Eighth-inning relievers save the day while closers get the saves
In each of Mike Adams' final three eighth-inning appearances with the Padres this season, he entered to face the heart of the opponent's order. Against the Marlins those Nos. 3-4-5 hitters were Logan Morrison, Hanley Ramirez and Gaby Sanchez. Against the Phillies Adams pitched to Chase Utley, Ryan Howard and Shane Victorino. And against the Diamondbacks he opposed Justin Upton, Miguel Montero and Chris Young. Eight of the nine made the NL All-Star team either this year or last.
All three times Adams retired the side without allowing any more than one baserunner, clearing each opponent's most dangerous sluggers before closer Heath Bell entered to face either the 6-7-8 or 7-8-9 slots of the lineup, a series of far less illustrious hitters that included Dwayne Wise, Bryan Peterson, Brett Hayes, Domonic Brown, Carlos Ruiz, Ross Gload, Willie Bloomquist, Ryan Roberts and Sean Burroughs. None has ever been selected to an All-Star team.
For his efforts Adams received three holds, an arcane statistic not found on the backs of baseball cards, while Bell received three saves, the relievers' sole counting number tracked by fans -- yet who had the more difficult job?
It wasn't a fluke of the National League, either. After being traded to the American League's Rangers, Adams' first two appearances -- again, both in the eighth inning -- began by facing the Tigers' 1-2-3 and 2-3-4 hitters.
"You notice that you go through stretches when you face a certain part of the lineup, but it makes no difference what part of the lineup you face," says Adams. "It's all kind of coincidental what part of the lineup I get."
Actually, there's a clear pattern at play. Adams has thrown 90 percent of his innings the past two seasons in the eighth, and is nearly 70 percent more likely to face an opponents' 2-3-4 hitters than its 6-7-8.
Nor is this just bad luck for Adams. Instead, it's a baseball-wide trend.
To be sure, the set-up man is nothing new, but his role is evolving and being valued in the game as never before. Last winter, Rafael Soriano, Joaquin Benoit, Grant Balfour and Jesse Crain were among those who signed lucrative multi-year contracts even though they neither start nor close. Teams now face a longer gap between starter and closer -- a gap that has increased one inning per game since 1988 -- thus increasing the importance of the set-up man.
But there's another oft-overlooked mathematical reason of average baserunners and batters faced explaining why the eighth, in particular, can be such an important inning.
That will be on display next month when every pitch and pitching change are placed under the postseason microscope. Each of the AL's playoff teams (including both the Rays and Red Sox, who are tied for the wild card entering play Tuesday night) and three of the five teams still alive in the NL have deployed the same dedicated eighth-inning reliever in front of the closer all season.
"It's really become that the set-up guy is as important as the closer," said Diamondbacks general manager Kevin Towers, whose rebuilding of Arizona's bullpen was key to their worst-to-first turnaround. "The way lineups turn over, the eighth-inning guy probably sees more of the middle of the order than the ninth-inning guys."
While there are myriad variables in any given game that can affect how a lineup cycles and the matchups a reliever faces, the data reinforces that, in the average game when a team has a chance to win, there is a slightly better chance it will face an opponent's best hitters in the eighth inning than in the ninth.
This season's rate of 11.3 baserunners per game equates to about 1.26 per inning, meaning the leadoff hitter of the eighth inning is often either the No. 3 or 4 hitter with the bottom of the order to follow in the ninth.
"It happened a lot with Benny last year," Rays manager Joe Maddon said, referring to Benoit, last year's eighth-inning reliever who now sets up closer Jose Valverde for the AL Central champion Tigers. "It was incredible how often it worked out exactly that way."
A few relievers are noticing it, too. In July, the Nationals' Drew Storen realized that his All-Star eighth-inning set-up man, Tyler Clippard, was often getting the more difficult assignment.
"It started with me mentally preparing for who I was going to face," Storen said, "and I was like, 'Clip's got to face 3-4-5 again. That's pretty funny.'"
Said Clippard, "After it happens over and over again, I'm like, 'Wow, I'm facing the best hitters in the league. What the heck? Give me some pinch hitters every now and then.'"
Shortly before the bullpen phone rings in Boston, eighth-inning man Daniel Bard will often consult the opposing hitters' scouting report the relievers keep with them, though against division foes like the Yankees, he's faced a few of the hitters so many times that a last-minute refresher isn't necessary.
"I seem to get [Nick] Swisher a lot," Bard said. "I've probably faced him 10 times this year. I know I've gotten [Mark] Teixeira a lot over the course of my career and [Robinson] Cano quite a bit. That's just coincidence, though."
No it isn't. The hitter Bard's faced the most this season is Swisher (seven times), who often bats sixth, and the two batters he's seen the most in his three-year career are Teixeira (16 times), who has been almost exclusively a No. 3 hitter since 2009, and Swisher (14 times).
Nowhere is this trend more pronounced than in the work of Adams, who in the last two seasons with the Padres and Rangers, has thrown 125 2/3 of his 139 1/3 innings in the eighth, securing 70 holds while maintaining an outstanding 1.61 ERA and 0.93 WHIP. Those numbers are even more impressive when you consider he is about 68 percent more likely to face an opponents' 2-3-4 than their 6-7-8.
It's no wonder that the Rangers traded a pair of prospects for Adams, although they aren't the only playoff-bound team that went to fortify their bullpen at midseason. Three weeks before Adams was traded, the Brewers made a deal with the Mets to acquire single-season saves record-holder Francisco Rodriguez so he could set up closer John Axford.
While Rodriguez recently griped that his new role doesn't include save opportunities the truth is he's doing important work for Milwaukee right where he is.
