I vividly remember the first time I heard about the concept of the "save" in baseball. My friend explained to me that a pitcher was credited with a save if he got the final three outs in a game where his team won by three or fewer runs. He neglected to inform me of all the other caveats and ways a save can be earned, but I was somewhat confused nonetheless. I asked why this pitcher got his own stat for doing something that several other pitchers had done before him (enter the game and get outs). While he talked about the closer "nailing down the win" in a tight ballgame, I couldn’t help but feel that this whole "save" thing was kind of fabricated (which it was).
Fast-forward a decade and the "save" seems just as arbitrary to me now as it did the first time I heard it. With the birth of the save came the birth of the closer, and with the birth of the closer came the market for such pitchers. That market has grown more and more inflated in recent years with free agent closers finding themselves some overly-lucrative deals, all to rack up a stat that was created out of thin air back in the 1960s. And while we’ve read time and time again how saves have distorted reliever prices on the free agent market, there’s been less emphasis on how saves have distorted arbitration salaries.
Luckily for us, we have a current crop of arbitration-eligible relievers, some closers, some not, that we can analyze to see the effects that saves have on salaries. Before we can compare, however, we have to acknowledge how reliever salaries are determined through arbitration. As Matt Swartz points out over at MLB Trade Rumors, they are generally derived from counting the number of saves or holds accumulated by the pitcher in the most recent (platform) season, with past save and/or hold counts also used for consideration. Other stats, like ERA and strikeouts play a supporting role. The reliever going through arbitration is then compared to recent, similar relievers and a salary is determined.
But we know this is a poor way to determine salaries. Simply counting saves or holds and applying a little ERA magic does not tell us the whole story of the pitcher under examination. Instead, it paints a partially true picture with a whole lot of noise and luck mixed in. Regardless, this is the process used today and it has some wild effects.
To see them, I looked for two contemporary pitchers with similar sabermetric profiles, one of them being a closer and one of them not, and searched for a difference in salary. Luckily there were two pitchers with very similar statistical profiles in Luke Gregerson and Steve Cishek. I used Gregerson’s 2012 season, when he went through arbitration for the first time, and Cishek’s 2013 season, his first time through arbitration, to compare. The table below highlight’s their similarities.
By comparing these two with saber-friendly stats, we see that Gregerson’s 2012 and Cishek’s 2013 were incredibly similar. The biggest discrepancy is in terms of WAR, which gives Cishek a strong edge. Why is this? It appears that WAR punished Gregerson for a left on base percentage of 87.1% and the benefits of pitching inside Petco. That aside, the things that pitchers control show that these two were indeed very similar pitchers in these respective seasons.
So how did the arbitration process award these two? Gregerson got $1.55 million in his first time through arbitration while Cishek, who is yet to agree to terms with the Marlins, is projected to earn $3.2 million in his first pass through arbitration. Why? If all other things are nearly equal, the answer is the 52 saves that Cishek has racked up since 2011. Gregerson has pitched in a set-up role during that time span and has accumulated only 13 saves. Clearly the arbitration process is rewarding Cishek for the saves he’s collected as the full-time ninth inning man in Miami, even though he is essentially the same pitcher that Gregerson was when entering arbitration. This is due to the player comparison aspect of the arbitration process that determines Cishek’s salary based on other closers’ earnings rather than just pitchers with similar characteristics/performance (like Gregerson).
A quick look through the projected salaries of arbitration-eligible relievers backs this up. We can find average-ish closers like Ernesto Frieri being paid triple what fully functional pitchers, such as teammate Kevin Jepsen, earned at the same arbitration stage. Of course Frieri and Jepsen are different pitchers and the former’s strikeouts are sexy (12.84 K/9), but the latter’s FIP (3.38) was nearly a half a run better in 2012 and 2013 and he’s far less home run prone since Jepsen’s middle-of-the-road ground ball rate (40.4%) is far superior to Frieri’s horrific one (23.6%). I’m not arguing that Kevin Jepsen should be the Angels’ closer, but I’d like to suggest that the performance gap between them isn’t worth the divergence in salary that Frieri’s saves create.
Teams can and should take advantage of this discrepancy. By using fixed-cost relievers in the closer’s role and utilizing young, arbitration-eligible arms in the set-up role, they can keep the cost of their young assets down. A possible example of this scenario can be found in Arizona where the team just acquired Addison Reed who is arbitration eligible for the first time in 2015. They currently have JJ Putz under contract for $7 million in 2014 and he’s slated for a high-leverage role, too. By using Putz as the closer and limiting Reed’s save opportunities, they may be able to keep Reed’s first and subsequent arbitration salaries down somewhat. Depending on Reed’s 2014 performance, the savings could be in the neighborhood of $2-3 million and for a team trying to contend on a tight budget, this could be a consideration.
While this might sound like a pipedream, we might have already seen it take place. Last year on the North Side, the Cubs used an incredibly mediocre Kevin Gregg to "save" games while they let a very good Pedro Strop set him up. If they had allowed Strop to pick up the saves after coming over from Baltimore, they would likely be paying him over $2.5 million in 2014 instead of the $1 million that he’s projected to earn this season. By suppressing Strop’s saves, they’ve also suppressed his salary and at right around a million dollars, Strop is one of the best non-minimum bullpen bargains in the game.
Rebuilding teams or contenders under tight payroll constraints could utilize this strategy effectively. By using affordable, fixed cost closers now, they can keep the cost of their younger assets down, providing payroll flexibility. Maybe this is what Kevin Towers and Theo Epstein have in mind, and if not, this is at least the first time you’ve seen their names mentioned in the same sentence. Either way, saves are clearly distorting arbitration salaries and teams may already be looking for ways around it.
. . .
Jeff Wiser is a contributor to Beyond The Box Score and co-author of Inside the 'Zona, an analytical look at the Arizona Diamondbacks. He occasionally blogs about craft beer at BeerGraphs and you can follow him on Twitter @OutfieldGrass24.