Does being proven mean anything?
By: Aidan Flynn

Throughout the years, I have often heard of the notion that teams need "proven" closers in order to succeed. After all, how else would the Yankees have won all those titles without Mariano Rivera? With the reputation of the ninth inning "being a different animal" and more pressure-intensive, it is understandable to see why many fans and more mainstream analysts think that way.

Common sense would say that only those that have "been there, done that" can know the intricacies of shutting down the last three outs of a ball game. How else could it be possible for one to "save" the game without having pure intestinal fortitude and balls of steel? While many still hold onto this belief, I believe the myth of the proven closer is one based more on narrative than proven fact.  
Even though Rivera and Jonathan Papelbon continue to close out games at an alarmingly successful rate, the volatility of major league relievers (that is, reliever performance tends to be unpredictable and erratic) is one that is quite dramatic and noticeable. This volatility is real and can be attributed to the regression to the mean phenomenon,  an occurrence seen when extreme performances tend to balance or stabilize over time.

While this tendency afflicts all players, major league relievers are especially subject to this because their limited innings do not allow for much stabilization over the course of a season. In other words, the outside factors that allow a pitcher to succeed one season (luck, health, weaker opponents, etc), tend to disappear or decline substantially in the following campaign. Of course, with any sort of trend there are exceptions (the aforementioned Rivera, Papelbon, and Craig Kimbrel), but the turnover rate among relievers is staggering.

A study done by Baseball Prospectus' Ben Lindbergh in the book "Extra Innings: More Baseball Between the Numbers", found that an average of 60% of the top 50 relievers one season fail to make the list the very next year. Then, when comparing how those very top 50 relievers did three years later, only 26% of remained on the list. Below lists the findings of Ben's research (click to embiggen)
For those that prefer a bit simpler numbers, just look at the saves leaders the past two seasons. Among those in the top ten in saves in 2011, only one (Kimbrel) made the list in 2012. Both show a massive turnover rate and something that prevents even "proven" closers from remaining "proven" for very long. It's not that they cannot handle the heat; it's that their true talents are masked by various factors sometimes not stabilized until years later. 
For those still that still doubt me, let's look at a comparison of two pitchers, one with the "proven closer" label, and another who has been criticized for lacking the ninth inning mindset necessary to close.

2012 Numbers

Pitcher 1: 19 saves, 8 blown saves, 63.2 innings, 8.34 K/9, 4.1 BB/9, 5.09 ERA

Pitcher 2: 25 saves, 8 blown saves, 84 innings, 8.04 K/9, 3.32 BB/9, 5.36 ERA

Clearly, both pitchers had a rough time closing out games, and perhaps unsurprisingly, both were relieved of the role by the end of the season. Pitcher 1 happens to be Heath Bell, who averaged 44 saves and a 2.36 ERA as closer for the Padres, but struggled in his first and only season with the Marlins. Pitcher 2 is Alfredo Aceves, who had saved only four games in his career before he was thrust into the role in 2012. One is clearly more proven than the other, yet neither one was any good at closer last year. 

Skeptics will say that despite the struggles, Bell has been successful in the past while Aceves has not. This, although true, is a foolish and  idiotic way to rationalize the proven closer. Essentially, one is assigning "the closer mentality" only after the fact. Since Bell pitched well (even though he perhaps was pitching over his head and was the beneficiary of good luck), he gets labeled with having  the bulldog mindset necessary to finish out games.

Aceves, while a good pitcher in the past, experienced some bad luck in 2012 (admittedly, as did Bell), and gets labeled as lacking the guts to close. Also, his apparent hyperhidrosis (abnormal sweating) probably didn't help his cause with the media and fans alike. This success or failure, whether luck or true skill, acts as a confirmation bias to the "proven" closer hypothesis. This bias completely disregards the actual talent level and makes unnecessary and flawed assumptions based on a limited sample size. More often than not, people mistake past success as proven and past failure as utter incompetence in the role. Clearly, it's just not that cut and dry.
Evidence of Aceves' hyperhidrosis?
Additionally, if the above evidence isn't damning enough, we see unproven pitchers succeed in these roles every year. This year alone saw the major league leader in saves (the Orioles' Jim Johnson had 51 saves) have previously limited exposure to the closer role. Johnson didn't just pitch decent in this "pressure-packed" role; he thrived in the midst of the Orioles' first pennant run in fifteen years. A situation in which most believers would think he would crumble, Johnson was light's out. And it wasn't just Johnson. First year closers Tyler Clippard, Addison Reed, and Tom Wilhelmson all saved at least 29 games with SV% rates of 85% or higher.

