Bunts Didn’t Take the Diamondbacks Higher

Maybe you’ve heard, but the Diamondbacks bunted a lot this postseason. I’m underselling it, of course — you definitely heard it. Maybe you saw Michael Baumann’s blow-by-blow recap of all the bunts. Maybe you read about Evan Longoria’s freelancing. Maybe you read Patrick Dubuque’s breakdown of teams trying to copy Arizona. Just yesterday, Russell Carleton actually went through all the math of which bunts are getting more popular league-wide in a typically excellent article. Maybe you read any of the countless other takes on it. But I have a different view. I think the Diamondbacks are being misrepresented. I think that it was more a case of a few opportunistic bunters than a team policy, and that their bunts didn’t alter the course of their offensive destiny much at all.

You’ve surely heard, in all that recounting, that the Diamondbacks led the majors in sacrifice bunts this year. It’s true! They also bunted a lot – either 68 or 69 times, depending on which bunt classification database you want to use. Corbin Carroll and Geraldo Perdomo combined to bunt 36 times this regular season. Perdomo was second in the majors in bunt attempts; Carroll was in the top 15.

Those two bunted with markedly different aims. Carroll was almost exclusively trying to bunt for a hit when he laid one down, or at least find a spot where reaching base was somewhat likely and a sacrifice was a good fallback option. I watched every one of his bunts; he never squared to bunt until after the pitcher started his motion. Carroll wasn’t particularly effective as a bunter, though. He singled on three of his first six bunt attempts this season, but after defenses adjusted to him, he made outs on his last seven. Here’s his best bunt, a great combination of base/out state and placement:

The defense wasn’t great there; Carlos Santana pinched in, but with a double play available and a runner on third, there was no one behind him to cover first. Even if Carroll had bunted into an out there, though, it was a good spot to try it, with a runner who could score and another runner who could advance.

That’s what you expect from a hitter as good as Carroll. When he’s bunting, you’d hope it’s in a good spot for a bunt, one where the rewards are high and the chances of getting on base are at least reasonable. Just one problem: Carroll didn’t make the net package work. He was 3-for-13 turning his bunt attempts into singles and produced negative WPA on balance. I like using WPA to analyze bunts, because it gives the hitter some credit for a successful sacrifice by taking context into account. But even with that benefit, Carroll’s bunting didn’t work out for him. He seemed to take that lesson to heart; after getting bunt crazy in June and July (eight attempts), he bunted only twice in August and once in September. Even in the playoffs, he only laid down two bunts, and I think they were both reasonable: a bases-empty attempt at a hit, and a savvy insurance run bunt, with a runner on second and no one out while the D-backs held a four-run lead.

I don’t think you can look at Carroll’s game and say that he’s bringing sacrifice bunts back. He’s clearly trying to bunt for a hit every time he bunts, and again, that makes perfect sense. If he were merely laying down a sacrifice, it would be a terrible decision; taking the bat out of Carroll’s hands is the last thing the Diamondbacks want.

Perdomo’s situation is different. He isn’t as good of a hitter as Carroll, for one thing, and the best Arizona bats come up after him in the lineup. During the regular season, he bunted early and often. Here’s one way of thinking about it: In bunt situations, with at least one runner on base and less than two outs, Carroll bunted 4.3% of the time. That’s high – 43rd in the league among hitters with 50 bunt situations – but it’s close to what I’d expect for someone with his speed and batting skill. Perdomo bunted 12.8% of the time, triple the rate.

No one who faced as many bunt situations as Perdomo bunted as frequently. The only hitters who bunted more frequently were bad: Austin Hedges, Johan Rojas, Zach Remillard, Martín Maldonado, and Nick Allen. Perdomo behaved more or less like a light-hitting catcher. I think his offensive skill set is better than that, but I also think it’s worse than Carroll’s, so it makes sense that he’d bunt more frequently.

In fact, Perdomo has done this since reaching the majors. He bunted just as frequently in 2021 and 2022. Teams are increasingly trying to take it away, and they’ve succeeded; he only converted four bunts into singles in 2023, a pretty miserable rate. To make the math work in the long run, particularly if you’re not bunting in a tie game in the late innings, you need to convert your bunt into a single around 30% of the time. It varies based on who’s on base, who’s up next, and all kinds of factors like that, but one third or so is a great rule of thumb. Perdomo was wildly short of that! He bunted 19 times in situations where he was making that exact tradeoff and only reached base on four of them (three singles and a successful squeeze play).

Give credit to Perdomo for one thing, though: he picked his spots well. Sacrifice bunts aren’t always bad decisions, even if they end in an out. Bunting to move two runners over late in a tie game increases your chances of winning if you start with no one out, for example. Another good rule of bunting: advancing the runner on first is less useful than advancing the runner on second. The best bunts put a runner on third with only one out. As that wonderful Carleton article points out, the very best bunts put a runner on third with only one out and are accomplished by poor hitters. That’s why Perdomo was such a frequent bunter, and why I’m so skeptical of the Carroll bunts. Overall, Perdomo lost WPA by bunting, but not a ton; he made up for his inefficiency at reaching base by picking good spots in general. With good hitters batting after him, it’s likely that WPA slightly understated his contribution, even.

I’ve focused on those two players because they were the only playoff contributors who were frequent bunters this season. Jake McCarthy, Jose Herrera, and Josh Rojas were third through fifth, respectively, on the team in regular season bunts put in play. Rojas departed the team in a trade this July, McCarthy missed the playoffs with injury, and Herrera played sparingly in relief of Gabriel Moreno. It’s not really fair to say that the Diamondbacks, as constructed in October, were built from top to bottom for sacrifice bunts. It would be more accurate to say that Perdomo bunts a ton, and that Carroll went for his fair share of sneak attacks.

