Craig Kimbrel was supposed to be on the decline. Starting in 2013, the man who was once the best reliever in baseball watched his ERA rise in four straight seasons. Kimbrel entered the majors as a revolutionary high-strikeout reliever, but over the course of seven seasons, he’d started to slip while the rest of the league caught up. In 2016, he dropped to sixth in K%, a 14th-place tie in saves, and all the way to a tie for 68th out of 135 qualified relievers in ERA-. Even great relievers tend to have short shelf lives, and Kimbrel, who looked like he was entering his age-29 season on the decline, was just the latest relief ace to lose his edge.
Then came 2017. Through 30 appearances, Kimbrel’s ERA is a career-low 0.85 ERA, and it’s not some fluke: Only 15 of the 111 batters he’s faced have reached base through a hit, walk, or hit-by-pitch. That makes for a FIP of 0.38. He’s allowed only three earned runs, and in those 30 appearances, Kimbrel has saved 20 games and increased his team’s win probability 29 times. In a ranking of guys announcing that they’re back, Kimbrel’s 2017 season falls somewhere between Randy Quaid in Independence Day and Keanu Reeves in John Wick.
Kimbrel weathered the tide of evolution, and has jumped back to the front of the line.
The modern relief ace evolved slowly over the past century or so. After World War II, the Yankees’ Joe Page, among others, started to come in to get tough outs in important spots, rather than just spell a tired arm. In 1950, a relief pitcher, Jim Konstanty, won the National League MVP award. After the second dead-ball era in the 1960s, a new set of multi-inning firemen emerged and the expectation that a starting pitcher ought to finish the game began to fade for good. In 1969, MLB started tracking saves. In 1974, the Dodgers’ Mike Marshall appeared in a record 106 games, averaging almost two innings per appearance en route to the first Cy Young win for a reliever since the award was instituted in 1956.
As heavy relief workloads started ending careers, the relief ace’s role was scaled back to fit the save statistic. Sparky Lyle was never the same after his 137-inning Cy Young campaign in 1977 and Marshall struggled for years after 1974 before recovering to set the American League appearances record with Minnesota in 1979. In 1990, Oakland’s Dennis Eckersley posted a 0.61 ERA, throwing only 73.1 innings in 63 appearances, and Chicago’s Bobby Thigpen set the single-season record with 57 saves. The modern closer was born.
Though many of the great closers of the 1990s relied on some degree of finesse, whether it was Mariano Rivera’s command or Trevor Hoffman’s changeup, teams quickly realized that a very specific type of pitcher could thrive in a clearly defined, high-pressure role if they had a hard fastball and a hard breaking ball. Eckersley’s rise under Tony La Russa was like breaking the sound barrier, and relief pitchers evolved the way jet- and rocket-powered airplanes did over the next 20 years: faster, louder, and prone to the occasional catastrophic failure. (If you really wanted to chase this metaphor, you could point out the visual similarities between Kimbrel’s unique arms-out lean on the mound and the drooping wingtips of the Mach 3 XB-70 bomber prototype.)
Since closers almost always started the ninth inning with nobody on base, the occasional walk wouldn’t hurt that much as long as they limited hits, particularly extra-base hits. From 1990 to 2000, a pitcher saved 30 games in a season with a BB/9 over 4.00 33 times. From 1988 to 1993, Mitch Williams averaged 29 saves a year for three teams while walking 6.4 batters per nine innings. Closers of the 1990s played with fire to a certain extent: Not only did Williams fall off a cliff, but while the Atlanta Braves won division title after division title in the 1990s thanks to Greg Maddux, Tom Glavine, and John Smoltz, they watched not one but two superstar closers — Mark Wohlers and John Rocker — suddenly lose the ability to throw strikes. Every baseball fan who lived through the 1990s can remember a pitcher like Williams or Rocker, able to command his fastball-slider combination just enough to get hitters out, only to suddenly lose everything as that command slipped.
But along with these failures came breakthroughs. In 1992, Cincinnati’s Rob Dibble became the first reliever with a K/9 ratio above 14 in a season with at least 40 appearances. From 1997 to 1999, Houston’s Billy Wagner did it three times in a row, and suddenly, the standard wasn’t just “throw hard”; it was “literally throw 100 miles an hour.” Armando Benítez, Byung-Hyun Kim, Eric Gagne, Brad Lidge, and Carlos Mármol all topped 14 K/9 by 2010, though all of them either flamed out quickly or flirted with disaster the whole way.
In 2011, Kimbrel topped 14 K/9 in his first full major league season, when he took over from Wagner as the Braves’ closer. The 23-year-old Alabamian immediately became the apotheosis of the one-inning power closer.
