The Peacekeepers

The 300-inning season is dead. It has been for a while. The last one was Steve Carlton’s 304-inning Cy Young campaign of 1980. The 250-inning season died a much swifter death, going from an average of 11 per year in the 1980s (even with a zero in the 1981 strike season) to just three per year in the 1990s. Even putting 10 in each of the two strike years of ’94 and ’95, it still gets to an average of only five per year. There have been nine total since 2000 and the last one was in 2011. The 200-inning season is in hospice at this point, taking a sharp turn for the worse in the last two years.

After averaging 35 a season from 2010-2015, we’ve seen just 30 in the last two years combined. This is not because pitchers as a whole have become soft. It’s quite the opposite, in fact. Velocities are way up, and secondary stuff is as nasty as ever. Several pitchers in today’s game could be groomed for 250- or even 300-inning seasons, but they simply aren’t asked to do that. Starters are filling a six-inning role. Go all out for about 100 pitches and get as close as you can to 18 outs. With that directive, you can attack everyone.

In previous eras, seven innings was the baseline for the front end of rotations, and those pitchers went out expecting to complete the game. A pitcher can’t go max effort from the first pitch and expect to throw eight or nine innings.

And make no mistake, part of the innings decline is just fewer starts in general. We are 30 years removed from a 40-start season, but the 35-36 start peak of the 2000s is inching downward, with only David Price in 2016 reaching the mark since 2011.

A key driver behind the change is teams paying more attention to the “Third Time Through” penalty, which is essentially a cool way of saying “fatigue.” Hitters start to gain an advantage the more they see a pitcher, especially as that pitcher is wearing down. The biggest jump is from the second to third time, which of course occurs in the middle innings.

While it is discussed more often than before in the online baseball culture these days, this phenomenon is not a new thing. The ever-brilliant Baseball Reference shows us that in 1987, batters went from a 94 to a 109 OPS+ with an additional 41 points of slugging percentage when comparing their second plate appearances to their third against a starting pitcher. The 2017 hitter pool went from 95 to 113. Narrowing our view, we see the number of TTO plate appearances in the last 10 years has dropped by 4,764, or about one per game.

As always, there are multiple factors triggering the changes in the game. Teams are more apt to pay attention to an individual pitcher’s TTO numbers and adjust their deployment of him accordingly.

Additionally, the rise of power relievers has allowed teams to feel more comfortable pulling their guys when they have flamethrower after flamethrower in the bullpen. The data go back only to 2002, but reliever fastball velocity has jumped four mph, although the way velocity readings are captured was changed last year, which essentially added a free tick, so it’s probably more like three mph. That’s still quite a leap in 16 seasons.

Over the last 30 years, relievers have garnered six percent more of the inning workload, up to 38 percent this past season. Most of the gains have come recently, too. It went up just one tick to 33 percent from 1987 to 1997, then plus two percentage points in the next 10 years, and finally a plus-three-percentage point jump over the last 10 years through last season. Instead of a return to the super relievers of yesteryear, the added workload is just going to more relievers churning through every team. The major leagues haven’t seen a pure 100-inning reliever since 2006 when Scott Proctor did it for New York in an AL-high 83 games.

Fun fact: Proctor threw in another 83 games in 2007 but  totaled only 86 innings, and his K-BB rate tumbled from 13 percent to five percent. He’d throw just 85 innings with a 6.59 ERA the rest of his career. Perhaps his subsequent collapse scared teams from using a reliever for 100 innings out of the bullpen. We have seen five 100-inning seasons by four primary relievers (80-plus percent relief appearances) since then, but each of them logged at least five starts:

Even 90-inning seasons pitched entirely in relief are scant in recent years with just six since 2010, the latest coming in 2014 when Dellin Betances tallied exactly that number for the New York Yankees. Josh Collmenter (92 IP in 2013), Anthony Swarzak (96 in 2013), Jim Johnson (91 in 2011), Matt Belisle (92 in 2010), and Tyler Clippard (91 in 2010) are the others. They all achieved the feat in fewer than 80 appearances. Pre-2010, seven pitchers had logged 90-plus innings, all in at least 80 relief appearances, with a peak of 94 by Salomon Torres in 2006.