That exact trend continued in the playoffs, and not to Perdomo’s benefit. His best bunts in the regular season came in no-lose situations: close and late games where sacrificing and reaching were both positive outcomes. In the playoffs, he didn’t attempt a single bunt in a tie game. He only attempted two bunts in the eighth inning or later, and both times the Diamondbacks already had a multi-run lead, which makes the whole play less important. But he bunted in the second, third, and fifth innings, and each one turned into an out. Don’t give up outs when there’s a lot of game left to be played:

Don’t give up outs when you’re trailing:

Especially don’t give up outs when you’re trailing if you only advance a single runner:

Those bunts are bad. Bunting when you’re down is bad! The math is easy to work out. If you’re trailing by one, the first and second runs are roughly equally important. The first gets you from 0% to 50% to win, and the second from 50% to 100%, assuming no further scoring. In a tie game, the first run is far more valuable than the second one. Plays that make the second run less likely to score are perfectly acceptable when you’re tied, and a disaster when you’re behind.

In the regular season, Perdomo acted according to that truism. He bunted seven times in tie games, and only four when the team was trailing. Carroll behaved similarly; he had five bunts in tie games and only four when Arizona was behind. But Perdomo bunted twice while trailing in the playoffs, and never in a tie game. Carroll at least only bunted with a lead – fine, but less exciting than bunting with a tie. It’s partially because he never had a chance, but for the most part, the team simply didn’t pick good spots to bunt.

Of the team’s 12 playoff bunts, Perdomo and Carroll accounted for eight. The others were largely improvised decisions, which worked to varying degrees. Ketel Marte dropped a sneak bunt single on the Dodgers that was never close to being an out:

That’s glorious, because it was a base hit. You shouldn’t bunt for a sacrifice early in the game, as our next contestant did. Even with runners on first and second and no one out, it’s a slightly negative win expectancy play. That can be mitigated if you’re a bad hitter and the batters coming up after you are much better. But Gabriel Moreno is a good hitter, so this clear sacrifice was a poor decision:

This bunt by Evan Longoria was worse:

Sure, he advanced the runner to second with only one out, but giving up an out to do so isn’t a good trade, particularly with the bottom of the lineup due up next. And Longoria had basically no shot of being safe at first, so the math was never going to work out. The only other bunt was Lourdes Gurriel Jr. dropping one down with a three-run lead in the eighth, which is a mediocre decision in a low-leverage spot.

The weirdest part of all of this bunt math? All those postseason bunts did very little to move the needle when it came to winning or losing games. All those bunts, the bad and the good together, combined for a grand total of -0.05 WPA, or 5% of a win. To be fair, Marte’s single was worth positive 5% all on its own, while the rest of the bunts cost the team quite a bit, but that’s how bunting works. If you never landed a single, it would be a terrible decision. The occasional successes make up for the larger set of small losses.

Maybe you don’t believe that WPA does a great job of capturing the effect of these bunts, so let me pitch it to you a different way. Most of the Diamondbacks’ bunts didn’t help them score. Seven of their bunts occurred in innings where they didn’t score a single run. Another four bunts came in innings where a batter who came to the plate after the bunter scored. In other words, the bases were getting cleared one way or another; the advancement from bunting didn’t produce any runs. In exactly one instance – Perdomo’s seventh-inning sacrifice in Game 2 of the World Series – a bunt led to a run that would not have otherwise scored if the following plays had still occurred in sequence.

Even stranger, all that regular season bunting didn’t actually help the Diamondbacks out much. They gained a fraction of a win, roughly 10% of one, across all of their regular season bunts. They finished exactly 15th in baseball by that metric. Their worst bunter, in terms of aggregate WPA lost? That’d be Perdomo, who cost the team 22% of a win with his bunts. Carroll was second with -0.19 WPA, or 19% of a win. In fact, McCarthy added half a win, with the vast majority coming thanks to a single bunt that took place in the fourth game of the season. For all the hullabaloo, Arizona’s bunting simply didn’t amount to much.

The best bunting team this year, per WPA, was actually the Royals. They didn’t bunt particularly often – 36 times, 14th in the league. The Rays were nipping at their heels, and they only bunted 26 times. Both teams did so well because they picked their spots and used the element of surprise. In fact, all the bunting in the majors combined didn’t amount to much. It sounds strange, but it makes sense if you stop to think about it. The whole point of a bunt is that it reduces variance; it removes the top end of great outcomes but also, in theory, the bottom end of abysmal outcomes. Nothing about a bunt is game-breaking.

This probably still isn’t convincing to people who love what the Diamondbacks were doing in October. Some people prefer that style of baseball, regardless of outcome. I’d just say this in response: The Diamondbacks didn’t generate the majority of their runs by bunting, either in the regular season or the playoffs. They generated negligible win value by bunting, in both the regular season and the playoffs. And before you point to their bunting as a reason that they outperformed their run differential, consider this: They batted .245/.315/.390 in low-leverage situations and .275/.362/.460 in high-leverage situations. If you could bottle that up and replicate it, you’d be onto something, but the bunts weren’t the secret sauce.

I wish I had more takeaways for you. I wish that different styles of play affected outcomes more. Somehow, though, it doesn’t seem to work that way. The Diamondbacks did some things differently. But those things just didn’t matter that much in the end – and for the most part, in fact, Perdomo was a one-man bunting revolution. He bunted more than six different teams. He and Carroll combined would have finished in a tie for 13th. The two of them combined for eight of the 26 total playoff bunts. The story here wasn’t that the Diamondbacks found a new way to play baseball; it’s that one of their players bunted a ton without great effect in either direction.