Lee Hancock understood anonymity.

The left-hander made 24 mainly unremarkable appearances as a middle reliever for the 1995 and 1996 Pirates, a pair of largely anonymous teams. Thus, it was fitting that Hancock was able to draw the perfect analogy between those relievers who are not in the glamour role of closer and those football players whose job it is to allow their teammates to rack up impressive statistics.

"We're like offensive linemen," Hancock said. "When we do our job, nobody notices. When we don't, we're not anonymous anymore."

Set-up and middle relievers have begun to get a little more credit in the last few years, primarily in the form of All-Star Game berths. Still, in the traditional box scores, the only accomplishments for which pitchers get positive mentions are wins and saves, which middle relievers earn only infrequently.

The lack of control that starting pitchers have over wins has been well-documented in the sabermetric community, but others seemed to catch on last year when the Mariners' Felix Hernandez won the American League Cy Young Award despite a 13-12 record. Relievers have even less control over their records, as evidenced by the Athletics' Brad Ziegler, who went 3-7 and tied for the AL lead in relief losses despite a fine 3.26 ERA.

Red Sox set-up reliever Daniel Bard is not necessarily obsessed with statistics for non-closers, saying, "I know how I'm feeling, how I'm throwing the ball, how the hitters are reacting and if I'm contributing to the team winning. That's how I know if I'm doing a good job or not."

Yet, as a reliever who usually hands the ball off to closer Jonathan Papelbon to start the ninth inning and will be eligible for salary arbitration for the first time at the end of this season, Bard does know that statistics matter. That is why he feels a certain amount of frustration over there not being any readily available statistics—such as those distributed to media members covering major-league games—that measure the value of non-closers. Papelbon has averaged 38 saves during five full seasons in the big leagues, but Bard doesn't have a glamour stat to show for his work.

The closest thing to a mainstream stat for set-up men and middle relievers are holds, which are listed in the box scores in USA Today, most of the major metropolitan newspapers, and on all of the major sports websites. However, it is difficult to readily find a list of hold leaders because the hold is not considered an official statistic by Major League Baseball.

"It's really one of the few stats a set-up guy can take into an arbitration case, but it's not a good stat," Bard said. "I think anybody who pitches in my role will tell you that it's flawed."

Pitchers who blow a hold opportunity are charged with a blown save. However, because pitchers who pitch in hold situations rarely get a chance to work in save situations, it greatly skews their save percentages.

Bard, for example, has notched just four saves in 147 career relief appearances. Yet he has been charged with 12 blown saves, giving him a conversion rate of just 25 percent.

"The whole blown save thing bothers me because it doesn't reward you for the times you come in with the lead and protect it but get no credit for it. Yet it penalizes you for the times you don't hold the lead," Bard said. "Unless you're a closer, you're always going to have a bad save percentage. Last year I came into a game with one out and a man on third base in the seventh inning and we were winning by one run. I gave up a sacrifice fly and then got out the inning with the score tied in a game we ended up winning but I still got a blown save."

STATS LLC has developed a statistic called Holds Adjusted Saves Percentage, which is calculated by adding holds and saves and dividing that figure by holds plus save opportunities. However, about the only place that stat can be found is in the annual Bill James Handbooks.

"That would be fair," Bard said. "The thing about the hold, though, is you can use it as leverage in arbitration, but it's not a good stat. I really think it works against you more than it helps you because (the team) can point to your bad save percentage in a hearing. "

Baseball Prospectus' Fair Run Average measures a pitcher's effectiveness by taking into account the number of runs he allows as well as inherited and bequeathed runners. While Bard admits he is intrigued by FRA, he believes it might be too hard to explain to a three-person panel of arbitrators in a hearing. His preference for a non-closer statistic is WHIP.

"I think WHIP is a really good stat when it comes to showing how effective a reliever really is," Bard said. "A guy may not have a really good ERA, but if he consistently has a WHIP of around 1.00 or below then you know he's doing good things. With the amount of luck involved with being a reliever, the biggest thing is keeping guys off base because then you're limiting your bad luck as much as you can. You're not going to win a lot of arbitration cases with WHIP, but at the same time, a general manager wants to build a team that maintains a good WHIP over a period of time."