Before Torres, Kent Tekulve’s 1987 was the last 90+ appearance season, and only Pedro Feliciano has done it since, entering 92 games in 2010 though tallying just 63 innings with the New York Mets. The success or failure of a few guys isn’t enough to determine the viability of a strategy, but I think using 80-plus appearances to log the innings is where trouble occurs. It’s generally not the raw pitch count that breaks a pitcher down but the pitches thrown while fatigued that do the damage.

This has led to the increased specialization of relievers. Not only have most been relegated to one-inning roles, but left-handers who specialize in getting left-handed hitters out might face only a single batter in a given outing. Feliciano is the perfect example of a recent-era LOOGY. From 2008 through 2010, he appeared in 266 games, logging just 175 innings. He led the league in appearances all three seasons at 86, 88, and the aforementioned 92. He missed two seasons due to injury and then pitched just 11 innings in 2013 before calling it quits.

With 90-to-100-inning relief seasons already seemingly outdated in today’s game, the 70-inning season could be next. After almost 45 pitchers per season completed the feat from the 1980s through the 2000s, there have been 27 per season in the 2010s.

Teams still have to fill the 1,458-plus innings per season, and to do so they are relying on more relief pitchers than ever. The league has set a record for relievers logging at least one inning in each of the last four seasons, topping out with 429 in 2017. Compare that to 360 in 2007, 252 in 1997, and 170 in 1987. Dialing it back all the way to 1957, there were 69 such relievers.

Think about that: 69 relievers were used in all of baseball. The Mariners used 34 themselves in 2017. Total relief appearances have nearly doubled since 1987, going from 7949 that year to 15,659 in 2017, by far a major league high.

The fungibility of relievers isn’t lost on teams, allowing them to use a couple of bullpen spots to shuttle  guys on and off the roster, but there is still room for something beyond the one-inning role. The specialization of relievers has shortened the appearances but upped the volume of total appearances. Relief appearances under an inning are way up, and relief appearances going beyond an inning are down.

Teams could lessen the burden on their minor league relievers by identifying a guy (or two if they have the resources) for the two-to-four-inning role. It would be a return to the role that was quite prevalent in the ’70s and ’80s. In fact, the multi-inning relief role didn’t really fall off until recently. The ’80s saw 2,500 per season, the ’90s saw 2,240 even with two seasons impacted by the strike, and there were at least 2,000 in the first five seasons of the ’00s. Since 2005, the average is just 1,634, but it’s on the rise after a dip below 1,400 in 2014, the first such dip since the late 1950s.

It seems weird that we don’t have any present-day versions of old-school firemen like Rollie Fingers, Rich Gossage, and Tekulve. Fingers averaged 115 innings per season in 69 games from 1974 to 1980 with zero starts. Gossage started 29 games in 1976 but put up just a 91 ERA+ in 224 innings before going back to the bullpen for an average of 93 innings in 55 games. He downshifted to 49 innings per season in 40 games from 1986 through 1994, which covered his late 30s through his final season at age 42. Tekulve made a career out of high-volume relieving, leading the league in appearances four times and averaging well over an inning per game with 1,437 innings in 1,050 relief appearances.

There were several others, too. Hoyt Wilhelm is the two-to-four-inning god with a major league-best 435 such appearances over his career. Guys like Gene Garber (373) and Lindy McDaniel (371) aren’t as heralded as the others highlighted thus far, as they tallied just one All-Star appearance between them in 40 years of playing, but both were solidly above-average firemen. Garber had a 117 ERA+ in 1,510 innings, and McDaniel was at 110 in 2,139. Tug McGraw gets remembered while Garber is essentially forgotten, but they were the same guy. McGraw had 1,515 innings of 117 ERA+. Garber had 38 more saves and matched McGraw’s eight percent K-BB rate.

