The Top Five Relievers in Baseball

How can you tell which relievers have had the most impact for their teams? Some might simply say “saves”, though that immediately eliminates any pitcher who hasn’t been given the opportunity to pitch in the ninth inning. It also doesn’t seem quite right to say that Huston Street and Tom Wilhelmsen, for example, have performed equally. Each has 11 saves, but Street has struggled with a 6.96 FIP while Wilhelmsen has been far more effective at 2.43.

Others might point to ERA, though it can be very unreliable over small amounts of innings, especially since it poorly accounts for ownership of inherited runners. We can see that with the case of Kansas City’s Tim Collins, who has actually been very good this year with a 1.75 FIP and 10 scoreless outings in 13 appearances. But thanks to one poor game earlier in the month, he saw his ERA jump from 2.79 to 5.59 overnight.

So what can we use that applies equally to all relief pitchers, regardless of role, yet also takes into consideration the primary job of getting outs (or not) in the most crucial situations? For that, we turn to shutdowns, a FanGraphs stat that attempts to measure the most basic question of “did a relief pitcher help or hinder his team’s chances of winning a game?”

Shutdowns (and its negative equivalent, “meltdowns”) works on the premise of win probability added. You can read the full description here, but it is essentially a context-based stat that identifies how important each play in a game was towards a team winning or losing.

This works perfectly for relievers, since the setup man who enters with the bases loaded in the eighth is often much more directly involved in the outcome than the closer who gets to start a clean inning in the ninth, even though it’s the latter pitcher who will get the “save”. Shutdowns are awarded when a pitcher increases his team’s win probability by at least 6 percent in a game, while meltdowns come when a reliever costs his team by at least that same amount.

Using shutdowns, who finds themselves on the top five list of most essential relievers this year? The answers, in some cases, may surprise you.

1. Mark Melancon, Pirates (17 shutdowns)
The top shutdown reliever in baseball so far is… who? Traded three times in less than three years and fresh off a 6.20 ERA and a long stint in the minors for the Boston Red Sox last season, no reliever in baseball has had a larger direct positive impact on his team’s fortune than the 28-year-old Pittsburgh righty. Relying on improved location and heavier use of his cut fastball, Melancon has been a revelation with the Pirates, posting an otherworldly 26-1 K/BB ratio. No pitcher with as many innings as he has tops his 64.1 percent ground ball rate, and when you’re getting strikeouts and grounders while almost entirely avoiding home runs and walks, success will almost certainly follow.

What of Melancon’s Pittsburgh teammate, Jason Grilli? The first-time closer has been outstanding as well, leading baseball in saves while sporting an excellent 32-5 K/BB mark of his own. But for as great as he’s been, Grilli ties only for just 18th in shutdowns since many of his saves have come with him starting a clean inning with a lead of a few runs, leading to low WPA. Melancon, on the other hand, has been in the thick of tight situations all season.

2. Edward Mujica, Cardinals (15)
Mujica didn’t begin the season as the St. Louis closer, and he didn’t get his first save chance until April 18, once Mitchell Boggs failed to adequately replace the injured Jason Motte. That keeps Mujica out of the top five list on the save charts, but he’s been so good that he’s been credited with a shutdown in 15 of his first 19 appearances.

Incredibly, Mujica is essentially getting by with just a single pitch, his lethal “split-change”, and he’s managed to control it with such efficiency that he hasn’t allowed a walk since April 3, in his second outing of the season. Stolen from the Miami Marlins for minor league third baseman Zack Cox last season, he’d spent most of the last seven years being decent but unremarkable for Cleveland, San Diego and Miami; now, thanks to reliance on his unhittable out pitch, he’s become one of the most dangerous relievers in the game.

t-3. Jesse Crain, White Sox/Mariano Rivera, Yankees (14)
Sometimes conventional wisdom and next-level thinking come to the same conclusion, and it hardly takes an advanced doctorate in statistics to point out that the immortal Rivera remains one of the best relievers in the game. That’s the case whether you’re using saves (second in MLB) or shutdowns (third).

The appearance of Crain ahead of Reed on this list shows how important the game can often be in the seventh and eighth innings, but the long-time setup man has quietly been an effective reliever in each of his three years in Chicago. After parts of seven years toiling in the Minnesota bullpen, Crain joined the White Sox prior to 2011 and has subsequently posted swinging-strike percentages of at least 11.6 percent each year, after never having topped 9.6 percent as a Twin. Crain’s name recognition remains low among casual fans, though that may not be fair to him.

 

t-5. Aroldis Chapman, Reds/Jim Johnson, Orioles/Addison Reed, White Sox
Reed has been steady all season long, only three times allowing more than a single hit in an outing while striking out 24 in 21 innings. Chapman, meanwhile, finds himself tied for 12th on the saves list with several other closers, one behind Street and one ahead of embattled Los Angeles Dodgers closer Brandon League, who is on the verge of losing his job.

That makes using saves problematic because it’s not adequately showing just how dominating Chapman continues to be, despite some homer trouble in Philadelphia last weekend. Although Chapman spent much of the spring preparing for an aborted attempt to move into the starting rotation, manager Dusty Baker has yet to allow him to pitch more than a single inning this year. That leads to oddities like the fact that lesser teammates like Sam LeCure and Logan Ondrusek have each pitched more innings in May than Chapman has. Despite the limited opportunities, few pitchers have managed to shut down the opposition as well as Chapman has, and the Reds may be better served by thinking outside the box to use him more often.

Johnson has not quite had the batted ball luck of the magical 2012 season in Baltimore on his side, seeing his BABIP rise from .251 to a more realistic .306. That’s largely why his ERA and FIP have each risen as well, but Johnson has counteracted that in part by increasing his strikeouts per nine innings from a poor 5.37 to a better 7.25 mark. Unfortunately for Johnson, he’s been hit hard in May — blowing three games in a row at once point — and he’s the only pitcher with at least 10 shutdowns who also has as many as five meltdowns, a very poor percentage. By comparison, Johnson led baseball with 46 shutdowns last season, but did so with just three meltdowns all year long.





Mike Petriello used to write here, and now he does not. Find him at @mike_petriello or MLB.com.

5 Responses to “The Top Five Relievers in Baseball”

You can follow any responses to this entry through the RSS 2.0 feed.
  1. libre says:
    FanGraphs Supporting Member

    Doesn’t a higher BABIP correlate with a higher K/9 out of necessity? That is, if fewer balls in play are turning into outs then shouldn’t there logically be a rise in K/9 without any improvement in K%? While Johnson’s K% has gone up slightly which makes this point kind of moot, I’m wondering if it’s still a point that should be made.

  2. Mike Petriello says:
    FanGraphs Supporting Member

    I don’t think they are necessarily connected. Since K’s are not “in play” the two metrics function independently of each other. If the end result is fewer outs overall, then the /9 becomes a different number to divide by, because fewer outs is fewer innings.

  3. kdm628496 says:

    i understand using shutdowns would produce a list of better pitchers than saves, but why not take it a step further and look at shutdowns minus meltdowns. or even another step and look directly at WPA?

    not criticizing, just curious. still a good article.

  4. Mike Petriello says:
    FanGraphs Supporting Member

    Thanks. Yeah, there’s a couple of ways to look at it. I didn’t like WPA as a whole because relievers are judged so much more on a game-by-game basis. SD-MD is an interesting perspective too, though.