any runners? How often does the hitter himself reach base on the sacrifice attempt, by way of an error or simply beating the throw to first? And so on.

In an attempt to incorporate as much of this information as possible, I used 2nd Guesser, an iPhone app created by the Mariners blog U.S.S. Mariner. The app allows you to quickly plug in a game situation and calculate how a sacrifice bunt would affect the RE of the inning. It accounts for the hitter’s on-base percentage, the probability of a bunt succeeding vs. not succeeding, the probability of a bunt leading to everyone being safe and the probability of a bunt resulting in a double play. It is not perfect, as it does not include all possible pieces of information. Some missing components are the hitter’s extra-base hit ability, the offensive ability of the player on deck, the speed of both the hitter and runner, the possibility of a double play when swinging away, and so on. It also uses a league-wide average for sacrifice bunt success rate, as opposed to individual bunting ability. But all in all, 2nd Guesser does a good job of giving a quick framework of RE. (Unfortunately, it is no longer available in the iTunes store, but I have access to it because I purchased it early enough.) So let’s take a look at last night’s bunts. As a reminder, I am ignoring the actual result of each bunt. I only want to look at what was known before Hurdle called for the bunt. (Note: Succeed+ indicates a sacrifice attempt that leads to the hitter reaching base as well.)

First up, the first inning. As mentioned earlier, with the Pirates trailing 2-0, Presley led off with a double and Tabata bunted him to third. As you can see, the RE was 0.14 runs higher if Tabata would have swung away.

We move to the bottom of the sixth. After Casey McGehee led off with a walk, Neil Walker bunted him to second. Walker may have been trying to bunt for a hit here, but I am including it because I am not sure. In this situation, the RE was 0.10 runs higher if Walker would have swung away.

In the seventh, Presley again led off with a double and Tabata squared to bunt. After getting ahead in the count 2-1, Tabata began swinging away and eventually singled. I included it as a bunt attempt because it seemed clear that the Pirates wanted to sacrifice initially. Since this was the same situation as the first inning, RE again was 0.14 runs higher if Tabata would have swung away.

The next inning, after Clint Barmes homered to tie the game, Garrett Jones pinch hit and walked. Presley tried to bunt him to second and reached on an error by pitcher Matt Belisle. RE was 0.10 runs higher if Presley would have swung away.

Tabata followed Presley and also tried to bunt. The bunting luck had run out on the Pirates by this point, as Tabata popped up to catcher Ramon Hernandez. This is our first wash, as the change in RE was the same for both bunting and swinging away.

In the end, the Pirates’ RE was 0.48 runs lower for the game because of these attempted sacrifice bunts. Sometimes it is important to remind ourselves that a manager has more information than we do. He knows his players better than we do, and he likely has a better read on what they can do at the plate. However, potentially losing a half a run per game by taking the bats out of the hands of your hitters is both significant and damaging.