Mike Marshall and Sparky Lyle both nabbed Cy Youngs as elite firemen, Marshall’s coming in an eye-popping 208 relief innings. Lyle threw 137 but probably had no business winning despite a sharp 2.17 ERA. Jim Palmer (2.91 ERA in 319 IP) and Nolan Ryan (2.77 ERA in 299 IP) both have reasons to be miffed about that 1977 Cy Young vote.

Marshall was hardly a clear winner in 1974, but appearing in two-thirds of his team’s games (106) and being dominant carries a lot of weight. Phil Niekro posted a 2.38 ERA in 302 innings and had to be scratching his head on finishing third. It’s not at all the point of this piece, but Niekro was ripped off multiple times when it came to Cy Young voting.

Willie Hernandez somehow won a Cy Young and MVP for his power relief season in 1984 when he posted a 204 ERA+ and 0.94 WHIP in 140 innings over a major league-best 80 appearances, 68 of which were games he finished, also the best in the majors. It was just something top relievers did a lot more in the past.

There have been 20 relievers with at least 200 appearances of two to four innings over the last 40 years, many with names you’ll be familiar with, but only one guy has topped 150 since 1997: Scott Sullivan (158). The most popular guys to log a decent number of two-to-four-inning appearances over the last 20 years would be Shigetoshi Hasegawa (132), Keith Foulke (114), and Ramiro Mendoza (99). Yusmeiro Petit (22 in 2017) and Erasmo Ramirez (20 in 2016) are the only two guys with at least 20 since 2014.

The playoffs could be ushering the role back into vogue. Tyler Kepner of The New York Times has identified a twist on the role that is worthy of a name change: The Peacekeeper.

To really be a fireman, though, there needs to be something to extinguish. In the modern game, especially in the World Series, relievers are more like peacekeepers than firefighters. You call for them before the blaze starts, just to keep things stable.

The 2000s postseasons have ushered in a wave of extending relievers beyond their regular season standards. Part of the increase in appearances of more than one inning is just playoff expansion offering more opportunities to do so, but even within that framework, we’ve seen a substantial surge the last two years.

Taking it a step further, there hadn’t more than 38 relief outings of two-plus innings since the 51 in 2004 until this year’s surge to 52. Those are actually the top two seasons ever by a long shot, with 1999 (39), 2003, 2005 (38), and 2016 (36) as the only others north of 35. Obviously, the playoff schedule is conducive to such deployment because the built-in off days make it easier to use your best relievers almost every single game. There are limitations in the regular season, when a team has to be cognizant of getting through 162 games without taxing its relievers to a point where they aren’t available in October.

Can we take a brief interlude to appreciate Mariano Rivera’s sheer brilliance? Everyone knows he was a stud, but the details may have been forgotten by even those who lived through his career. He logged more than 80 innings only twice in the regular season through his storied career, but he has more than twice as many two-plus-inning appearances in the playoffs than anyone else, with 33.

In those appearances, he had an insane 0.52 ERA and 0.63 WHIP in 69 innings of work. His overall playoff line is just stupid: 0.70 ERA, 0.76 WHIP, and 17 percent K-BB in 141 innings. He allowed a whopping 13 runs in his 96 playoff appearances. McGraw (15), Fingers (13), and Gossage (10) are the only others with at least 10 playoff outings of two-plus innings. Andrew Miller’s eight are already tied for sixth most with Dick Tidrow.

We won’t see this role brought from the playoffs back into regular-season play overnight, but there is a place for it in today’s game. While strict adherence to the save has waned some, it is still a driving force behind the bullpen management of most teams, so a lot of teams’ ace relievers won’t necessarily be picked for this role. However, many guys who fall short of the five-six role of a starter show they have more than just an inning in the tank. Common shortcomings include the aforementioned TTO penalty as well as no third pitch, a sharp platoon split, and/or sustained injury issues.