Rumors and Rumblings:

While Joakim Soria was removed as the Royals' closer on Monday in favor of rookie Aaron Crow, the move is only temporary. In a perfect world, the Royals want Crow to show he is capable of closing then allow Soria to get on track and close some games before trading him at the July 31 non-waiver deadline… The Rangers would consider trading outfielder David Murphy for pitching help, though they would prefer to deal center fielder Julio Borbon… The Rockies are becoming more concerned about right-hander Ubaldo Jimenez's drop in velocity, realizing that the root of the problem is more than the cut cuticle on his thumb that caused him to go on the disabled list early in the season… Marlins right-hander Josh Johnson won't return from the DL next Tuesday as hoped and will likely be out until at least the middle of the month. Meanwhile, shortstop Hanley Ramirez has admitted his slow start stems in part from lower back pain he has been feeling for more than a month… Blue Jays first baseman Adam Lind is playing in extended spring training games and should return to the lineup sometime next week… Cubs right-hander Matt Garza is expected to be activated from the DL and to start Sunday, and Tigers left-hander Phil Coke is expected to follow suit next Thursday.

Scouts' Views:

Indians right-hander Fausto Carmona: "He's slipped into most of his old bad habits: no command, letting innings get out of control, getting rattled on the mound. He's a very frustrating pitcher. You know the talent is there, but he beats himself too many times."

Rays right-hander Wade Davis: "He's really taken a step backward this season. His command within the strike zone hasn't been good, and he's leaving too many pitches over the heart of the plate. His velocity has also been up and down. He's still young, so I wouldn't give up on him, but he's going in the wrong direction."

Mariners third baseman Chone Figgins: "I'm just baffled by how this guy has gone off the cliff in two years in Seattle. He was a pretty good player in Anaheim. He looks totally lost at the plate, like he has lost all his confidence. I don't understand it."

Rangers catcher/designated hitter Mike Napoli: "I don't understand why no one will commit to making him an everyday designated hitter. I know he has deficiencies behind the plate and that he tends to be really streaky as a hitter. He has a great eye, though, and when he gets hot, look out. I'd love to see what he could do with 600 at-bats."

Braves center fielder Jordan Schafer: "He's added an element of excitement to that team since being called up. He can really fly and he looks a lot more prepared to be in the major leagues than he did when he was up two years ago. He makes better contact, he's taking more pitches and he's more consistent in the outfield now."

Thank you for reading

This is a free article. If you enjoyed it, consider subscribing to Baseball Prospectus. Subscriptions support ongoing public baseball research and analysis in an increasingly proprietary environment.

Subscribe now
You need to be logged in to comment. Login or Subscribe
Interesting that players are aware of needing to justify statistics to arbitrators (and wary of using advanced metrics for that reason).

It seems to me that a) arbitrators should know the industry and what's happening at BP and elsewhere sabremetrically, and b) agents should present their player's best case and explaining statistics may be a key part of that. Surprising that it's not the norm.

I did a quick scan for other 2012 arbitration eligible players in search of an easy-to-digest statistic for their agents to make their case:

Jeff Mathis and Eli Whiteside (how to measure non-hitting value of catchers...the "Nichols Metric?"...or in Whiteside's case, the ability to step in in Posey's absence...the "Organizational Soldier Metric?")

Elvis Andrus (how many runs does his defense save...the "anti-Jeter Metric?")

To a lesser extent, Paul Janish, Daric Barton, Dexter Fowler, and Andres Torres (Strat defensive ratings/the Fielding Bible?)

Relievers sharing the stage with Bard include set-up men Mike Adams, Joey Devine, Luke Gregerson, Jason Motte, and one of my favorites, a guy with a career whip in 140+ innings of under 1.00, more than a strikeout an inning and a K/BB ratio of 4.82 (2nd only to Strasburg among top 15 whip pitchers with at least 50 career IP), an ERA of 2.73, and an on base against of .252.

