It’s not really going out on a limb to claim that the Royals’ trio of stellar relief pitchers was instrumental to the team’s surprise World Series run last year. Kelvin Herrera, Wade Davis, and Greg Holland collectively threw 64% of the Royals’ relief innings during the 2014 playoffs; 
that number shoots up to 77% in games in which the Royals were competitive, discounting their two blowout losses in the World Series. The trio allowed just five runs in 40.1 innings, not surrendering a single lead throughout October as Kansas City finished one agonizing run short of a title.
That dominance led to a great quip from Royals’ manager Ned Yost: “After the sixth inning, my thinking is done,” he said during the postseason.
Through the first 36 games of this year (all stats in this column are through Friday’s games), Yost hasn’t had to do much more thinking—I’ll refrain from making any wisecracks about the oft-criticized manager—as the three have picked up right where they left off. Herrera allowed his first home run since July 2013 and was suspended for throwing at a batter’s head, but he still sports a 1.93 ERA; Davis has tossed 16 scoreless innings with six saves in as many opportunities; and Holland has allowed just one run while converting seven of eight save chances.
Overall, the Royals’ bullpen has posted a tidy 1.61 ERA and held opponents to a .175 batting average—both the best marks in baseball and both on pace to set MLB records. The lowest bullpen ERA in history (minimum 100 innings pitched) comes courtesy of the 1968 Dodgers, who took advantage of their home park’s mountainous mound in the stingiest run-scoring environment since the Deadball Era to post a 2.22 ERA. And the Royals and Astros are both on pace to best that record this season.
By ERA-, an advanced metric that in effect neutralizes the 1968 Dodgers’ contextual advantages by modifying ERA based on ballpark and league averages, the best-ever relief performance comes from the Dodgers in Eric Gagne’s 2003 heyday. A whopping four teams this year—Kansas City, Houston, St. Louis, and Los Angeles—are well below that record mark through the season’s first month and a half, and though it’s premature to expect that pace to continue, the achievement is impressive nonetheless.
Those four teams, along with the Mets and Yankees, are the top six teams by bullpen ERA this year and double as the six division leaders thus far. That kind of correlation is unprecedented—under the six-division format, the most division winners in the top six in bullpen ERA in a single year is four—and while it can’t be expected to hold up so neatly for the rest of the season, it’s worth examining how important these team’s stout relief performances have been for their early-season successes.
For Kansas City, the exploits of the Herrera-Davis-Holland trio, as well as of their lesser-sung bullpen brethren, have been necessary due to the team’s set of dreadful starters. Of the five members of the Royals’ Opening Day rotation, only Edinson Volquez has an ERA below 5, and Kansas City’s bullpen has accumulated the fourth-most innings of any outfit this year trying to limit the damage.
The Dodgers’ bullpen ranks third in the majors with a 2.28 ERA and has similarly helped mask a mediocre rotation. Los Angeles’s cadre of fourth and fifth starters—the team has cycled through six already—have combined for a 5.12 ERA and just three quality starts in 13 appearances. Among the team’s projected starting rotation, Hyun-jin Ryu hasn’t pitched an inning this season, Brandon McCarthy is already out for the year, and Brett Anderson is a disabled-list visit waiting to happen. Beyond the leading duo of Clayton Kershaw and Zack Greinke, the Dodgers’ staff is worryingly thin, making the need for a consistent bullpen all the more pressing.
The Astros rank second in relief ERA (2.07), and their fifth starters have been even worse, surrendering nearly as many runs as they have recorded innings pitched (8.79 ERA) and, having made it through five innings just once, taxing the bullpen every fifth day. This group has allowed more earned runs in 28.2 innings than Houston’s entire bullpen has in 113 frames.
Beyond the six division leaders, 2015 as a whole has seen a surge in relief pitchers’ effectiveness. This season is on pace to have the second-largest gap between relievers’ and starters’ ERAs as well as the second-highest percentage of innings pitched by relievers of any year in MLB history. Better pitchers throwing more often, naturally, leads to even fewer runs allowed.
The relief advantage is increasing, and nowhere is that trend more apparent than in looking at the recent rise in strikeouts. Sixteen of the top single-season performances measured by strikeouts per nine innings (K/9) have occurred since 2010; for the best career ratios, eight of the top nine hail from active relievers (it appears that Rob Dibble, the lone exception, was ahead of his time).
In nearly every year over the last half-century, relievers have posted a higher K/9 rating than starters, but the difference between the two—in essence, the gap between the strikeout rates of relief and starting pitchers—has increased with every successive set of seasons.
