Could Tim Lincecum be the next Willie Hernandez? (via Dirk Hansen)
This has been a postseason dominated by relief pitching. In the National League, Kenley Jansen worked multiple-inning magic to get the Dodgers past the Nationals and was dominant when given a lead in the NLCS against the Cubs. Aroldis Chapman also made some rare outings of more than three outs in anchoring the Cubs bullpen. And of course, Andrew Miller was the ultimate fireman in first two rounds, reeling off 11.2 scoreless innings in six appearances out of Cleveland’s bullpen, squelching rallies and eating innings whenever a tight spot arose.
Miller, in particular, has captured attention this postseason, with his performance inspiring many to open their minds beyond the conventional wisdom of bullpen usage, with its defined roles and single-inning usage patterns. Amazingly, Miller needed just six games this postseason to rack up a plus-0.83 Win Probability Added. Surely the extra off days in the postseason schedule have assisted in allowing Miller to make so many long appearances, but this performance makes one wonder why such a strategy doesn’t get more run in the regular season. After all, at this rate, the postseason version Miller is adding about a win per seven appearances; it would take just 46 such appearances to match Zach Britton‘s 2016 league-leading 6.1 WPA out of the bullpen.
Such a reliever, over the course of a full season, could be one of the most valuable players in the league. In 1984, Tigers relief ace Willie Hernandez proved it, serving as a multi-inning fireman-closer hybrid to win the American League Most Valuable Player Award. He appeared in 80 games, recorded 32 saves, and finished with a 1.92 ERA over 140.1 innings. Hernandez may not have quite handled a starter’s workload, but considering the importance of his innings — he had a 1.42 average leverage index over those 80 appearances — Hernandez was every bit as valuable as an ace starter. He finished with a plus-8.58 WPA in 1984. In the FanGraphs WPA era, dating from 1974 to the present, Hernandez’s 1984 season tops all relievers and is second only to Dwight Gooden‘s 1985 season (9.46 WPA) among all pitchers.
Win Probability Added is a better statistic for storytelling than for prediction for a lot of reasons, primarily the fact that WPA is so context-dependent. Starter WPA is dependent on how much run support the lineup can provide. Hitter WPA depends as much on how many opportunities come with runners on base as it does ability to get hits. And reliever WPA is the best example of all — Miller doesn’t compile his WPA this postseason without his Cleveland teammates giving him a lead through the game’s early innings in six games against some of the best teams in baseball.
The best reliever season of all-time by FanGraphs WAR, for instance, is Bruce Sutter‘s 1977, in which he held opponents to a 1.34 ERA and 1.61 FIP over 107.1 innings. But that Cubs team finished at an even .500, in stark contract to the 104-win 1984 Tigers — with fewer leads and big situations to protect, Sutter’s WPA finished at 5.18, more than three wins below Hernandez’s record setting mark.
Hernandez’s WPA says, though, that a talented reliever capable of going multiple innings can turn a good team into a great team, and a great team into a 100-win World Series champion, as the 1984 Tigers were. Hernandez won the MVP for his role on that team, and while there were other worthy candidates — Cal Ripken hit .304/.374/.510 and played a brilliant shortstop for the Orioles; Alan Trammell hit .314/.382/.468 and stole 19 bases for the Tigers; Lloyd Moseby played a marvelous center field for the Blue Jays and supplied a .280/.368/.470 batting line. But I find it hard to argue with anybody who put Hernandez above all those players on their MVP ballots. Between his workload and the importance of his innings, he clearly produced immense value.
Here are just a few ways to describe just how far above and beyond the typical closer duty Hernandez went: He pitched from the seventh inning (or earlier) to the games’ end on 10 different occasions, losing just once. He threw as many as four shutout innings in a single outing, including more than three shutout innings eight times. He struck out five batters or more in four different appearances. He allowed just six home runs all season and never more than one in an outing. His first 11 saves were all of at least two innings.
His only blown save came on a sacrifice fly which scored an inherited runner in a meaningless late-September contest against the Yankees. He entered as early as the fifth inning and pitched as late as the 12th. Hernandez accomplished things by himself that would take three or four relievers in today’s game, and did them with the effectiveness of today’s very best closers.
Still, the model didn’t catch on. Per the Baseball-Reference Play Index, Mark Eichorn is the only other reliever to throw as many innings as Hernandez since then. Only 12 have even thrown more than 120, and none later than 1990. The only full-time relievers to go more than 100 innings in the new millennium have been middle relievers or mop-up men — Scott Proctor, Scot Shields, Steve Sparks, Guillermo Mota and Scott Sullivan, none of whom would ever be confused for Willie Hernandez reborn.
Hernandez’s arsenal today also differs heavily from the typical closer mold of today, with its 100 mph fastballs and hard-biting sliders. Instead, Hernandez lived off his famous screwball, taught to him by Cy Young winner Mike Cuellar in the Puerto Rican winter league, and a cut fastball that gave him the ability to pitch inside despite lacking the top end velocity we expect nowadays. Hernandez also flashed a curveball against left-handed pitching, and having such a deep arsenal of off-speed options helped him be so effective in multiple-inning appearances. Even though he was never quite able to repeat his 1984 in effectiveness or workload, he made the All-Star game in 1985 and 1986 behind his continued success pitching more than an inning per appearance.
But perhaps the biggest reason why the Tigers were able to use Hernandez is such a unique way was the presence of another top level reliever on their roster in Aurelio Lopez. Lopez was an All-Star in 1983, a season in which he appeared in 57 games and went 9-8 with 18 saves and a 2.81 ERA over 115.1 innings pitched. When Hernandez needed rest, the Tigers were able to plug Lopez into that same role, even if his 2.94 ERA in 1984 paled in comparison to Hernandez’s 1.92 mark. Having two relievers of such talent is a luxury few teams have had before or since.