Unlike most relief pitchers, Kimbrel is not a failed starter. From the moment the Braves drafted Kimbrel out of an Alabama junior college in 2008, he was an elite closer. Of Kimbrel’s 575 professional pitching appearances, only one — a three-batter rehab outing at Triple-A in 2016 — was a start.
Kimbrel struck out 56 batters in 35.1 innings across three levels in 2008. He struck out 103 batters in 60 innings across four levels in 2009, and by the end of 2010, he’d struck out 40 batters in 20.2 innings with the Braves, en route to a 4–0 record and a 0.44 ERA.
There are differences between Kimbrel and Wagner, namely throwing hand and breaking pitch (Kimbrel throws a knuckle curve, while Wagner used a slider), but the effect is much the same. The game plan is also the same: Establish a hard fastball and work off it to a hard, diving breaking pitch. A hitter can sit on one or the other, but the two pitches are so different in speed, location, and movement that it’s next to impossible to look for one and adjust to hit the other. Over his career, Kimbrel has thrown about three first-pitch fastballs for every first-pitch curve, and that ratio narrows to a little less than 2-to-1 with two strikes. Wagner was a 4-to-1 first-pitch fastball guy, and with two strikes his fastball-slider ratio was almost even.
PITCHf/x didn’t exist for Wagner’s best years, but in 2010, Wagner posted a 13.5 K/9 and a 1.43 ERA, so he was still an outstanding closer. That year, Wagner’s fastball averaged 96.3 mph and topped out at 100.3, while his slider averaged 83.4 mph with 4.7 inches of vertical break and 2.5 inches of horizontal break. In 2011, Kimbrel’s fastball averaged 96.9 miles an hour, topping out at 100.7, and his 87.4 mph curve broke down 5 inches and over 5.9 inches.
If you’re a closers-belong-in-the–Hall of Fame type of person: Wagner is a Hall of Fame–caliber pitcher, and Kimbrel, a 23-year-old juco guy with two and a half years of pro experience at any level, took over for him without missing a beat or breaking a sweat.
From that point on, Kimbrel was the best reliever in the National League. In four full seasons with Atlanta, Kimbrel made four All-Star teams, led the National League in saves every year, and (counting a ninth-place tie with John Axford in 2011) led all NL relievers in Cy Young voting each year, including two top-five finishes. From 2011 to 2014, Kimbrel’s 252 ERA+ led all MLB relievers with at least 200 appearances. He and Aroldis Chapman were the only relievers to strike out 40 percent of the batters they faced over that stretch.
Then, in April 2015, Kimbrel was traded to San Diego and his numbers started to slip. Even in pitcher-friendly Petco Park, Kimbrel posted a career-high ERA and career lows in saves, strikeouts, and innings pitched. Traded again the following offseason to Boston, Kimbrel posted a still higher ERA, threw still fewer innings, saved fewer games, and struck out fewer batters.
The league was catching up. Kimbrel broke the all-time K/9 record in 2012, but Chapman, who debuted the same year as Kimbrel, broke it in 2014. By Kimbrel’s debut year in 2010, there had been only nine 14 K/9 seasons in the 100-plus-year history of Major League Baseball. Since Kimbrel left Atlanta in 2015, there have been eight. In 2017, seven different relievers are on pace to strike out 15 batters per nine innings, including guys nobody had ever heard of before this year, like Tommy Kahnle and James Hoyt.
The expectations of the role are changing as well. Nowadays, when even Miami’s Kyle Barraclough can record half his outs by strikeout irrespective of command or control, it takes stuff and command to stand out. Another fastball-slider guy, Andrew Miller, blew past 14 K/9 last year not only while posting a walk rate of just 3.3 percent, but while pitching as a Mike Marshall–type fireman, sometimes for multiple innings at a time. In 2017, Kenley Jansen took Miller’s “14.9 K/9 while figuratively not walking anyone” and raised it to “15.2 K/9 while literally not walking anyone.”
And yet Kimbrel, after two years of looking quite ordinary, is once again pitching as well as any reliever in baseball. He’s throwing his fastball harder (98.7 mph on average), with more arm-side movement (7.1 inches) than ever before, and is therefore generating whiffs on 41 percent of opponents’ swings on his fastball. That’s a career high, and apart from 2012, when he posted a 16.7 K/9 ratio and a 1.01 ERA, he’s never even come close. As you might expect, Kimbrel has also cut his walk rate by almost two-thirds since 2016, from 13.6 percent to 4.5 percent. To put that into context, his walk rate went from 133rd out of 135 qualified relievers in 2016 to 13th out of 170 in 2017.
