In 1996, the New York Yankees won the World Series thanks in large part to an exceptional bullpen. If the Yankees had a lead after six innings, manager Joe Torre would hand the ball to Mariano Rivera (then just 26 years old and still a few years from becoming Mariano Rivera) for the seventh and eighth innings. Then John Wetteland, the All-Star closer who led the AL with 43 saves that year, would pitch a clean ninth inning to preserve the victory.
Rivera made this partnership possible; many teams had one Wetteland-quality one-inning closer, but few had two, and fewer still had a second relief ace who could throw closer-level innings two at a time. It was said at the time that Wetteland and Rivera were so good that it was impossible to score off them; “shortening the game” was a common description for what Rivera and Wetteland could do. Just as in the standard one-inning closer model popularized by Oakland’s Tony La Russa and Dennis Eckersley in the late 1980s, the Yankees divided the game into phases and approached each phase differently. The first six innings served to set up an advantageous position for a sprint to the finish; Rivera allowed the Yankees to sprint for longer.
“Shortening the game” is a good tactic for teams with the relievers to pull it off, and some have extended the sprint by employing four or five knockout relievers at a time; the 2014-15 Kansas City Royals, for instance, shortened the game to five innings by consistently turning to the bullpen from the sixth on.
But even the Royals used a traditional starting rotation; in 2014 and 2015, Kansas City’s starter lasted less than five innings just 47 times in 324 regular-season games, and in 33 of those starts, the pitcher got knocked out after allowing at least as many runs as innings pitched. Teams that followed in Kansas City’s footsteps leaned even more heavily on their relievers. Last year, the Tampa Bay Rays, who became the first team to make widespread use of the opener, pulled the starter after fewer than five innings 88 times. The Brewers did it 42 times, 12 times in September alone. In the playoffs, Milwaukee pretty much sprinted all game, every game; relievers pitched 60 2/3 of the Brewers’ 93 2/3 postseason innings.
The 2018 Oakland Athletics won 97 games despite pulling their starter before the fifth inning 53 times, and not having a single pitcher qualify for the ERA title. When they got to the AL wild-card game, they started reliever Liam Hendriks as an opener. Setup man Lou Trivino entered the game in the second inning and threw three innings, and closer Blake Treinen entered in the sixth inning with Oakland trailing by three runs.
Pitcher usage trends in baseball aren’t always linear, so it’s possible that someday the pendulum will swing back and starters will take up more of the slack than they are right now, but in the short term, workhorse starters are being replaced by groups of relievers who throw at maximum effort for one or two innings at a time.
It’s not enough to just have one or even two relief aces whose primary purpose is to protect a lead late in the game; key bullpen roles and assignments are becoming more fluid and numerous, requiring as many as half a dozen hard-throwing relievers to fill them. Stacking high-quality relievers isn’t about shortening the game anymore. It is the game.
It’s no longer enough to find one closer for the ninth, then another closer-quality reliever for the eighth, and so on; relievers have to be ready to enter the game within an inning, and record more than three outs at a time. Therefore bullpens can no longer be considered a chain of individuals coming in to perform discrete tasks—a modern bullpen is a team within a team, like an offensive line or a center back pairing.
“It’s absolutely a unit,” said Seattle right-hander Anthony Swarzak, who pitched in loaded bullpens in New York, Chicago, and Milwaukee before joining the Mariners. “I remember with the Yankees in 2016, they had [Aroldis] Chapman, [Andrew] Miller, and [Dellin] Betances. It was a three-headed monster. It’s absolutely a weapon in the game that you look collectively at those three guys, and a pretty strong one. When you try to break it down into individuals, I don’t think it has as much punch.”
Sometimes, a more fluid bullpen truly means that any pitcher could pitch in any situation, depending on leverage. Treinen, for instance, saved 38 games but also made 19 appearances of four outs or more and pitched 25 times outside of save situations, earning the win in nine of them. Oakland had the third-best bullpen in baseball last year according to wins above average, and the best according to win probability added, and was able to shuffle pitchers around because just about everyone was comfortable throwing in many different game situations, often for more than one inning at a time.
“The strength of the bullpen was we had a bunch of guys out there who were capable of going multiple innings,” Hendriks said. “I think generally that makes things easier. … Look at the Rays. The Rays are stockpiling guys who can go multiple innings right now, and they had a pretty decent bullpen last year as well. We’re doing the same sort of thing. It helps shorten games.”
One advantage of the closer model is that it’s easy to manage. Closers know when they have to start loosening up based on the score and inning, and each pitcher settles into a role he’s comfortable in. Doing what Oakland or Milwaukee did last year, or Cleveland in its run to the pennant in 2016, takes a little more managerial finesse, both tactically and in terms of man management.
