Blurring the bullpen:
Events of 2016 won’t scramble bullpen roles
The frequent use of Andrew Miller in the playoffs spurred debate, but traditional roles appear safe.
SARASOTA, FLA. Zach Britton was clad in a sharp, navy blue suit and green tie, not his Baltimore Orioles uniform, when he accepted his 2016 Mariano Rivera reliever of the year award in October, commemorating one of the greatest relief pitching seasons in baseball history.
Just before picking up his hardware, Britton took a detour across the grass at Wrigley Field to chat with one of his best friends in baseball, who was dressed a bit more casually.
Andrew Miller was preparing for Game 4 of the World Series, a night he added two more innings of relief work to a postseason that cast Miller as a convention-shattering playoff beast. Or so it seemed. Miller struck out 30 of the 73 batters he faced in the playoffs while entering games as early as the fifth inning — fueling a Cleveland Indians’ run that saw them nearly win a World Series largely on the backs of three relievers deployed by a manager unbound by convention.
It sparked hope among statistical purists that other managers would take a cue from Terry Francona and cast aside traditional bullpen roles, using relief aces in the highest-leverage situations rather than innings that fit predetermined job assignments.
Once the adrenaline-fueled nights of autumn give way to spring’s monotony, however, and the reality of the seven-month grind ahead comes into view, bucking convention looks a lot more daunting.
Certainly, no one’s expecting relievers to pitch with Miller’s October frequency and tenacity, a performance aided by extra off days and the coming offseason.
Yet a cross section of players, managers and analytics-driven executives agrees that even mildly altering reliever routines to maximize leverage over 162 games is a more complex task than it appears.
Britton smiles when asked if a disconnect remains between the executive suite — or the highinformation fan pining for the closer in the seventh inning — and the pitchers summoned to record the final outs of a game.
“Not every game is scripted the same way,” Britton told USA TODAY Sports. “Analytics are obviously bringing a lot to the game that’s very beneficial.
“But you have to realize there’s humans playing a game who don’t always feel the same every day. If you’re in the front office, sometimes you don’t know those things. Not that it’s a bad thing, but you’re running data upstairs, you’re not communicating with the players every day.”
Even within the most progressive front offices, the conflict is palpable. Balancing the desire to simply pitch your best pitchers in the most crucial situations while ensuring their health is a delicate tango.
“There’s definitely a difference between the ideal way to run a bullpen and what the pitchers can physically handle,” said Pittsburgh Pirates general manager Neal Huntington, who helms one of baseball’s most data-driven front offices. “The human element probably doesn’t get the respect it needs and deserves, because the same two pitchers, statistically, might not be the same in terms of recovery, of being able to (pitch) three out of four or five out of seven games.
“For us, that’s why reliever volatility is still so challenging. Guys who pitch well tend to pitch a lot. And then there’s a different set of guys who pitch well the next year who pitch a lot.”
The Orioles would be thrilled if their bullpen — tops in the American League with a 3.40 ERA — avoided that volatility.
Britton, 29, was nearly perfect in 2016, posting a 0.54 ERA in 69 games and giving up 38 hits, including one homer, and striking out 74 in 67 innings. He converted all of his 47 save opportunities.
It was an unfortunate twist that he was last remembered for the game he did not pitch — the AL wild-card game vs. the Toronto Blue Jays, when manager Buck Showalter did not want to insert Britton in a game that was tied on the road.
Baltimore’s season ended when long reliever Ubaldo Jimenez gave up an 11th-inning homer with Britton in the bullpen, a decision for which Showalter eventually expressed great remorse.
Britton was reduced to watching the playoffs and frequently texting Miller, whose legend grew after he struck out 14 of 26 batters to earn AL Championship Series MVP honors. Their offseason chats included whether Miller
single-handedly shifted the lateinning paradigm with his October dominance.
Britton’s conclusion? Not so much.
If a club inserts a top reliever as early as the seventh inning of a close game, it’s far likelier he’ll be asked to throw multiple innings if the result remains in doubt. And that will curtail availability.
“Do teams want to take away their best reliever for a two- to three-game period because they’re using him in that role?” Britton asked. “Would you rather have a guy throw one inning, three days in a row, to affect three games rather than one?