The latter can never be fully eliminated. Guys who get hurt starting still can get hurt in relief, especially if they are overworked. The key isn’t to extract more outings from these potential Peacekeepers, but rather more innings per outing. Some of the more recent reliever workhorses were run into the ground by appearing in half or more of their team’s games.

Of course, some guys just have a limited number of bullets, whether they throw a 90-inning season one year or a pair of 45-inning seasons across two years. That doesn’t justify how Proctor was handled, but I want to stress that normal burnout and injury risks still exist, even with well-curated Peacekeeper roles.

Potential benefits include being able to shift a reliever roster spot back into a bench player, which would be especially useful in the National League (since it insists on allowing pitchers to bat for some ungodly reason). It wouldn’t hurt in the AL, either. The game is seeing teams more open to platoon roles again, so cultivating Peacekeepers would foster that strategy in both leagues. Of course, there’s also the natural benefit of shifting some innings from your worst relievers to your best. A team needn’t have just one PK if the assets are there.

Let’s be honest, this whole article is an elaborate ruse encouraging a team to recreate The Nasty Boys of the early’90s Reds. Norm Charlton actually had more starter innings than relief ones during the 1990 World Series season, but he added 50 relief innings to the 185 from Rob Dibble (98) and Randy Myers (87). Houston, I’m looking at you to get this done! I’ve actually gone through every team to find its  Peacekeeper candidate(s).

AL East

Tanner Scott, Baltimore Orioles: The 23-year-old lefty got an ever-so-brief look in September (two appearances) before heading to the AFL, where his spotty control wreaked havoc. The O’s may already be grooming Scott for the role, as he made 24 starts, but went more than three innings just twice. The easy gas (97-100 mph fastball) aids his 29 percent K rate in the minors, but a lack of command has his walk rate pushing 17 percent.

Joe Kelly, Boston Red Sox: It seems like he should already be filling this role, but he can’t seem to turn his impressive raw stuff into consistent results outside two BABIP-fueled sub-3.00 ERAs in 2013 and 2017.

Chad Green and Adam Warren, New York Yankees: They may be planning to put Green into the role, as he’ll enter spring training as a starter, and if he can’t consistently make it five-plus innings, then he’ll likely be back the multi-inning relief role that saw him log 69 innings just 40 appearances last year. The Yankees are equipped to run double or even triple Peacekeepers (Dellin Betances?), but their ridiculously deep pen means they don’t have to tax guys like Green and Warren.

Nathan Eovaldi, Tampa Bay Rays: He’s returning from Tommy John surgery this year, so this could be a way to limit his innings while still leaving a decent total to build on in 2019 should he return to fulltime starting. He’s had platoon and TTO issues throughout his career, making him an ideal fit for a swingman type of Peacekeeper, spot starting him against lefty-light lineups and giving him multi-inning relief roles otherwise to maximize his 65-80 pitch potential.

Joe Biagini, Toronto Blue Jays: A volatile run in the rotation–five starts of under four innings and five-plus runs; five quality starts, four of which were seven-plus innings–tanked his ERA, but the skills were the same. He seems well equipped to consistently make it through lineups one time.

AL Central

Carson Fulmer, Chicago White Sox: I don’t see him being able to consistently command his stuff for a starter’s workload, but it’s also too deep and good an arsenal to waste on getting just three outs at a time.

Andrew Miller, Cleveland Indians: The prototype right now. He logged a career-high 18 appearances of more than an inning, and it would’ve been higher if not for a knee injury that ate a month and a half of his season.

Shane Greene and Buck Farmer, Detroit Tigers: Greene is ostensibly their closer, but are you really a closer if you close for one of the worst teams?! He showed some promise in the rotation before thoracic outlet syndrome (TOS) surgery, though he had less than an inning per outing his 2017 since to 24 of his 71 appearances were under an inning. Farmer has some compelling swing-and-miss stuff but gets hit around entirely too much. Maybe tightening up his arsenal for shorter, more explosive outings could unlock another level.