Indeed, nearly every playoff-bound team or contender has a lockdown eighth-inning pitcher. Teams with excellent starting pitching -- the Braves and Phillies this year, for example -- are likely to face slightly different combinations because opposing lineups won't have had as many baserunners. Similarly, a great lineup like the Yankees' will typically have an extra runner or two per game.
See the below chart for a side-by-side comparison of a set-up men with their closers on baseball's best teams this year, with a listing of plate appearances for each batting-order position. The most frequently faced batters are bolded for a visual illustration of the trend. (The Cardinals and Phillies were excluded below because of a lack of consistency with their eighth-inning relief pitching over the course of the season, though Antonio Bastardo has since claimed that role in Philadelphia.)
The difference usually isn't extreme but for most teams it's identifiable.
For evidence beyond this season, consider the careers of closers Mariano Rivera and Trevor Hoffman, who pitched long enough to serve as sufficient sample sizes for his phenomenon. In his work as Yankees closer since 1997, Rivera has faced an opponents' Nos. 6-7-8 hitters most frequently -- 493, 503 and 516 plate appearances, respectively -- and an opponents' Nos. 2-3-4 hitters the fewest times: 396, 391 and 420 plate appearances, respectively.
Hoffman, who was the primary closer mostly for the Padres and also the Brewers from 1994 to 2009, had a similarly discernible spread during that time, even if it wasn't as severe. He most commonly faced an opponents' 6-7-8 hitters (448, 444 and 445 plate appearances) and least often faced Nos. 1-2-3 (401, 393, 388 plate appearances).
While the interpretation of such data might reinforce the notion of using a bullpen by committee -- the premise of throwing one's best reliever against the opponents' best hitters, no matter the inning -- managers such as Maddon and pitching coaches such as the Braves' Roger McDowell all emphasized that the stress of the ninth inning remains different enough that a single, dependable closer should still be relied upon for the final three outs.
The point here is not to denigrate the closer but, instead, celebrate that burgeoning niche of a second relief star who can pitch the eighth inning no matter what type of hitter is at the plate.
"The eighth-inning guy that can be more neutral against both lefties and righties is invaluable," Maddon said. "If you have an eighth-inning guy that's really much better against righties or much better against lefties, you're going to have some problems, especially if you're talking about the meaty part of the batting order."
Then there is the trickle-down effect. Teams with a dominant eighth-inning guy have the luxury of using their situational relievers in the sixth or seventh innings.
"You don't really see matchups in the eighth inning anymore," Orioles pitching coach Rick Adair said. "Those eighth-inning guys are irreplaceable when they have the ability to get lefthanded hitters and righthanded hitters out. That sixth inning becomes more critical. That's when you first see the matchups."
Texas assistant general manager Thad Levine said the club noticed that their eighth-inning reliever more often tangles with an opponents' big boppers, and it was particularly evident last year when veteran reliever Frank Francisco mired the thickets of 3-4-5 in the eighth inning more than any other lineup position (and often No. 6 if he allowed a baserunner), whereas then-rookie closer Neftali Feliz was most apt to face an opponents' 7-8-9.
This year the club lacked that eighth-inning presence until acquiring Adams at the trade deadline.
"It helps a lot, big time," Rangers pitching coach Mike Maddux said. "We used to use Mark Lowe and Darren Oliver to mix and match, but now with Michael being able to anchor down the eighth, we can use those guys in the seventh inning or the sixth inning, whatever the case may be."
There has been a slight increase in recent years in the likelihood that a team's top set-up man is charged with more difficult matchups than the closer. There are fewer baserunners on average in today's lower run-scoring environment than there were during the height of the Steroids Era a decade ago. As mentioned, this season each team has had an average of 11.3 net baserunners per game, calculated by taking all the ways a player reaches base (hit, walk, error, etc.) and subtracting all the ways he makes an out on the basepaths (caught stealing, double play, etc.). At the height of the offensive era in 2000, there was an average of 12.8 net baserunners per game.
Thus in an average eighth inning in 2000, a team would most frequently have sent its 5-6-7 hitters to the plate; now it's 4-5-6. The effective difference between the two innings is the difference between a No. 4 hitter and a No. 7 hitter. In 2011 the average No. 4 hitter homers every 25 at bats and has an OPS of .792; the average No. 7 hitter homers every 38 at bats and has an OPS of .697. That may not seem significant, but the margin for error with a reliever is small, so when a club's only home run threats are stacked in the middle, that makes a major difference.
Another way to assess the frequency with which hitters bat in which inning is to study the total number of batters faced in games won in nine innings and work backwards. Such a game (i.e. a nine-inning win) best replicates the situations a club is most likely to use its eighth-inning set-up man and closer.
The mode (the most frequently occurring outcome) has been 36 batters faced each year since 2000, the height of the Steroids Era, when it was 37 batters faced. Facing 36 batters means the No. 9 hitter made the game's final out in his fourth plate appearance. In most games that means a closer typically faced the 7-8-9 or 6-7-8-9 hitters and the eighth-inning reliever faced 4-5-6 or 3-4-5-6.
And the data is a bell curve: while 36 batters faced is the most likely outcome, the next two most common results were 35 and 37, which would only shift the likely batters faced by one in either direction.
For a closer to face an opponent's 3-4-5 hitters in the ninth inning, the other team needs to have had either 29 batters or 38 batters through eight innings -- in other words, either five men need to reach base in the first eight innings or 14 men; five is well below average and 14 is a bit above average.
As people around the game are taking note of these numbers, there is one distinct group to whom they are of secondary importance: the pitchers themselves. In fact, most late-inning relievers are oblivious to any overarching trend governing which batters they face -- and most don't care, either.
"It doesn't really matter," said the Yankees' David Robertson, who became an All-Star this year for his work setting up Mariano Rivera. "Once [the eighth-inning] comes, you're going in and you've got to get the job done."