It’s common knowledge that MLB analytics departments are reducing the role of the starting pitcher and shifting its workload to the bullpen. At the same time, unintended consequences of this shift - primarily the reliever exposure effect - have been documented and (as Ben Lindbergh covered last postseason) seemed to be gaining momentum. Has that momentum carried over into 2024? Do bullpens operate differently now?
What is the exposure effect? The idea is similar to the “times through the order penalty” (TTOP): the more at-bats a hitter gets against the same pitcher during a game, the better the results get for that hitter. Hitters thrive with familiarity.
The TTOP typically isn’t applied to relievers over the course of one game - relievers aren’t likely to see the same hitter twice in a single game. But that’s not true over the course of one series, where a reliever might be brought in for the same hitters two (or even three) times. When a hitter gets to see the same pitcher in the same series, a familiarity effect pops up. Ben Lindbergh covered this concept in the context of the playoffs last October. Cameron Grove dug into the data in 2022. It’s hard to compare starters to relievers - they pitch differently. But I think it is fair to say that the same-series reliever exposure effect is comparable to the TTOP for starters.
Why is it relevant in 2024? Teams should be recognizing that minimizing the TTOP for starting pitchers (through shorter outings) comes with a tradeoff. In passing a bigger chunk of innings to your bullpen, you’re creating additional opportunities for opposing hitters to develop familiarity with the same relievers - particularly when you play matchup and line up each reliever against the stretch of hitters he profiles best against.
The reliever exposure effect is not a huge deal in a three-game series against a non-division opponent that you see once a year. But it can matter in a four-game series or against a team you play again in a week or two.
So this is a thing now? Are teams re-designing their pitching plans? Not yet. Or, not universally. The Astros (and perhaps the Yankees) both showed us in their opening four-game set that they’re probably not factoring reliever exposure into their gameplanning. Here’s why.
The closers on each team (Josh Hader and Clay Holmes) each pitched three out of four games in this series (last Thursday, Saturday, and Sunday). Both managers let typical “closer rules” dictate their use of each pitcher. For the Astros, that meant Hader faced the following hitters in each game:
Thursday: Stanton, Rizzo, Volpe
Saturday: Cabrera, Grisham, Torres, Soto
Sunday: Trevino, Cabrera, Torres, Soto, Judge
The hitters in bold had faced Hader less than 24 hours prior to their Sunday at-bats against him. Soto’s at-bat on Sunday was a memorable one. He worked Hader for 8 pitches and shot this to left field for the go-ahead RBI:
Impressive. Oswaldo Cabrera and Gleyber Torres aren’t Juan Soto, but neither went down easy against Hader. Check out the way Cabrera took this 0-2 slider:
That take looks a little too comfortable for an Oswaldo Cabrera-Josh Hader matchup. It looks more like a Juan Soto take. When Cabrera saw Hader for the first time in this series (less than 24 hours earlier), he didn’t look so comfortable:
That’s what familiarity can do. Does all this mean that Josh Hader shouldn’t pitch on back-to-back days? Of course not. But it does mean his manager can help him out by picking his spots. For example, Hader’s Thursday outing came with the Astros down a run in the 9th on Opening Day. That’s a low-leverage situation for Hader to be used in. With three games remaining in the series, it also gave three Yankee hitters a bonus look at him. So inserting him into that spot is hard choice to justify.
Managing reliever exposure also requires flexibility in the bullpen hierarchy. Had Hader not pitched in Houston’s Opening Day loss, he and Bryan Abreu also could’ve swapped the 8th and 9th inning on Sunday with a very limited exposure penalty.
So that’s the Houston angle. What about Clay Holmes’ matchups against Astros hitters?
Thursday: Dubón, Altuve, Alvarez, Tucker, Bregman
Saturday: Diaz, Abreu, Peña, Dubón
Sunday: Peña, Caratini (hit for Meyers), Altuve, Alvarez, Tucker
Peña had a good at-bat to lead off Holmes’ Sunday outing. Altuve and Alvarez each took appetizing chase pitches before retiring on hard-hit outs. All had previous looks at Holmes earlier that weekend.