The problem isn't the lack of sufficient pitchers ready to step in the role. It's that, quite often, the failures are given more publicity with the successes, giving the notion that only "proven" guys can succeed. In turn, front offices can succumb to the pressures of the media and fans by going out and spending money on a high-priced closer, just for the sake of a false sense of security. Even if the closer pitches poorly, it is seen more favorably because it shows the fans and media that ownership was "trying" despite having clearly better in-house alternatives. Not only is it bad business, but it only further adds to the legend of the "proven closer".

Often, when inexperienced pitchers are assigned the role of closer, they don't have a very long leash. Given its rising status as a prominent position, closers that fail to do the job (even with limited opportunities) get wrongfully branded as being incapable of the job.

I find it ironic, that before their actual performance is looked at (example: maybe pitcher X just isn't very good), people automatically assume that that pitcher just doesn't have the stones necessary to close. This probably bothers me most of all when I hear people discuss the need for the "proven" closer.

This last point of mine is more opinion than fact, but I always felt that if those that could make it to the major leagues and have the stomach to play in front of 40,000 drunk, obnoxious, obscene, fans every night, with even more watching on live television, they wouldn't be too fazed by three outs at the end of the game. Besides, those that are fazed by that stuff probably don't last long enough to even get the opportunity to close. These people have literally spent their entire lives playing baseball, specifically training their bodies and minds to withstand the pressures of the major leagues. 

Call me cynical, but I have a hard time believing a couple of outs, one of thousands made in their careers, should be any different.
Is the "save" what it used to be?
         Nasty does not begin to describe the stuff Aroldis Chapman and Craig Kimbrel possess.  The two young fireballers have turned into two of the most dominant relief pitchers in all of baseball.  Notice I said “relief” pitchers, not closers.  Although technically they are perceived as closers, they should not be used in the ninth inning exclusively to “close” out games.  Kimbrel is leading the National League in strikeouts per nine innings (K/9), while Chapman is right behind him at 16.35 K/9 and 15.60 K/9 respectively.  So, if your team has a one run lead in the 7th inning with the bases loaded and 1 out, why would a manager not go to his best strikeout guy? This is because of the save statistic and the fact that closers are defined by how many saves they can record. 

            Nowadays, you see a closer, someone who is generally regarded as the team’s best reliever, come into a game more often when the team has a 3 run lead at the start of the ninth inning.  The amount of pressure in these situations is drastically different.  One way to measure different pressure situations is with leverage index. Leverage index determines the weight of a particular situation due to the inning, score, and outs present at the time. This simplifies pressure-packed or pressure absent at a certain point during the game. For example, the leverage index for a reliever coming in the bottom of the ninth with nobody on and nobody out with a three run lead is 1.0, or in other words, a low leverage situation. On the other hand, a bases loaded, down two, nobody out scenario in the 7th assigns a high leverage number of 4.1. "Firemen" like Chapman and Kimbrel should be used when the pressure is at its highest point in the game. Yes, more often than not, this situation arises before the ninth inning, often representing the turning point of a ball game. However, some teams have done a better job of maximizing their pitcher's ability in pitching higher leverage circumstances. Guys like Fernando Rodney of Tampa Bay and Jim Johnson of Baltimore have emerged as the game’s best relievers and have become poster children for pitching with increased pressure. For example, Johnson leads the league in games entered with high leverage present (leverage index greater than 1.5). On the other hand, Rodney leads all relievers in situational wins saved, a computation of win probability added divided by the leverage index or pressure at a given time. Also, both pitchers rank in the top ten in terms of pitching with high leverage index. In other words, they have pitched the most in pressured packed situations. By maximizing a pitcher's ability and bringing him when the "fire" is its hottest, teams are better able to preserve and secure victories. Again, it is not a coincidence that these teams are either in the playoffs or were in contention until the last week of the season (Tampa Bay).

            Clearly, pitchers that can come in and shutdown a rally or strikeout somebody out when necessary is crucial to any team’s success.  A shutdown or meltdown would be a better way to look at a reliever’s success rather than just look at the save statistic. Shutdowns and meltdowns offer an alternative to the save statistic by simply analyzing whether a pitcher helped or hurt his teams' chances of winning. If a pitcher improved his team's chances of winning by at least 6%, he gets a shutdown. If he hurts his team's chances by 6% or more, then he is awarded a meltdown. This has no silly rules (rule 10.19) to follow and treats all relief pitchers equally (instead of favoring closers like the save statistic implies). Not surprisingly, the aforementioned “closers” have some of the best shutdown ratios in the league (Kimbrel: 37 SD, 4 MD, Chapman: 41 SD, 6 MD, Johnson: 46 SD, 3 MD, Rodney: 34 SD, 2 MD). Yet, even with these excellent numbers, it does not necessarily mean that pitchers are capitalizing on their talent. Would you rather have a middle relief journeyman pitching in a high leverage situation, or a talented flamethrower capable of getting a much needed strikeout? I think the answer is quite obvious and just further illustrates why the current status quo must be changed.