Nate Karns, Kansas City Royals: He’s returning from TOS surgery, so his innings will have to be managed, and he’s always had major TTO problems. He has more than 150 innings in just one pro season, too.

Tyler Duffey, Minnesota Twins: The Twins are essentially already using him that way, with 25 of his 56 appearances going beyond an inning. He added velocity and whiffs in the bullpen while maintaining an above-average 1.6 groundball-to-flyball ratio.

AL West

Joe Musgrove and Chris Devenski, Houston Astros: Musgrove has shown a sharp drop-off around 50 pitches, but that’s enough for four innings when he’s on. He was great out of the pen in 2017, with 31 innings of 1.44 ERA and 0.86 WHIP in 23 outings. Devenski threw essentially two innings per appearance in 43 relief outings in 2016 (84 IP), but he seemed to burn out for spells with a heavier outings workload. The added depth in their bullpen could allow the Astros to go back to the 2016 usage with Devo.

Nick Tropeano, Los Angeles Angels: Some prospect analysts tabbed him for the bullpen as he was coming up with Houston, and a 2016 Tommy John surgery has fueled that idea even further. His change-up has always been a challenge, though his 48-point platoon split isn’t egregious.

Yusmeiro Petit, Oakland A’s: He’s the unheralded long-relief stud, with 229 innings of 116 ERA+ in 138 outings (three starts) over the last three seasons.

Erasmo Ramirez, Seattle Mariners: Ramirez has already dabbled with the role (2016: 91 IP in 64 appearances) and could be a ready-made option for the M’s…unless he’s needed in the rotation, which seems likely barring some more moves.

Mike Minor, Texas Rangers: I know they just signed him to transition back to starting, but I’m dubious about that plan. He was so good in relief (and did get 20 one-plus IP appearances) last year and has only  one really successful starting season back 2013.

NL East

Mike Foltynewicz, Atlanta Braves: This role is tailor made for Folty. Power stuff that you dream of using for 180 strong innings, but he lacks polish, the ability to thwart lefties, and stamina. Just 60 innings would be a waste, but 150 feels like a stretch.

Jose Urena, Adam Conley, and Jarlin Garcia, Miami Marlins: Urena is another poster boy for the role. He rode a .249 BABIP to some success in 2017, but it masked the arsenal and stamina issues that have him better suited for three or four innings at a time. Conley and Garcia are lefties with some promise, though likely not enough for 165 innings.

Rafael Montero ands Robert Gsellman, New York Mets: Montero was a control pitcher coming up, only to struggle mightily with finding the zone in the majors. He seems fit for a 50-pitch role that should allow him to see a lineup once. Even in his strong cup of coffee to end 2016, Gsellman had trouble the third time through a line-up. Meanwhile, nothing really worked in 2017, including a 240-point jump in OPS the third time through.

Vince Velasquez, Philadelphia Phillies: Velasquez is another guy who leaps to mind any time I start talking about this role. His flashes of success are memorable, but health has been an issue throughout his career, and it’s time to consider the move into super reliever. If he excels and seems to be primed for longer outings, they can take the Kris Medlen 2012 route and shift him back into the rotation later in the season.

A.J. Cole, Washington Nationals: He’s proven all he can at Triple-A, and the rotation seems full, at least to start the season. Getting just 60 innings would be a waste, as Cole has proven himself capable of logging at least 100 innings.

NL Central

Mike Montgomery, Chicago Cubs: Logged 21 starts over the last two seasons, but he’s kind of already in the role with 126 innings in his 72 relief appearances. That said, he doesn’t want to be, as he’s really looking for a full-time starting role.

Raisel Iglesias, Michael Lorenzen, and Amir Garrett, Cincinnati Reds: Maybe the team with the first iteration of The Nasty Boys is primed to come out with the 2.0 version? They already semi do it with Iglesias, even in the closer’s role. Lorenzen has done his best work in his 141 innings of relief (over 111 outings). Garrett was great or awful with very little in between, tallying 14 starts in 16 appearances. They’ll likely give him another real shot at starting.