“The game is fast,” said Oakland left-hander Ryan Buchter. “When you’re out there—and I’m not saying it happens to [Bob Melvin]—it sneaks up on you, and with these young managers these days, if they don’t have any managerial experience it sneaks up on them and it gets hard.”
Orchestrating the timing of a call to the bullpen is harder than it looks. Unlike in a video game, pitchers need advance warning, often an inning or more, in order to stretch and warm up, and much of the exertion of pitching comes from throwing warm-up pitches. That not only means a pitcher can’t just stay warm indefinitely, it means that a pitcher who warms up but doesn’t enter the game—a practice known in sophomoric baseball parlance as “dry humping”—will often be almost as fatigued the next day as a pitcher who threw an inning out of the pen.
Skippers who know how to manage their bullpens tend to avoid dry humping; if a reliever gets warmed up, he comes in. In Game 4 of the 2016 World Series, for instance, Indians manager Terry Francona had his relief ace, Andrew Miller, warming up to protect a three-run lead in the seventh inning. While Miller was getting ready, Cleveland stretched the lead to six, but Miller came in anyway and threw 27 pitches over two innings; if he was warm, he might as well pitch.
But even if a manager knows which buttons to push, and when, he has to consider the people attached to those buttons. Relief pitchers exist on a continuum of flexibility; some closers get amped up to pitch one inning with a lead, and one inning only. Hendriks, on the other hand, enjoys coming in with runners on base and doesn’t care much what his role is “because I get to the field every day expecting to pitch.”
We tend to think of baseball managers as coaches or leaders, but in this respect, the title of “manager” means the same as it would in a department store or an office setting. The manager is responsible for not just creating a plan of action but communicating it to his subordinates in a way that gets everyone to pull on the same end of the rope. This has always been true, but it’s even more important when a pitcher might throw in the eighth inning one night and the fourth inning the next. Most pitchers can adapt to a new role, but they adapt more easily if the coaching staff keeps them informed.
“If you’re being used in a certain role and things start to get switched around, we’re humans and our brains start to go a certain way,” Buchter said. “So even if you’re on a hot streak and all of a sudden you don’t pitch the inning you’re supposed to, the first question that comes up in your head is ‘Did I lose my job?’ or ‘Am I not pitching this inning anymore?’ I think everyone pitches better with confidence. It’s not anybody’s arm or anybody’s body, it’s their confidence that starts to go first.”
But once the manager and the coaching staff get buy-in, staffwide confidence is contagious, as is the willingness to put the team first. Buchter and Trivino both pointed out how important it was for team chemistry that Fernando Rodney, a 16-year veteran with 325 career saves, accepted a middle relief role.
“When you see guys come in like that and take a role in the seventh inning, and he’s happy every day,” Buchter said, “all the younger guys see that and go, ‘Oh, crap, this is what it’s all about. Winning is more important than anything.’”
Hendriks, an eight-year veteran who’s pitched for four different teams, smiled wryly and shook his head when asked whether Oakland’s high bullpen morale is the norm. But he went on to explain that an individual pitcher’s inflexibility doesn’t necessarily spell disaster for the entire unit.
“There are some things that come into it that aren’t always about helping the team, but sometimes it’s just like, ‘I need a set role. That’s the way I pitch better. It’s going to help the team in the long run but I need to be rigid,’” Hendriks said. “It’s all about the type of person you are—there’s nothing wrong with being set in [your ways], but it’s a little easier on a manager when you’re able to be thrown around anywhere from the first to the ninth inning.”
In some respects, Miller’s extreme durability and flexibility, on the giant stage of the 2016 postseason, spoiled onlookers and created an unrealistic expectation for what a relief ace can do. Not everyone can perform at a high level for more than an inning on back-to-back days. Some closers, even great ones, struggle when entering an inning with men on base. Sometimes it’s just a matter of comfort, but sometimes individual closers have specific flaws: They don’t control the running game well, or their arsenal leads to an unusually high number of walks or wild pitches, which aren’t a big deal with the bases empty but can be the difference between victory and defeat with runners on base.
In other words, different relievers still have different roles, but they’re based on the leverage index and number of runners on base rather than the inning. Craig Kimbrel isn’t less valuable than Miller—he’s just better suited for a slightly different job.
“You need two guys, really,” said Hendriks. “You need the guy who’s going to be ready no matter what, and you need the closer who can go out there and get the last outs.”
Hendriks is mostly right—teams do need a closer and a fireman, like Cody Allen and Miller on the 2016 Indians. But that’s not enough anymore; they also need three or four other hard-throwing relievers to fill in the innings in between. The good news is that closer-quality relievers are easier to find now than ever. The average leaguewide fastball velocity has gone up 10 years in a row, and is now nearly 2 miles an hour faster than it was in 2008.