Kimbrel is back to his old self, but beyond that, he’s innovating. He’s throwing more pitches in the strike zone than ever — 52.7 percent — and pitching backward. While for his career Kimbrel has started batters off with a fastball 75 percent of the time, this year he’s throwing first-pitch fastballs only 60 percent of the time. And that 2-to-1 fastball-curve ratio with two strikes is now almost 3-to-1. It’s hard enough to hit either of Kimbrel’s pitches when they’re coming in a predictable pattern; imagine not even being able to guess.
Kimbrel’s 2012 season was one of the best ever for a one-inning closer, and based on rate stats, he’s just as good now as he was then, maybe a little better. The Red Sox have World Series ambitions pretty much every year, and after acquiring Kimbrel from the Padres for four players, including two top-60 global prospects, those intentions — as well as Kimbrel’s role in a hypothetical World Series run — have never been clearer.
Despite playing almost his entire career for very good teams, Kimbrel has never had a signature postseason moment. In 2010, he featured in all four games of Atlanta’s NLDS loss to the Giants, and in 2011, he blew a save to the Phillies on the final day of the season, knocking the Braves out of the playoffs as part of the most dramatic night in recent regular-season history. In 2012, Kimbrel threw a scoreless inning in the wild-card game, but only after the infield fly heard round the world turned the game against Atlanta.
In 2013, Kimbrel waited in the bullpen, arms folded and apparently trying to shoot lasers out of his eyes at manager Fredi González, while David Carpenter blew an eighth-inning lead in an NLDS elimination game against the Dodgers.
In 2016, Kimbrel made his first playoff appearance post-memeification, but the Red Sox never got close enough to Miller’s Cleveland squad during the three-game sweep for a closer to make a difference.
Last year’s postseason knocked Boston back a few steps, but it was a body blow for the concept of a one-inning closer in the playoffs. Not only did Miller reinvent the fireman role, Jansen delivered a pitch-until-you-drop outing in Game 5 of the NLDS, throwing 51 pitches in 2.1 innings. Chapman, who didn’t want to pitch in the Miller role, nevertheless got at least four outs in five of his last six appearances in last year’s playoffs, including a 42-pitch eight-out save in Game 5 of the World Series. Could Kimbrel do that?
To a certain extent, that’s between Kimbrel and manager John Farrell. While some relief pitchers, like Miller, are up for anything, others — even analytically inclined ones like Minnesota’s Glen Perkins — would rather have a more defined role. Kimbrel’s 2013 postseason stinkeye certainly indicated a willingness to pitch outside the ninth inning, but how far outside the ninth inning remains to be seen. Even still, his role is already changing.
Kimbrel has 19 career appearances in which he recorded more than three outs. Seven of them came in his four-plus years with the Braves, while 10 have come in his year and a half with the Red Sox. Kimbrel is on pace for his first season ever with more innings pitched than appearances, thanks in part to a one-week span, from May 30 to June 6, in which he recorded three four-out saves.
Farrell expressed a reluctance at using Kimbrel for more than an inning at a time, while Kimbrel said the downtime between innings took some getting used to. But assuming he gets used to that, there’s no reason Kimbrel couldn’t get four- or even five- or six-out saves with regularity. Each of Kimbrel’s three most recent four-out appearances came on at least two days’ rest, and innings are kind of a crude measure for a pitcher’s workload anyway. An inning can mean as few as three pitches or as many as 30 or more.
It’s important to remember that because so few of Kimbrel’s opponents reach base, and because he’s throwing more strikes than ever, his innings don’t last as long as a normal pitcher’s. None of Kimbrel’s 30 outings this year has stretched beyond 33 pitches or seven batters faced, and he’s hit 30 pitches only three times. By comparison, last August 9, Kimbrel threw 37 pitches in 0.2 innings, an outing that would’ve been his longest of 2017 in terms of pitches, but tied for his second-shortest in terms of innings. So far this year, big league pitchers have thrown 30 pitches or more in one inning or less 133 times.
For comparison, Kimbrel has thrown an average of 15.6 pitches per inning this year, while Milwaukee’s Corey Knebel has thrown 17.3, which isn’t very much on an outing-to-outing basis, but over the course of a season it means Kimbrel can throw about 11 percent more innings than Knebel without throwing more pitches. On a typical closer’s workload, that’s an extra six or seven appearances — or, if you want to look at it this that way: another 20 or so four-out saves.
Even if the occasional four-out save is the limit of Kimbrel’s flexibility, he’s still throwing more innings, and in higher-leverage situations, than he ever did with Atlanta. Kimbrel, the perfect culmination of the last phase of relief ace, has finally adapted, and now looks, once more, like the future.
