Time To Step It Up, Dodger Bullpen

wilson-bullpen-2014-04-30What’s the best way to evaluate a major league bullpen? There’s not really an obvious answer in the way that I would go to WAR or wRC+ for hitters, or FIP or xFIP for starting pitchers. It sure as hell isn’t wins, obviously. ERA really doesn’t work there, either, as we saw just a few days ago when Brian Wilson was atrocious and the runs were all charged to Hyun-jin Ryu. It’s not saves, which exclude most of the non-closing relievers, and it’s absolutely not the horribly flawed “blown saves,” which put setup men in the position of only being able to blow a save, not attain one.

I think part of it is that relievers tend to be in situations that seem like they impact the game so much more than other players, are; that is, when a starter gives up a run in their first inning, it’s generally not a huge deal, especially if they’re fine for the next six or seven innings. If a reliever does, it’s often a disaster. People want game situation context in their reliever value, even if that’s somewhat unfair, in the same way that success with runners in scoring position for hitters is neither a predictable nor repeatable skill. (“Shut up,” say the 2014 Cardinals.)

Still, no one would argue that the Dodger bullpen has been good, and you can’t just throw up your hands and say that no stat captures it correctly. I suppose I’d lean towards two . First, there’s WAR, which isn’t perfect here, partially because it’s a counting stat and not every bullpen has the same number of innings thrown, but is at least directionally accurate, and isn’t a good look for the Dodgers:

bullpen_fwar

If you prefer something more context-based, I also like “shutdowns” and “meltdowns,” FanGraphs creations that that credit or demerit a reliever if they add (or cost) six percent win probability, no matter what inning it comes in, which puts closers and non-closers on an even scale. (A fuller description here.) The Dodgers have the 17th most shutdowns; they have the 12th most meltdowns. But in terms of net shutdowns, or how many more shutdowns a team has than they do meltdowns, that’s not necessarily great for the Dodgers either:

shutdowns-meltdowns

This is, of course, not what we expected from a bullpen that came into the season with:

— One of the four or five best closers in the game in Kenley Jansen
— A formerly great closer coming off a wonderful end to 2013 who other teams wanted as their closer, Brian Wilson
— A solidly effective lefty in J.P. Howell
— Two young talents with big league success in Chris Withrow & Paco Rodriguez
— Two former All-Stars who needed merely to hold down low-leverage setup roles in Chris Perez & Brandon League
— An ageless utility reliever who seemed to be getting better as he got older in Jamey Wright
— 
Young talent in Triple-A just waiting for a chance in Jose Dominguez and Yimi Garcia

And… it just hasn’t been working, for a variety of reasons. Wilson was injured and has mostly been terrible, as we all wait to hear that he’s been pitching with a bad arm. Withrow missed plenty of bats but couldn’t find the plate before blowing out his elbow. We’ve discussed Jansen’s issues (or non-issues, depending on your perspective) endlessly. (Spoiler alert: He’s still awesome.) Rodriguez has been in Triple-A for the last month, and while he’s missed bats, his walk rate is concerning. Perez has been a disappointment by metrics both old and new; Wright has just sort of been there, rarely terrible or great, to the point that I have just about no immediate memory of anything he’s done at all, and now that the rotation is at full strength, Paul Maholm has found his way to the bullpen as well.

90topps_jphowellIt hasn’t been all bad, of course. Howell has generally been outstanding, and League, last night’s misstep aside, has actually been a nice rebound story after last year’s disaster, but overall, things aren’t getting better. In March/April, the bullpen had a 3.35 FIP; in May, it’s 4.26. Unfortunately, this is going to end up being more a recap of what’s happened than it is a prescription for how to fix it, because there aren’t obvious solutions. One would hope that with a healthier rotation than what we’ve seen to date, when Clayton Kershaw & Ryu both missed time and Maholm, Red Patterson, and Stephen Fife had to fill in, will help with longer starts and less bullpen abuse. (Only two teams have thrown more bullpen innings.)

If Wilson needs a disabled list stint, I’m fine with that. If Perez or Maholm gets DFA’d, I’m totally fine with that. (People lump Wright into that, but there’s no need to, especially since he can go multiple innings.) It’s just that a shakeup for the sake of a shakeup isn’t usually a great idea. Everyone wants Rodriguez back because of how great he was for 80 percent of last year, but he was awful in September, mediocre in April, and hasn’t necessarily been stellar in Triple-A. (With the usual “small sample, Albuquerque” caveats attached.) Colt Hynes continues to have a great K/BB, but as Chris Jackson constantly reminds me, that comes wrapped around a lot of loud hits. Dominguez’ arm is intriguing, but he’s also walked 10 in just 14.1 Triple-A innings, and for a bullpen that already has the majors’ second-highest walk percentage, that’s not something you want to add.

Maybe we will see Garcia soon, who has been seemingly very good — warning, totally scouting the stat line, which is almost always dumb, and I’ve never seen him pitch — for the Isotopes. But really, my worry here is the same as yours; that if things don’t improve, Ned Colletti is going to end up trading Zach Lee or Chris Reed or someone potentially useful for Fernando Rodney or Tommy Hunter or Hector Rondon or whoever this year’s Octavio Dotel is going to be. For the sake both of winning ballgames now and ensuring something dumb doesn’t happen later, this bullpen really needs to step it up. The rest of this roster isn’t perfect, but this is clearly the biggest sore spot on the team right now.

About Mike Petriello

Mike writes about lots of baseball in lots of places, and right now that place is MLB.com.