Keeping the Astros Bullpen on the Right Track

The Houston Astros, who made it to the World Series thanks at least in part to a bullpen that led the majors in xFIP (4.06) and placed second in ERA (3.75) during the regular season (the unit’s 11th place-FIP was still good, if a bit more pedestrian), saw four of its relief arms enter free agency last month: Will Harris, Collin McHugh (who threw innings as both a starter and reliever this season), Joe Smith, and Héctor Rondón. Here’s how those four stacked up in 2019:

Between them, the Astros’ four free agent relievers threw a little more than 32% of Houston’s 555 relief innings in 2019, and about 35% of their right-handed innings (505). That’s because Houston got an astonishingly small number of relief innings out of lefties in 2019: 49 and a third — the fifth-lowest such total in a decade — 35 of which came from Framber Valdez, who pitched only three times after September 1.

That imbalanced composition made Houston’s 2019 ‘pen unusually reliant on right-handers with a demonstrated history of getting left-handed hitters out. Lefties still aren’t close to the majority of all batters faced league-wide, of course, but they are 40% of the total, and so it behooves teams to have a plan for when they step into the box. The Astros did: Five Houston relievers — Ryan Pressly, Roberto Osuna, Cy Sneed, McHugh, and Harris — were better at retiring lefties than righties in 2019, when the league’s average tendency for relievers was the opposite:

The Astros’ top offseason priority should probably be their starting rotation, with Gerrit Cole seemingly extremely likely to depart (though Jim Crane is making noises about taking a run at him). They’ll also need to replace Robinson Chirinos and Martín Maldonado at catcher, where Yasmani Grandal may make sense. But the chart above suggests that retaining Harris and McHugh, at least, should be a priority for Houston as well. Letting Rondón and Smith walk will leave about 85 innings and Smith’s strong performance to replace, of course, but this year’s relatively strong relief market (10 relievers are projected for at least half a win, including Harris and McHugh) means there’s ample opportunity to do so if Houston is willing to spend a little money.

It’s not clear that the Astros will in fact spend — our RosterResource payroll page for Houston estimates the Astros 2020 payroll at $221 million with their luxury tax payroll estimate higher, and both are in excess of the initial $208 million luxury tax threshold. But if they do choose to spend, and in particular spend on their bullpen, they’ll have a number of intriguing options to choose between. Chris Martin, lately of the Braves, has the height the Astros like in their pitchers (he’s 6-foot-8), was significantly better against lefties than righties last year (allowing a .239 wOBA against them, versus .318 to righties), and has above-average spin on his fastball. Sounds like a Houston reliever to me. Robbie Erlin, Jake Diekman, and Will Smith could also be intriguing for different reasons (fastball spin rate, lefty splits, and overall competence respectively), but if I were Houston I’d feel pretty satisfied with an offseason that included signing Martin and retaining Harris and McHugh.

Despite getting beaten on a good pitch in the World Series, Harris will likely command a hefty premium this offseason as a number of contending teams seek bullpen help and take note of his sterling performance for the Astros over the last half-decade. The median crowd estimate you gave for his services was two years at $7 million a year — Kiley predicted two years at $10 million a year, which I think is somewhat more likely — but at either price, I think the Astros would be silly to let him walk, particularly given Rondón, Smith, and McHugh’s concurrent free agencies.

Like Aroldis Chapman, who just extended his time with the Yankees, Harris is well into his 30s and lost a mile per hour or so on his fastball and curveball in 2019. Those factors will probably keep the offers mostly to two years, as you projected, though I wouldn’t be surprised if the winning team ends up being the one that guarantees a third year. If that is the case, his new team should take comfort in the fact that Harris, like Chapman, adjusted to his declining velocity this season by increasing the rate at which he threw his breaking ball (in Harris’ case, that curve), and found success throwing that pitch out of the zone for strike three, or to steal a strike on the second pitch of a sequence, as we can see in this chart from Baseball Savant (cutters are in brown, curves in blue):

Bullpens aren’t everything, of course, but they’re of outsized importance in the postseason, where the Astros’ poor relief performances played a major part in their loss to the Nationals. As Houston stares down the decisions in front of them this offseason — the pursuit of Cole probably foremost among them — they’d do well to set a little bit of money aside for two of the players who helped carry them as far as they got last year, and perhaps a little bit more for one or two who can help them do more of what they did so well in 2019.