The Tyranny of the Save

Allow me to take you back to the evening of September 5, 2017. The Blue Jays were in Boston for the middle game of a three game set, as they played out the string and the Red Sox moved towards the division title. Two ships passing in the night, heading in opposite directions. But behind seven shutout innings from Marco Estrada, the Jays scratched their way to a tenuous 2-0 lead, and it looked like for one night at least the script was to be flipped.

With Estrada done at 108 pitches, for the 8th inning manager DeMarlo Hale (Gibby was back in Texas on personal business) tuned to Dominic Leone. He took all of seven pitches to set down the side in order: a strikeout, a shallow fly ball, and a popout. Leone being one of the best options in the 2017 pen and clearly "having it" on the evening, naturally for the 9th Hale sent Leone back out to finish the job.

I jest. Of course not: 9th inning, save situation, have to go to the Proven Closer™. So in came Roberto Osuna, in the midst of a very rough five week stretch and who had thrown 26 pitches two days prior in a blown save. He issued a leadoff walk, a double, and then a pair of RBI groundouts in the course a 29 pitch outing. And so the game went on...and on...and on...until the Jays were walked off in the 19th inning. If anything, that's probably what one remembers about this game.

It was ultimately meaningless in the grand scheme, but in a season of full frustrations and disappointments, it made me as mad as I can remember being at any point. It's not that Osuna was a bad option, or that Hale did something unusually dumb. Indeed, there's probably not a manager in the game who would not bring his closer into that situation (if he has a closer of high ability). And that's entirely the problem.

That move was the paragon of "managing to the save", taking out an effective pitcher simply because it's the 9th inning and a save situation when either it's at best a marginal upgrade or the situation doesn't require it. Leone was in the midst of a dominant season, the most effective Blue Jays reliever in the second half. Most significantly, he was on his game that night. Pitchers vary from outing-to-outing, even the best ones have off nights - churning through relievers is a good way to find the guy who doesn't have it on a particular day. Taking out a guy who got three outs on seven pitches is just asking for trouble.

The Tyranny of the Save

More broadly, this type of management is emblematic of the save distorting baseball. Sportswriter Jerome Holtzman didn't think that existing statistics accurately evaluated relief pitchers, so he developed the criteria for saves. Unfortunately, as is too often the case, the cure was worse than the disease with Holtzman's folly (as Keith Law coins the save) being simply another junk stat divorced of real value.

It would be one thing if the save statistic was benign, a mere curiosity. But it's an official MLB stat and in the almost 50 years since that became the case, it's become the measuring stick for relievers. And in the process, the tail has come to walk the dog. In the heyday of the "fireman", the best relievers were the leaders in saves because they were most often on the mound when small leads needed to be protected, and finished out the game. As it should be. Saves were a by-product of a process to best deploy resources to win games.

That causation has been turned on its head with the rise of the Capital-C Closer who is explicitly reserved for save situations and thus almost invariably the 9th inning. And so one is more likely to see the team's best reliever used up three runs in the 9th than up one run in the 8th. Or see an effective reliever removed to make way for the Closer. And that's the tyranny of the save. It would be bad enough it were limited to impacting game management, and while it's the most pernicious effect of the save, it's hardly the only one. Let's look at a few others.

The Hall of Fame

Last month, Trevor Hoffman was elected to the Hall of Fame in his third year on the ballot, with 79.9% of the vote. He was undoubtedly a great one inning reliever, his change-up one of the great pitches in MLB history, and he was quite durable. Yet for as good as he was, with his 71 ERA-, he only threw 1,089 career innings. Over his career, Hoffman saved about 178 runs (189 taking out his last season) compared to league average, which is quite paltry by Hall of Fame standards. In fact, it makes for an interesting comparison to some other relievers.

At the end of the day, the job of a pither is to get outs and prevent runs. Dan Quisenberry pitched almost as many innings as Hoffman, and was his equal at suppressing runs - and was summarily 5%'d off the ballot in his first year. John Franco threw 15% more innings, same result. The difference? Saves, of course. 601 for Hoffman, just 424 for Franco and 244 for Quisenberry. The only favourable comp for Hoffman is Bruce Sutter, one of the most marginal inductees to begin with, and who also owes it to the save. Sutter ranked second in career saves behind Rollie Fingers in 1985, after which his career was over.

Hoffman's election is particularly interesting/galling coming on the same ballot on which Johan Santana was 5%'d off the ballot. We can debate whether on the merits Santana did enough to belong in the Hall of Fame, but over his career he pitched 2,025 innings to a 74 ERA-. Hoffman's innings were more important on average, but not enough to make up the difference in volume or degree of difficulty. To borrow a line: just as Ginger Rogers did everything Fred Astaire did, except backwards and in high heels; Santana did everything Hoffman did, but turning a lineup over three times every five days.

And that's to say nothing of Billy Wagner, who languishes around 10% on the ballot. A contemporary of Hoffman, despite pitching almost 20% fewer innings he still managed to prevent more total runs than Hoffman (196 vs. 178, ontext neutral) in the same role. But he only racked up 70% as many saves.

Salary Inequity, or how 95-3 trumps everything

Last month, the Jays and Aaron Sanchez agreed on a $2.7-million contract for 2018 as a first-time arbitration eligible player. Roberto Osuna was in the same position, and despite losing his hearing earlier this month, will make $5.3-million in 2018, or just shy of double Sanchez. That's quite interesting, for as much of a stalwart as Osuna has been, his career totals barely exceed what Sanchez did in 2016 alone:

That comparison alone makes the disparity in 2018 salary quite disproportionate, but it gets even worse. Sanchez has also accumulated 59.1 dominant innings in the bullpen, to the tune of a 1.67 ERA. Add that those to 2016, and Sanchez has 21% more innings at a (slightly) better rate of run prevention. Even if the rest of his starting innings are valued at zero, Sanchez has been the more valuable player and should be the one with the bigger salary (nevermind half).

You've probably guessed where I'm going, but the one area where Osuna has Sanchez hands down? Saves, 95 to 3.

***

Game management, historical recognition, and remuneration. The save is an octupus whose tentacles extend everywhere. The sooner its reign ends the better. I'm pretty confident in saying that the Earth will not spin off its axis if a Proven Closer does not come into a Save Situation.