“That’s the question: Are games affected more important to you than the innings pitched?”
There’s no question at all for Showalter.
Blessed with a deep bullpen featuring a diverse array of arms and deliveries, Showalter ran it to near perfection and never rode Britton too hard or too often. The left-hander did not pitch three days in a row until working four consecutive days July 20-23. But that sequence was preceded by three days of rest. Britton pitched three consecutive days only one more time.
Showalter is convinced the cost of maximizing Britton’s leverage would far outweigh the benefits.
“It’s all stuff that’s really good on paper by people who make it sound real good, but it won’t work. It won’t work,” Showalter said. “It’s one thing to have a great weapon and a good pitcher like Zach — and then you abuse it, and you don’t have it anymore.
“That’s pretty stupid, isn’t it? Those things are bandied about by people who really don’t know.”
Relievers also bear a burden that doesn’t show up in the box score — instances in which they warm up but don’t pitch in a game. Managing more closely to leverage would likely increase those situations, since it can change so quickly.
Say a starting pitcher is cruising with a 4-1 lead in the bottom of the sixth inning, then hits the leadoff batter and gives up a bloop single. The leverage index leaps from 1.1 to 2.7, just shy of the “very high leverage” plateau of 3.0.
The team’s top reliever is ordered to warm up, quickly. But the starter finds his footing, inducing a double-play ground ball, and suddenly the leverage tumbles back to 1.4, on the low end of medium.
Does the reliever come in anyway, even though higher-leverage spots might emerge later? Or does he sit back down, a fireman with no inferno to extinguish?
“People definitely don’t take that into account,” said Tyler Thornburg, acquired in December by the Boston Red Sox to be their primary setup man. “If you get hot one day and don’t go in, it takes away that day of rest, almost.”
Showalter says there will never be another relief season as perfect as Britton’s 2016, and he might be right. It’s almost impossible to top what Britton gave the Orioles without pushing a reliever’s physical limits.
Britton led AL pitchers with 6.5 Win Probability Added, a statistic that measures how a player’s actions positively or negatively impacted his team’s chances of winning. That ranks him ninth among relievers in single-season WPA.
Tops on that list is Detroit Tigers reliever Willie Hernandez, who posted a 8.65 WPA in 1984 — a total exceeded only by the greatest seasons of starting pitchers Lefty Grove (1930 and ’31), Dwight Gooden (1985) and Bob Feller (1940).
Hernandez pitched in 80 games and saved 32 while logging a staggering 1401⁄ innings. He won the 3 AL Cy Young Award and the MVP for the eventual World Series champions — and the Tigers certainly made him earn it. Fifty of Hernandez’s 80 outings came with zero or one day of rest. In August, he pitched 12 times in an 18-day span, totaling 192⁄ innings. 3
The Tigers had 104 victories — and won the AL East by 15 games. By April 1987, Hernandez complained he didn’t have any strength in his arm as he headed to the disabled list. By August 1989, at 34, his career was over.
It’s unlikely that level of arm abuse would happen today, particularly for an MVP-caliber performer and especially as relievers are viewed as not just arms but assets. The bar for most lucrative reliever contract was raised thrice this offseason, capped by Aroldis Chapman’s $86 million deal with the New York Yankees.
And therein lies another disconnect: leverage and compensation.
Miller cashed in as a setup man in free agency, signing a four-year, $37 million deal with the Yankees before 2015.
Those not yet eligible for free agency aren’t so fortunate. Miller’s former bullpen mate, Dellin Betances, lost his arbitration case Saturday, getting $3 million instead of the $5 million he sought, along with a now-infamous tongue-lashing from Yankees President Randy Levine for dar- ing to aim so high. For now, saves still pay. “None of those arbitrators want to be the guy that just blew up the reliever market,” Britton said.
For now, most will have to settle for fame over fortune. Britton’s non-appearance in the wildcard game dominated the discourse in the early part of October. By November, it was Miller and Chapman in the spotlight as they entered games earlier and earlier.
Soon, 30 managers will choose to ignore or apply those lessons, to tweak or maintain the status quo.
“There’s definitely an opportunity to increase leverage and understand the most important out of the game might be in the sixth inning,” Huntington said.
“We are also still of a mind-set that the three toughest outs of a game are the last three outs.”