Even more striking is that, in a listing of the largest gaps between relievers’ and starters’ strikeout rates, the years from 2010-2014 almost uniformly sit at the top.
The six division leaders this year exemplify this trend as the Dodgers, Yankees, and Astros sit atop the bullpen K/9 leaderboard. And for the Dodgers, already on pace to post the best such mark in league history, this rating only figures to improve: closer Kenley Jansen, who holds the third-best career K/9 mark in history, pitched for the first time this season on Friday and struck out four batters in one inning—I’ll repeat: four batters in one inning—of work.
But the most interesting and surprising team of the bunch is the Astros, so consider them as a test case.
In general, it doesn’t make sense for a team with an otherwise porous roster to chase a strong bullpen; an ace closer can’t do much when he never has a game to save or a lead to protect. Accordingly, teams out of playoff contention shed bullpen arms for prospects at the trading deadline each season, and the Braves took that philosophy a step further this year by preempting the July deadline and trading Craig Kimbrel before the season had started.
During their recent rebuilding years, Houston never bothered to populate its pen with high-quality arms, instead just jettisoning the occasional impact reliever to contending teams in July. The team’s numbers suffered accordingly: last year, no reliever who threw more than 10 innings for the Astros had an ERA below 3.
A team’s record in one-run games can fluctuate wildly from year to year and is often flukish, but the quality of a team’s relievers likely plays some role in a team’s one-run performance. From 2011-2014, then, the Astros parlayed one of the worst bullpens in the league into one of the worst records in close games each year:
Houston decided to roster a collection of MLB-worthy arms this year, though, and the results have been transformative. Of the Astros’ six pitchers with double-digit appearances this year, four are new to the organization, and Luke Gregerson, Will Harris, Pat Neshek, and Joe Thatcher have combined for a 1.96 ERA and 9.65 K/9 rating en route to shutting down opposing offenses late in games.
Relying on the newly staffed pen and a stable corps of dependable arms, Houston in 2015 has jumped all the way to the top of the one-run standings. With a 9-2 record in one-run games and a 3-0 mark in extra innings, the Astros’ ascension to the top of the AL West is in large part a function of their newfound close-game prowess.
Oakland, comparatively, is mired at the bottom of the AL West because of its league-worst 1-12 record in one-run games; perhaps coincidentally, the Athletics’ bullpen ERA (5.11) ranks 29th, with Oakland’s relievers allowing three runs more per nine innings than their Houston counterparts. Take out the two teams’ diverging records in close games and the Astros’ 10.5-game lead over the Athletics shrinks to a more manageable 1.5.
It’s easy to forecast that regression is coming and that Houston’s one-run winning percentage is doomed to fall. But that 9-2 record is already banked; FanGraphs currently projects the Astros, Angels, and Mariners all to finish the season with the same number of wins, and in a division that tight, every win that Houston can tally before regression hits could prove vital.
Of course, a good bullpen per se is neither sufficient to propel a team to the playoffs nor is it necessary for a team to make an October run. The losing streaks from the two New York teams over the past week show that even pens full of relievers doing their best Ernie Shore impressions can’t make up for offenses that score 16 combined runs across nine games.
Last year, to give just one example, the Padres’ pen posted the second-best ERA in the majors while the Dodgers’ ranked in the bottom third at 22nd overall, yet Los Angeles coasted to the NL West title. Overall, since the addition of the wild card in 1995, teams finishing in the top six in yearly reliever ERA have averaged a 90-72 record, with 53% making the playoffs. (Running the same analysis using the top six teams by ERA- returns remarkably similar results: an average record of 89-73, with 50% of the teams in question qualifying for the postseason.)
Applying that average to this season would suggest that of the six current division leaders, only three will qualify for the postseason. Among those six teams, only the Cardinals and Dodgers have a better-than-50% chance to win their respective divisions, with the wild-card route serving as the more likely option for the other four teams, per FanGraphs.
But for an organization like Houston, which hasn’t played in October in a decade, even a coin-flip chance of playoff baseball would be welcome. Of the 88 ESPN staffers who made preseason predictions for the website, not one touted the Astros as the projected division champion, and FanGraphs gave the team just a 15% chance of making the playoffs before the season began. Those odds are now up to an even 50%, and Houston has a 4.5-game division lead, the second largest in the majors.
You might say that having a set of reliable bullpen arms will be a bit of a relief for the Astros as they try to stay atop the division.