“I think it’s just like any other sport,” Trivino said. “You look at sprinters from the 1940s, [compared to now], guys are just getting better. It’s not that athletes are better—conditioning and the practice regimens are better. We know more now about how to throw hard than we did 10 years ago.”
Unlike the gradual progression of something like the 100-meter-dash world record, modern baseball is also full of success stories in which individual pitchers make one mechanical tweak and turn into fireballers overnight.
“That’s exactly what happened [to me],” Trivino said. “It’s incredible. I’d warm up and throw as hard as I can and it’d be 88 or 90. When I made that tweak, I was nice and easy and I looked at the scoreboard and it said 91. I said, ‘Oh baby, here we go.’”
An 11th-round pick out of Slippery Rock University in Pennsylvania, Trivino went from throwing in the high 80s to averaging 98.2 miles per hour on his four-seamer last year. That’s the eighth-highest average fastball velocity among the 191 relievers who threw at least 40 innings last year, and Trivino held that velocity over 73 innings, the 21st-greatest workload among MLB relievers in 2018. The adjustment came so easily that Trivino might as well have been bitten by a radioactive spider, yet his story is far from unique.
“If you make an adjustment and it just feels good, you know it’s going to come easy, then you can build on it and work with it,” Swarzak said. “When you try an adjustment that doesn’t come as fast, generally that’s not going to be the adjustment that gets you to the next step. Now if it’s a pitch, then that pitch takes time to develop, but if it’s an adjustment to your delivery and it’s just not coming yet, then it’s probably not going to work.”
The Twins drafted Swarzak out of high school, where he threw a four-seamer and a curveball, but developed him as a sinker-slider guy and encouraged him to pitch to contact. By 2017, he was back to his original repertoire and posted career bests in K/9 ratio (10.6) and ERA+ (187) over 77 1/3 innings and earned a two-year, $14 million contract from the Mets in free agency, though he struggled to stay healthy in 2018 and was flipped to Seattle in the Robinson Canó deal.
Swarzak chalks up much of the nastiness of current big league pitching to the proliferation of pitch tracking technology and high-speed cameras. Not only does the additional data make it easier to evaluate and teach pitchers how to throw hard or spin a better breaking ball, the new technologies allow coaches to present pitchers with hard evidence of what they’re doing right or wrong, which increases trust between players and coaches.
“A coach used to be a good coach 15 years ago because he had experience in a lot of different areas, and could bounce some things off a player to see what stuck,” Swarzak said. “That style of coaching is finding its way out of the game now because there’s too much left to chance.”
The end result of this process is a veritable army of hard-throwing pitchers—youngsters and rejuvenated veterans alike—filling out big league bullpens. In addition to high-profile trends like popularization of the opener, teams are leaning more on their bullpens generally. Since 1908, there have been 12 individual seasons in which one team’s relievers have combined to throw more than 600 innings; eight of those came in 2018.
Bullpen-heavy pitching staffs are more effective in some cases, but because baseball’s financial structure still rewards high-workload starters over middle relievers, they’re also cheaper to run. The two teams that finished 2018 with the most innings pitched by relievers, Tampa Bay and Oakland, had the lowest and third-lowest payrolls in baseball, respectively. Arbitration, which Buchter called “a broken system,” still rewards relief pitchers for racking up saves, even as teams recognize players’ value through other means in free agency.
“I’m sure that’ll figure itself out,” Hendriks said. “You look at Andrew Miller. He didn’t rack up that many saves last year but still got a nice contract this offseason. Teams are starting to look at what we’re doing on the field now, it’s not just, ‘Oh, you have 45 saves, we have to give you this much.’”
But pitchers have to make it to free agency—six or seven years of MLB service time—in order to cash in, which isn’t a given. Throwing as hard as relief pitchers do now is bad for the human body, and an ill-timed injury can slash a pitcher’s career earning potential.
“On the mound, I literally just try to throw everything as hard as I can now, because if you throw really hard”—here Swarzak’s eyes widened as he leaned forward for emphasis—”you’re going to have a job. It’s almost as dangerous to sit up here and back off, throw it 90, and then what? OK, maybe I can play catch when I’m 60, but who wants to play catch when they’re 60? I want to throw hard now.”
Increased reliever usage is just one of many impetuses for the game’s economic structure to change, and the game’s economic structure is just one of many things that will have to change about baseball as relief pitchers take on a larger share of the workload. As bullpens like Milwaukee’s and Oakland’s become more common, batters will have to find a way to score runs as the game gets shorter.
“I’d hate to be a hitter,” Trivino said. “I don’t know how they do it. I couldn’t do it.”