Brent Suter, Milwaukee Brewers: Suter had the biggest change in performance the third time through when compared to his first two times, with a 497-point OPS shift. Batters tattooed him for a 1.127 OPS the third time through with a zero percent K-BB rate.

Tyler Glasnow, Pittsburgh Pirates: He may already be starting the season in the bullpen, and I’ve long been dubious of his ability to hold up as a starter, so this role seems like a worthy middle ground.

Sam Tuivailala, St. Louis Cardinals: I always felt Trevor Rosenthal was more than a one-inning guy. That was years ago, though. He’s now recovering from TJ surgery and not even a Cardinal any more. Tuivailala hasn’t figured out how to get whiffs from of his quality stuff yet, but he sharply cut his walk rate to a career-best six percent and got some multi-inning looks.

NL West

Randall Delgado, Arizona Diamondbacks: No platoon split, solid stuff, and plenty of multi-inning relief outings. He hasn’t reached 80 innings as a primary reliever yet, though taking away some outings in exchange for extending some others definitely could get him there.

Chris Rusin, Colorado Rockies: Rusin quietly has had back-to-back strong seasons essentially already in this role. He got fewer than four outs just six times in 2016 (84 IP in 29 appearances) and then basically went half-and-half in 2017 en route to 85 innings in 60 appearances. Rusin, a lefty, hasn’t shown a platoon split, either, allowing him to make it through most lineups at least once. Antonio Senzatela definitely had moments in the rotation, but in the end might be best served to log 90 relief innings and give the Rockies a right-handed counterpart to Rusin.

Ross Stripling and Brock Stewart, Los Angeles Dodgers: Let’s be honest, the Dodgers have a million options for this role. They basically have Rich Hill and Kenta Maeda in the role already for the early innings, so these two would give them someone to pass the baton to for a couple innings before Pedro Baez comes in! Just kidding…I just wanted to trigger Dodgers fans. But seriously, Hill and Maeda leading to Stripling and Stewart leading into Kenley Jansen is an effective way to keep the bullpen from being taxed since they only have one starterwho will regularly go six-plus innings, Clayton Kershaw.

Dinelson Lamet, San Diego Padres: Just two pitches, a massive platoon split, and a huge fall-off when facing the lineup a third time make him a prime candidate for this role if he can’t improve on these issues. Just because the idea is a two-to-four-inning role doesn’t mean these pitchers never can go beyond that. If the Padres have some righty-heavy lineups on the schedule, they can slot him into the rotation. He had 10 quality starts in 21 tries. He also allowed five-plus earned runs in five starts. The swingman iteration of this role is probably the best fit for Lamet. The Padres can get 110-130 strong innings from Lamet if properly deployed instead of just blindly starting him.

Chris Stratton and Ty Blach, San Francisco Giants: A righty/lefty duo that definitely deserves more than an inning per outing but don’t have the tools to consistently log five innings, let alone going beyond that.

…

Some teams obviously are better positioned to deploy a Peacekeeper than others, though I’m sure every organization could identify a guy or two in its system who could pull this off. I was focused primarily on pitchers who have logged major league time. Doug Thorburn has been railing on inefficient pitcher roles for years while also understanding it won’t change overnight. Limiting guys to six- or one-inning roles leaves a lot of quality inning unused and probably pushes some guys beyond their limits, yielding poor innings that would be better off in the hands of another pitcher.

The playoffs are showing how well this can work in a controlled environment and tight time frame, and there is a path to replicating it across six months. After all, it’s nothing new to baseball. Closers were birthed from an expanded relief role. If Hollywood can constantly recycle the same handful of ideas, baseball teams can bring in this idea and better maximize their resources.

References & Resources

• Baseball-Reference
• FanGraphs
• Tyler Kepner, The New York Times, “Next! Dodgers and Astros Are Quick to Pull Pitchers”
• Bill Brink, Pittsburgh Post-Gazette, “What will the Pirates do with Tyler Glasnow?”