This, too, is an underrated aspect of Miller’s success this postseason. Would Terry Francona have been so brave with Miller in the early innings if Cody Allen, a stud closer in his own right, hadn’t been waiting in the wings? The fact that Cleveland had somebody in the more traditional closer’s role opened the door for Francona to experiment with the rest of his bullpen usage. Credit to Francona for realizing this, though — when the Yankees had Miller in the same bullpen with Dellin Betances and Aroldis Chapman, the trio was largely locked into their own innings, with Betances taking charge of the seventh, Miller the eighth and Chapman the ninth, even on days when pivotal moments appeared in the fifth or sixth.
Not every team will have the relief talent on hand to take advantage of such a strategy. But it is worth noting that many pennant winners in recent memory have had at least two pitchers capable of holding down the closer’s role. Cleveland’s opponent in the World Series, the Cubs, have Hector Rondon, who saved 30 games with a 1.67 ERA in 2015, to go with Aroldis Chapman. The Royals of 2014 and 2015 had the likes of Greg Holland, Kelvin Herrera and Luke Hochevar backing up Wade Davis. The even-year-magic Giants had the pairing of Sergio Romo and Santiago Casilla, as well as the ace-in-the-hole of Tim Lincecum in 2012 and 2014.
Lincecum, in fact, is probably the most interesting example of the kind of player who could thrive in such a role. As incredible as Lincecum was as a starter earlier in his career, it was clear by 2012 that his body was struggling to hold up under a starter’s workload. Batters hit a ludicrous .316/.389/.605 against Lincecum in the sixth inning in 2012. But when he came out of the bullpen for the Giants that postseason, he was dominant in five of his six outings, recording four shutout appearances of at least two innings and earning the win in Game Four of the NLDS against the Reds with 4.1 innings, six strikeouts and just one run allowed as he shut Cincinnati down from the fourth inning through the eighth.
As enticing as it might be to have a pitcher like this to put out fires throughout the regular season, the reason it doesn’t happen is fairly simple: Any pitcher good enough to excel in such a difficult role is already either closing or starting.
The Giants continued to try Lincecum in the rotation even after his arm started to give out despite his postseason success in the bullpen. Miller began his career as a highly hyped starter and was given plenty of chances to fail there before he was transitioned into the bullpen. Even Wade Davis, who had success as a multiple-inning reliever with the Rays, was first tried as a starter after he was dealt to Kansas City and then pigeonholed into a one-inning relief role (a luxury the Royals could afford with their plethora of amazing relievers).
Money surely plays a role as well. What player — and perhaps as importantly, what agent — is going to want to give up a shot at becoming a major league starter, where even also-rans like Ricky Nolasco and Ervin Santana can make $60 million or more on the free agent market? The highest-paid closer in baseball right now is David Robertson, who signed a four-year, $46 million deal with the White Sox in 2015. Craig Kimbrel ($42 million over four years) and Ryan Madson ($22 million over three years) are the only other relievers on contracts totaling over $20 million right now.
That’s the major rub in actually implementing this strategy that we’ve seen can be so effective. Even if you have a manager and front office that’s progressive enough, and a deep enough bullpen, and a strong enough rotation, you also have to find an unusual pitcher with the stuff to handle high-leverage situations and the endurance to effectively pitch multiple inning, and who isn’t stubborn enough to resist placement in this new, awkward role. So while it’s tempting to blame those who run teams for being too stuck in their own ways to implement a new and arguably more optimal bullpen strategy, we need to acknowledge the roadblocks.
Aut at the same time, I have no doubt that the current way we have of sorting pitchers as either full-game starters or one-inning relievers is not the way to get the most out of everybody’s unique talents. Many many pitchers fail as starters not because of their problems in the first three innings, but because their stuff drops off and they get punished the second time through the order. Others excel as relievers because their stuff is significantly better when they can throw at 100 percent without worrying about pitch counts.
Josh Kalk found for The Hardball Times in 2008 that relievers generally don’t experience a dropoff in stuff in their second inning as long as they aren’t forced to throw too many pitches in their first frame. If a young pitcher is struggling with endurance but dominating in shorter outings, perhaps this is the role he should be groomed for.
Unfortunately, I doubt we’ll be seeing too many pitchers developed particularly for the role Andrew Miller has occupied this postseason. There are so many obstacles in the way, and it would require almost a full overhaul of the way minor league pitcher development is set up, not to mention a re-imagining of the relief pitcher market — somebody who showed he could play this fireman role for a full season should be worth more than even what David Robertson earned as a free agent. But given the endless shortage of starting pitching faced by major league teams, there’s no reason to expect teams to stop trying everything they can to get their young prospects to succeed in the rotation.
Still, as Miller himself has said, his performance this postseason is proving the value of an elite reliever outside the ninth inning. Sometimes a game hinges on eliminating a rally in the sixth or seventh inning, or shutting down the heart of the order when they come up all in a row in the eighth. Such a relief ace may never be common just because the talents necessary to pitch as well as Miller has this postseason or Hernandez did for the entire 1984 campaign are simply too rare.
But it’s hard to watch the likes of Miller and Jansen dominate over multiple innings and wonder what could be if such a strategy was used over 162 games. In a world where every team is trying to discover its own competitive advantage, perhaps some team out there will discover a way to make a Hernandez-esque relief ace its next big thing.
References & Resources
Josh Kalk, The Hardball Times, “Do relievers struggle if they throw multiple innings?”