A non-closer, his team could argue that he has blown more save opportunities than he has converted and that he has failed in some key situations (including the playoffs last year, and, for that matter, last night).

Any guesses on who this is? ;-)

Sergio Romo
Indeed. Having a fine little career for himself so far.
Is Bard aware that the Red Sox haven't gone to arbitration with any player in about a decade?
John, I know it's not your focus, but I'm a little surprised that the Bard portion of this article does not make mention of the work that's been done with the WPA (win probability added) and LI (leverage index) statistics, at Fangraphs and elsewhere. Even if our understanding of how to properly credit a reliever for the leverage of the situations in which he is inserted remains imperfect, these stats have undeniably made an important contribution to the discussion. Indeed, I think they point the way to paydirt much more so than "Hold"-related stats (any stat that is derivative of the save stat is probably not progress), or even WHIP and FRA (both of which remain useful, of course, but neither of which reflects how the context in which a pitcher is used may affect the value of his performance). It may be out there already, but "paydirt" -- to me -- would be a stat which adjusts not just for game state context, but also for quality of opposing batters faced.
It sounds like arbitrators need to be educated, and, failing that, replaced.
I know this isn't the most glamorous, or even advanced, stat, but one when I watch Phillies games is "inherited runners scored/stranded" (pretty sure there is even a percentage in there). To me, a middle reliever is supposed to come in and get out batters. Should he come in with 2 runners on in a high leverage situation, and he strands both of them, he has done his job. Even if a reliever enters with his team up 10-2, and he allows a runner to score, he has not done his job. I know it sounds archaic, but that is a stat I think middle relievers should be judged on
To be fair, though, those stats should take into account where those runners are, and what the situation was when the reliever entered the game. In your example, suppose the middle reliever enters the game up 10-2 with a runner on 3rd and nobody out. If he gives up a ground ball, and the infield is playing back, is it REALLY the pitcher's fault that the runner scores? Conversely, if the bases are loaded with 2 outs, the reliever can very easily get credit for three runners stranded.
Exactly. This is what the more advanced metrics, like WPA and LI, attempt to do, using (I think) Markov chain-based analysis of changes to "game states," of which there are only 27 (from nobody on two outs, all the way to bases juiced nobody out).

WPA and LI attempt to go further, by accounting for the score as well, and what that means about how each event affects the ultimate outcome of each particular game. Which I think is fascinating. But to me, this might also run the risk of trying to capture "clutchness" in a statistic, a noble but maybe misguided goal. A better way, in my view, would be to focus solely on the expected runs in an inning from those "game states" (and the changes which result to expected runs following each batter faced), rather than on the probability of winning that particular game (and the changes which result in that probability following each batter faced). It's sort of a corollary to the Pythagorean formula, in that we should be more interested in runs scored and allowed (and the Wins and Losses that can be inferred from the same) than in actual wins and losses.

I don't know if any of what I just wrote made sense, but it makes sense in my head. Has to count for something.
I believe that what statistics the arbitrators are allowed to consider is negotiated as part of the collective bargaining agreement. It's not a matter of educating the arbitrators; they are told what statistics they may base their decision on and are not allowed to do their own research.
Not correct, actually. You're correct that the issue is collectively bargained, but this isn't what the current CBA says.

Check out Article VI(F)(12), which describes the "Criteria" for salary arbitrators to follow:

"The criteria will be the quality of the Player's contribution to the Club during the past season (including but not limited to his overall performance, special qualities of leadership and public appeal), the length and consistency of his career contribution, the record of the Player's past compensation, comparative baseball salaries..., the existence of any physical or mental defects on the part of the Player, and the recent performance record of the Club including but not limited to its league standing and attendance as an indication of public acceptance.... Any evidence may be submitted which is relevant to the above criteria, and the arbitration panel shall assign such weight to the evidence as shall appear appropriate under the circumstances."

The only meaningful limit to the presentation of evidence is time. One hour per side, plus a half hour each for rebuttal. So it *is* a matter of educating the arbitrators, but one must do so rather quickly. :-)