For the Yankees, avoiding Holmes’ overexposure would’ve been an unpopular move. He pitched in high-leverage spots in all three games he entered. To shift his appearances to minimize his exposure would have required a medium-to-high leverage inning from a less-trusted contributor like Victor González. Boone didn’t make that decision, but he paid a price for making the conventional decision. Like Hader, Holmes had to overcome both familiarity (facing the same set of Astros hitters) and fatigue (his third outing in four days) on Sunday. And because Holmes pitched Sunday, he wasn’t available for a save situation that came up the next day against the Diamondbacks.
Keeping track of reliever exposure is complicated - it means a lot of additional pre-game preparation for managers. But that doesn’t mean it isn’t worth doing.
It gets worse
The Astros had another opportunity to leverage the reliever exposure effect on Sunday - this one on the hitting side. You’ll notice above that Mauricio Dubón faced Holmes twice prior to Sunday’s game (Thursday and Saturday). And yet, with Dubón and Caratini on the bench on Sunday, manager Joe Espada chose Caratini as his pinch hitter against Holmes.
There are a few reasons why Espada might’ve picked Caratini. He’s left-handed, so he has the platoon advantage. Dubón is also valuable as a pinch runner (which is how he ended up coming into the game). But whatever the actual reason was, Dubón’s familiarity against Holmes was clearly not enough of a factor for Espada to justify him instead of Caratini in that spot. Maybe it should have been.
But he’s already hot!
Here’s an example from another four game series - this one between Pittsburgh and Miami.
In the middle of the 6th inning of Game 3 of this series, Pittsburgh went to its mid-inning setup reliever (Ryan Borucki) to hold onto a 5-2 lead. It was Borucki’s second appearance in the series.
Borucki got out of the 6th inning having thrown only two pitches. Leading 5-3, Pittsburgh’s bullpen was quiet; Borucki was going to head back out for the 7th.
Then the Pirates offense erupted. After a walk, two singles, and two doubles, the score was 9-3 and no longer required a high-leverage reliever like Borucki. With two outs in the inning, Pittsburgh began to get another reliever (Hunter Stratton) loose.
Manager Derek Shelton could have handled the bottom half of this inning in two ways. He could’ve let Borucki have it; Borucki had warmed up, entered the game, and only thrown two pitches. You wouldn’t want to waste all of that effort.
The other route would be to yank Borucki from a game that had been blown open. That approach would limit his exposure to Miami hitters ahead of the final game in the series.
Shelton chose an approach in between these two options. He let Borucki face one hitter in the 7th and then brought Stratton in (it’s possible Stratton wasn’t ready to start the inning).
After Borucki was yanked from Game 3, Stratton faced three hitters: Christian Bethancourt, Luis Arraez, and Josh Bell. The next day, Pittsburgh entered the 7th with a 7-6 lead and Arraez, Bell, and Jake Burger due up for Miami. Borucki came in for his third appearance of the series, but Shelton’s decision the day before meant that Borucki didn’t have to worry about exposure to any of those hitters.
What does this example highlight? The exposure effect further limits the number of bullets high-leverage relievers can fire. If your closer hasn’t pitched in three or four days, handing him the 9th inning with a 3-run lead might seem like a harmless move. But in Game 1 of a series, it’s probably a bad idea.
Where from here?
There’s a lot we don’t know about the exposure effect. Just how big is it, exactly? What’s the specific mechanism behind it? Does the amount of exposure (pitches seen) matter? How long does it last?
These are all questions that can be answered with publicly available data. While there isn’t much evidence that teams are making pitching decisions based on the exposure effect yet, soon enough, we’re going to see pitching plans (regular season and postseason) reflect this type of research. The “arms race to nowhere” that occupies front offices just got a little longer.
Have a great weekend!