The mirage of the "proven closer"

I was originally going to title this "the myth of the proven closer," but that isn't quite accurate. They do exist, but they are nowhere near as common as you might imagine. Yes, there are the Mariano Riveras and Trevor Hoffmans, who put up 30+ saves, season after season. But they are the exception, rather than the rule, even if they are the exception which has some GMs chasing after the next proven closer, as if it were a unicorn that pooped rainbows. Here's the reality. How many pitchers have put up 30 saves in each of the past three full seasons, 2011-2013? That would be one, the Braves' Craig Kimbrel.

Let's crank the machine back to 2011, take a look at the closers with most saves that season, and how they performed over the following couple of seasons. Here's the list, with the number of saves for each year. If they were in the top twenty that year, the background is green; if they were in the top forty, it's yellow, and if they are outside the top forty, the background is red.

The decay is obvious. Of the 20 closers with most saves in 2011, only seven were also on the same list the following year; four had a decent number of saves, but almost half (nine) more or less fell off the map as far as closing was concerned. There were a variety of reason for that, injury, ineffectiveness and retirement among them, but it goes to show, saves are a "skill" that can fluctuate widely from year to year. Those top twenty averaged 38 saves in 2011, but less than 50% of that (17) the very next season. Two years on, it's a sea of red: the average was down to a dozen saves, with only one-quarter of 2011's "top closers" having more than fifteen.

Studies have shown that something similar happens, regardless of the metric used, and the conclusion is more or less the same. Experience of being a closer, does not seem to make you a better one. Our siblings over at Beyond the Box Score looked at this using stats from the 2013 season. They took the relievers with the most save opportunities, and divided them into two groups, those with significant amounts of saves previously ("proven closers") and those without.

As BtBS's Neil mentions, "It's important to recognize a bit of selection bias here. The unproven closers only get into the sample if teams stick with them, so players like Phil Coke who were briefly given the closer reins and then lost them don't make it into the top 30 in save opportunities. That said, over the last two seasons, half the teams in the league have found pitchers to fill the closer role who previously had no experience... We can very clearly see that experience is not a necessary condition for success in the role." Or, put another way, "If you put a good reliever into the closer role, he will succeed."

It's also worth mentioning that the closer hasn't actually improved teams' chances of winning to any significant degree. Back in 1960, when the "save" was barely a twinkle in Jerome Holtzman's eye [he invented it the previous year, but it wouldn't become an official MLB statistic until the end of the sixties], teams going into the ninth inning with a lead won the game 94.7% of the time. More than half a century later, with the closer now an established part of the landscape, teams going into the ninth inning with a lead in 2013, won the game... 95.1% of the time. That's about one extra victory every three years or so.

This isn't to say that the Addison Reed for Matt Davidson trade was necessarily a bad one. Right now, with Davidson hitting .172 in Triple-A, it certainly doesn't seem like we got shafted, though it's worth pointing out that the third-baseman is still only aged 23. But it's an interesting contrast to Kevin Towers' other acquisitions of relievers, in particular J.J. Putz. When we signed him as a free-agent in December 2010, he had picked up only five saves over the previous two seasons (though he had been Seattle's full-time closer in 2006-07), but had a career ERA+ of 136. A good reliever became a good closer again, notching 77 saves for us in 2011-12.

When we dealt for him, Reed had a much less impressive ERA+ of 103, though his FIP of 3.30 was comparable with Putz's pre-Arizona figure of 3.35. Reed's FIP last year was better still, at 3.17 - but it's worth noting that still didn't even put him in the top fifty for relievers (min 50 IP); it sits between Casey Fien of Minnesota and Robbie Ross of Texas. Don't worry, I've not heard of them either. But it does suggest there wasn't much in Reed's performance to suggest he was an outstanding reliever or would be a great closer. And while I don't think he's as bad as his current numbers imply, the cost once he hits arbitration is going to be artificially high due to those shiny saves.

I accept there is a case to be made for having a closer: I can see there is something psychological about the ninth inning, and also that some players are happier with pre-determined roles, in terms of things such as in-game preparation. But the evidence seems to indicate the best closers are simply the best relievers. Closers will always come with a premium, so it make more sense to acquire (whether through trade or free-agency) very good pitchers who aren't currently closers, as the team did with Putz or Brad Ziegler; the latter certainly didn't cost us a top 100 prospect. Back in 2009, Steven Goldman wrote for Baseball Prospectus

Though the instability at the top of the reliever corps is great, the pitchers that move onto the list have to come from somewhere. Quite often, they come from the minor leagues, be they prospects or journeymen. This means that a GM's best option is often also his cheapest option... Even more indicative statistics, like H and K/IP, are subject to enormous fluctuation due to the increased bearing of luck on pitchers who throw only 50-90 innings. No team has ever bought a great bullpen. Those come about through luck - one or two front-line arms are backed by a series of second-line, bargain-bin refugees who happen to click behind them.

We've seen our share of "bargain-bin refugees who happen to click." Will Harris would be the most recent example. Sporting an ERA over eight in 20 appearances with the Rockies, he was nails for the Diamondbacks last season, with a 2.91 ERA and 2.74 FIP. This year, he allowed nine earned runs in 8.1 innings, and punched his own ticket back to Reno. That could be small sample size, as in AAA, he has given up only two ER in 13.1 innings at the time of writing. But Harris does show, you don't need to spend to get pieces that can be highly valuable and effective out of the bullpen, for at least one season.

Arizona do currently have one of the more expensive bullpens in baseball. When Wendy Thurm looked at the breakdown of team payrolls before the season, our relievers were costing just shy of $18 million. Only seven franchises spent more in raw cash than the Diamondbacks, and except for the Royals, every other team shelling out more, had a payroll of at least $130 million. It's not an enormous amount of money, certainly: but I can't help feeling that there may be better places for us to be spending it.