How to Fix Arbitration for Relievers

Last week, Dellin Betances lost his arbitration case after his agents attempted to argue that he should be paid like an elite reliever, which, of course, he is. But because arbitration is based on historical comparisons and mostly rely on traditional metrics, Betances wasn’t able to overcome his lack of saves, which is effectively the deciding metric for how much a reliever will get in arbitration.

This is a problem for Major League Baseball. The ideas about traditional bullpen usage are finally breaking down, and increasingly, teams are looking to deploy elite relievers in situations before the ninth inning. But if you’re a young pitcher, and you know that the system the league uses to value your performance depends almost entirely on how many saves you rack up, there isn’t a good incentive for you to agree to that kind of role. The way the system is setup, the best young relievers are financially motivated to try and move into the closer’s role as quickly as possible, because that’s the only bullpen role that arbiters put a significant value on. The arbitration system is effectively propping up 1990s-style bullpen usage, and it’s going to hinder the buy-in from players on a better way to deploy relievers during the season.

As Ken Rosenthal wrote, maybe the best answer to this problem (and the many other problems with arbitration) is to just get rid of the process entirely, which costs both teams and agencies thousands of hours of work for no real purpose. As Rosenthal notes, it wouldn’t be that difficult to design an algorithm that could determine the salaries of pre-free-agent players, and could take into account more meaningful metrics than the ones generally considered by the arbiters making the decisions now.

But at this point, that’s a pipe dream. Dumping the arbitration system might be a long-term reality, but Dellin Betances probably won’t still be pitching by the time that actually has the chance of happening. So how do we fix arbitration before one of the game’s truly great relief pitchers spends the next three years getting his pay docked simply because he’s not pitching the ninth inning? It’s probably easier than we might think.

As a quick bit of background, both sides are allowed to use any piece of data that appears in The Bill James Handbook; the league and the MLBPA have agreed that the numbers published in that annual publication are sufficiently credible to be presented during the case. Arbiters are given access to the BJH, and it’s on the table in the room when arguments are presented.

Helpfully, leverage is one of the metrics in the BJH, so both sides can actually fix the reliever valuation problem without overhauling the current arbitration system. All that has to happen is the league and the PA jointly encouraging the arbiters to move from saves to leverage index as the metric of determining whether a player was an important reliever or not.

Leverage index is already brought up in the arguments; it’s not a foreign concept to the arbiters. But as their rulings show, they’re still leaning heavily on save totals to determine comparisons. By simply moving the arbiters off of saves and onto leverage index for finding comparisons, the results would improve dramatically.

For instance, Betances had an leverage index of 1.77 when he entered the game last year, which was the 19th highest in baseball. He had a higher entrance leverage index than Aroldis Chapman or Kenley Jansen, for comparison. And that wasn’t a one year fluke; his 1.64 pLI from 2014-2016 ranked 28th highest in baseball.

And because Betances wasn’t just working one-inning stints, he’s actually been even more valuable than just the average of the leverage when he’s come into the game. Over the last three years, Betances ranks 9th in MLB in high-leverage innings pitched, and his .240 wOBA allowed in high-leverage situations ranks 12th-best of the 79 relievers who have thrown at least 30 high-leverage innings since 2014.

Leverage index isn’t some scary voodoo, but because it currently takes time to explain to the arbiters why it’s an important number, agents have to determine how much of their 90 minutes they want to allot to educating non-baseball-experts on a number that isn’t as common as saves. And because teams like the Yankees are individually incentivized to try keep their payrolls down, they’re not going to simply concede to using leverage index as a save-replacement metric when it works against their argument.

But overall, the league would benefit from flattening the payscale between pre-free-agent relief pitchers, rather than simply letting closers collect a lion’s share of the arbitration rewards. If both sides collectively agreed to tell arbiters to use leverage index as the primary way to find usage comparisons, mediocre save-compilers — I’m looking at you, Shawn Tolleson — wouldn’t end up making more money in arbitration than a truly elite reliever like Betances. Closers whose numbers relied primarily on saves would get a bit less, and elite setup guys used regularly in high-leverage situations would get a bump up, reflecting the reality of how baseball teams actually value these guys.

This wouldn’t even cost Major League teams more money, as they’d simply be redistributing money given to ninth-inning specialists to their entire bullpen, so we’re not even arguing that MLB teams should increase pre-FA reliever salaries in total. But by switching from saves to leverage index (or even just high-leverage innings totals), the arbitration system could more accurately reflect the game as it’s being played in 2017, and teams could ask their young relievers to pitch in ways most beneficial to the team without also asking them to limit their earning potential in the process.

It’s an easy switch, using a number already published in the book that both sides have agreed to use in arbitration. In 2017, there’s no real reason to still be using saves as a way to determine which relievers pitched in the most important spots. Everyone has better data, and there’s a simple way to move the arbiters off of saves for determining reliever salaries. Let’s just use it.