Rethinking Reliever Roles is fun and alliterative

For this off-day in the Nats’ schedule, I’m going to dust off an idea I first wrote about eleven-years ago as a potential solution for the Giants related to the declining effectiveness (at the time) of Tim Lincecum, albeit with some updates since then. I have long thought that there are potentially better ways to build bullpens than the Tony LaRussa-inspired modern version, with six or seven arms that are (except in rare cases) only capable of pitching one inning, with one or two guys at the end of the line who can go multiple frames in case of extra inning games or a starter who has to be pulled really early. Almost every team follows this model to the letter and has for at least thirty years. The only difference is with the 26-man expanded rosters that NL teams carry eight relievers now.

But there have been two big changes to major league pitching in general – one to starters and one to staffs as a whole – without a corresponding adjustment (by most teams) in how they manage bullpens (or rotations, for that matter). The first is the decline in innings pitched per start across the majors, which held steady around 5.5 throughout the steroid era but has dropped close to half an inning since the pandemic. That change has of course resulted in relievers having to pitch more innings (in effect, one more full-time reliever’s worth of 80 innings in a season), which up until the middle of last year most teams solved by carrying fourteen pitchers on their roster and churning through relievers with minor league options (hello, Tampa Bay!). However, because of the Rays’ actions (and other teams as well), teams are now limited to thirteen pitchers on their staffs and five options per player per year (for those with available options). Yet with the (mostly) one-inning model, that means pitchers who frequently play the role they do because they physically can’t maintain a starter’s workload are being asked to work three and sometimes four days in a row, or four times in five nights, or in both ends of a doubleheader.

Operating from the premise above, I propose a different strategy for building a major league bullpen. Rather than fill the pen with mostly guys who generally only go an inning or so, keep four or even five guys who are capable of turning a lineup over once (averaging, say, somewhere between 1.6 and 2.4 innings per appearance) while also making sure that none of them ever pitch on consecutive nights. With four or five, you could stack two of them in any given game while giving the others the night off, and still having two of the remaining three or four available for an inning in reserve.

But how to develop those types of relievers? It starts in the minors, and the development team has to work to make sure that as many of their young pitchers as possible have a third pitch that is at least average and allows them to be relatively effective at getting out hitters from either side of the plate. That in turn means identifying potential draftees and signings who have the beginnings of those tools, and not just big arms who only maybe have a second pitch because they can rely on just throwing smoke most of the time (which the Nats’ current regime has all too frequently been guilty of doing).

There are Nationals both past and present who were or are capable of pitching multiple innings in high-leverage situations. Tyler Clippard averaged almost four outs per appearance from 2009-11 (when he had to pitch a lot because the most reliable Nats’ starter was usually John Lannan). Mason Thompson is currently averaging 1.30 outs per appearance, even with his May downturn (which happened as a direct result of Davey Martinez going back to the well too frequently). Paolo Espino, whether as a starter or a reliever, has frequently been excellent the first time through an order and eminently hittable the second time (he should never even start a third time through the order for the rest of his career). Jake Irvin allowed just one run in his first ten-plus innings as a starter, but he profiles better as a reliever, and perhaps might be perfect for such a role. Thad Ward has history as a starter but might be better served in his post-Tommy John recovery having a reduced innings count. Further down, current starters with checkered injury histories or inconsistent stat lines as starters (salutations, Cole Henry, Jackson Rutledge and Andry Lara!) loom as potential candidates.

Let’s not forget the rest of the bullpen. This style would envision a one-inning closer and two or three other principally one-inning relievers, at least one of whom would also theoretically be trustworthy in high-leverage situations. Those would be ideal places to stick your fireballers, all the more so if you have more than one multi-inning reliever along the lines of Espino, soft tossers with pinpoint command. Then you could mix styles in a game in which you need three relievers, going soft-hard-soft or hard-soft-hard.

With the way pitchers are handled these days, teams are generally asking their starters to cover roughly 800-850 innings per season, with the bullpen taking 600-650, and in some cases up to 700. Having a couple of bullpen arms who are capable of taking, say, 90-110 innings each (perhaps as high as 120 if they get 65 appearances at just under 2 innings per appearance) plus two more in the 80-90 inning range eats up 340 innings on the low end and 400 on the high end, leaving 200-300 innings for your two best one-inning flamethrowers, the other short relief options, injury call-ups, and AAAA filler.

Why not build a potential 2025 pitching staff along these lines? Start with the rotation, including the 99.99% likelihood that Mike Rizzo drafts Paul Skenes second overall in a couple weeks and he is indeed of the Stephen Strasburg/Gerrit Cole mold and is ready for the 2025 Opening Day roster. Let’s also not speculate on potential free agent signings/trades/waiver claims so that we can avoid going down too much of a rabbit hole. Such a rotation could possibly be:

MacKenzie Gore (age 26 in 2025)

Josiah Gray (age 27)

Cade Cavalli (age 26)

Paul Skenes (age 23)

Jackson Rutledge (age 26)

That’s not a bad rotation at all, but it is also full of guys who haven’t proven themselves as workhorses yet in their careers, with a combined five starts of seven innings at the major league level (three by Gray, two by Gore). Ergo, a bullpen with several multi-inning arms would be helpful. Here is a possible eight-man bullpen, followed by depth arms in the minors (all currently in Wilmington or higher – apart from Jarlin Susana – so there’s a reasonable belief that they could be in Rochester or at least Harrisburg in 2025):

Hunter Harvey (age 30)

Kyle Finnegan (age 33)

Tanner Rainey (age 32)

Cole Henry (age 25)

Jake Irvin (age 28)

Jose Ferrer (age 25)

Thaddeus Ward (age 28)

Mason Thompson (age 27)

Joan Adon (age 26)

Jarlin Susana (age 21)

Evan Lee (age 28)

Dustin Saenz (age 26)

Jackson Tetreault (age 29)

Zach Brzykcy (age 25)

Jake Bennett (age 24)

Andry Lara (age 22)

Matt Cronin (age 27)

Amos Willingham (age 26)

Mitchell Parker (age 25)

Rodney Theophile (age 25)

Alemao Hernandez (age 25)

Andrew Alvarez (age 26)

Kyle Luckham (age 25)

If you would prefer to envision Susana as the closer (with any or all of Finnegan/Harvey/Rainey traded) and Brzykcy or Cronin or Bennett in that top eight in place of someone else, I’m not going to fight you on that – in fact, I would probably prefer it. There are others over whom the Nats would theoretically still have team control, but we’re going to ignore them because they’re either waiver wire castoffs who would be on the wrong side of 30 (Cory Abbott, Jordan Weems) or perpetually injured (Aldo Ramirez), or otherwise don’t have a clear path to still being in the organization in two years.

That bullpen only has one lefty (Ferrer, who would likely be a one-inning-only type), but the organizational depth chart accounts for Lee, Bennett, Saenz, and Cronin, all of whom except for Cronin could become multi-inning guys or potentially even starters. The organization could also use the regular long-relief roles as a way to ease young pitchers into the majors before moving them into the rotation (except in cases of outliers like Paul Skenes), but then of course the Nats would have to have a strong developmental pipeline, which we all know has been an issue for the entire Rizzo era. In fact, let’s pivot and do that, jettisoning two of the three over-30 relievers with the presumption that they have been traded or otherwise allowed to leave.

New bullpen:

Jarlin Susana (age 21)

Hunter Harvey (age 30)

Jose Ferrer (age 25)

Cole Henry (age 25)

Mason Thompson (age 27)

Jake Bennett (age 24)

Jake Irvin (age 28)

Thaddeus Ward (age 28)

Here I have separated the primarily one-inning guys from the multi-inning options, with a lefty in each category, although if you prefer Brzykcy to Ferrer I am once again not going to argue. In this hypothetical we are assuming Susana as the closer over Harvey. Now, how to break down the innings from this group? A standard one-inning closer who stays healthy the whole year is going to throw roughly 65 innings, so we’ll put that down for Susana, and the top set-up guy will typically throw a few more, so let’s give Harvey 70. Leaving aside Ferrer for the moment, I have put the multi-inning relievers in the order of their likely ranking (this is of course assuming a full return to health for Henry, which I know is a BIG assumption). Ideally Henry would throw about 100 innings, Thompson 90, and Bennett (being groomed for a starter’s workload and responsibilities) 120. That gets you to 445 innings from five relievers, and if you assume another 70 from Ferrer/Brzykcy you only need to get 140 more in total from Irvin, Ward, and the organizational depth (in rough order of importance: Willingham, Lee, Adon, Saenz, Cronin, Lara, Parker, Tetreault, with the last three perhaps more likely to be the first guys called up to the rotation for spot starts and injury replacements).

With that many innings being eaten in bulk by five or six guys in the bullpen, the Nats could do things like keep Skenes’ and Cavalli’s innings down around 150 or fewer for the regular season in expectation of a possible playoff run, or find it easier to give a guy like Rutledge a mid-season break, all without losing much in the way of effectiveness from the pitching staff as a whole. It could also be the optimal way to extract maximum value out of a pitcher as fragile as Henry, basically asking him to average five outs for sixty appearances in a season, with the stipulation that he never pitches on consecutive days (but that’s okay because you have Bennett and Thompson and Irvin who can all pick up the slack if the next day’s starter is ineffective!).

Whatever the personnel, I think an organization could do well running their bullpen in this fashion, building to that point by focusing on the following organizational principles:

  • Drafting/signing pitchers over throwers (some organizations like Cleveland already do this).
  • Preaching the importance in the minors of having secondary pitches that are effective against hitters from both sides of the plate.
  • Working as a starter as long as possible and not transitioning to a multi-inning relief role before the AA level.
  • Building strong organizational depth so that as guys either graduate from long relief to starting or flunk out of the role entirely, there are options to replace them.

Perhaps this model wouldn’t work for every team, but I’m fairly certain it would work for some if only they gave it a shot, because it would appear that with the new roster rules (about how many pitchers a team can carry) combined with the sharp decrease in starters’ innings across the majors, burning through arms an inning at a time is an unsustainable practice. Your thoughts and comments are welcomed.

This entry was posted in Bullpen. Bookmark the permalink.