How Well Are Baseball's Best Relievers Being Used?

Are managers getting the most out of their key relievers?

For decades, we've pulled our hair out and grinded our teeth as we watched our favorite closers and set-up men attempt to get the final outs of a ball game. The closer is regarded by most managers to be the most important pitcher on the relief staff. However, this doesn't imply that the closer is pitching the most meaningful innings. It doesn’t even imply that the closer is the best reliever on the team.

Managers will often send in their closers in the ninth inning, up by three runs, to acquire the coveted save. This particular example isn't a high-leverage situation though. A high-leverage situation in which a manager’s best reliever should be pitching, for example, could be bottom of the ninth with the bases loaded and the score tied, or the eleventh inning with your team up by one run.

When a manager consistently uses his best reliever in medium- or low-leverage situations, he's using his bullpen in a manner that isn't optimal. To examine how baseball’s best relievers are being used by their managers, let’s find the reliever on each time who has the lowest FIP, with at least 22 innings pitched, and no games started. For each reliever, we'll determine the percentage of his innings pitched that are high leverage, medium leverage, and low leverage.

To define these terms, we will use Fangraph's leverage index. This system depends only on the inning, score, outs, and number of baserunners in a given state. At its most simple level, leverage index defines high leverage states as high stress situations while low leverage situations are less stressful situations that occur all the time in baseball. A medium leverage situation has a leverage index of approximately 1, a high leverage situation an index of 1.5 or greater, and a low leverage situation an index of less than 1. For reference, approximately 60% of game states have a leverage index less than 1 (e.g. low leverage), while 10% of game states have a leverage index greater than 2 (e.g. high leverage).

Our list of relievers, with data current as of June 26th (approximately the halfway point for many teams), is as follows:

TeamPitcherFIPIP% of High Leverage IP% of Medium Leverage IP% of Low Leverage IP
MarlinsSteve Cishek*1.6932.246.9426.5326.53
IndiansCody Allen*2.8933.245.5429.7124.75
BravesCraig Kimbrel*1.6230.143.9632.9623.08
PhilliesJonathan Papelbon*1.6031.143.6221.2635.12
TwinsGlen Perkins*1.9532.142.2729.8927.84
MarinersFernando Rodney*2.5130.139.5631.8728.57
PiratesTony Watson1.8736.238.1840.9120.91
CubsHector Rondon2.1829.137.5015.9146.59
DodgersKenley Jansen*1.9133.236.6331.6931.68
RedsAroldis Chapman*0.642234.8542.4322.72
AngelsJoe Smith*3.1332.233.6731.6434.69
RoyalsWade Davis1.4933.233.6649.5116.83
White SoxRonald Belisario*3.183731.5337.8430.63
AthleticsSean Doolittle*1.023730.6328.8340.54
RaysJake McGee1.703330.3035.3634.34
RangersJoakim Soria*0.7225.229.8724.6845.45
TigersJoba Chamberlain2.3032.226.5339.8033.67
Blue JaysBrett Cecil2.326.125.3232.9141.77
PadresJoaquin Benoit1.8431.225.2644.2130.53
RockiesAdam Ottavino3.163524.7636.1939.05
Red SoxAndrew Miller1.453124.7335.4939.78
BrewersZach Duke1.7931.124.4727.6647.87
MetsCarlos Torres2.9947.223.0827.2749.65
OriolesRyan Webb2.7035.121.7031.1347.17
GiantsJeremy Affeldt2.3426.116.4636.7046.84
DiamondbacksOliver Perez3.2030.214.1323.9161.96
CardinalsPat Neshek2.0631.213.6836.8549.47
NationalsCraig Stammen2.2841.112.9016.9470.16
YankeesDellin Betances1.084410.6142.4246.97
AstrosJosh Fields2.5527.19.7628.0462.20

* indicates a reliever who is a designated closer as of June 26th

Of the 30 relievers on this list (organized in descending order based on the percentage of the reliever’s innings that are high leverage), 12 of them would be considered full-time closers. The top-six relievers on this list who see the highest percentage of their innings as high leverage are all closers, while none of the bottom 14 are closers.

While managers have a tendency to use their closers in unnecessary situations, we still see that closers are used for high leverage situations more often than other types of relievers. However, even the best closers are still pitching in medium or low situations the majority of the time. Even Steve Cishek, the Marlins closer who tops this list, pitches in medium or low leverage situations 53.06% of the time. In fact, there isn’t a player on this list who has over 50% of his innings go towards high leverage situations. Aroldis Chapman, possibly the most dominant reliever on this list, has a 0.63 FIP, a 0.74 xFIP, and 1.3 WAR in just 22.2 innings pitched. But 65.15% of his innings pitched aren’t high leverage.

We also notice that certain managers prefer closers with lower FIPs, while other managers prefer closers with lower ERAs. The Dodgers use Kenley Jansen as their closer and have given him the most high leverage innings of any Dodgers reliever at 12.1. Despite a 4.01 ERA, Jansen has maintained a 1.91 FIP as of June 26th.

Alternatively, Adam Ottavino has a 3.18 FIP and a 4.46 ERA, but isn’t being used as the Rockies closer. Instead, LaTroy Hawkins is the designated closer for the Rockies, and has a 2.67 ERA despite a 3.91 FIP (and an even worse 4.43 xFIP). Therefore, Hawkins has seen 12 high-leverage innings pitched (the most of the Rockies relief staff), while Ottavino has seen 9 high-leverage innings (the third-most of the relief staff, behind Rex Brothers' 9.2). And only 24.76% of his innings have been high leverage.

Even if the pitcher with the lower FIP is most likely performing better, some managers are apprehensive to give high-leverage innings to pitchers who have been unlucky, whether it be from an unusually high BABIP or high HR/FB rate.

Joba Chamberlain, arguably the best of the Tigers relief staff, has seen just 26.53% of his innings go towards high-leverage situations because the designated closer for the Tigers is Joe Nathan. Nathan has the most high-leverage innings pitched on the Tigers relief staff, and 40.70% of his innings have been high leverage despite his 4.69 FIP relative to Chamberlain’s 2.36 FIP.

Josh Fields has the lowest percentage of high-leverage innings at 9.76% due to a 5.93 ERA that has been inflated by a .371 BABIP. Because of his misleading ERA, Bo Porter has given Fields just 2.2 high-leverage innings relative to Chad Qualls’ 9.2. Fields should see his ERA decrease in the second half due to his extremely high K/9 rate of 12.18 and, if Bo Porter is smart, an increase in his percentage of high leverage innings pitched.

One of the biggest crimes on this list is Craig Stammen, who sees a massive 70.16% of his innings pitched go towards low-leverage situations. Despite a 2.28 FIP, 2.83 ERA, and a 2.99 xFIP, Stammen has been rendered virtually useless by Matt Williams, who uses him in unimportant situations.

Managing a bullpen is one of the most complicated aspects of baseball. Managers are hopelessly unwilling to deviate from the “closer” system that mistakes medium and low-leverage situations for high-leverage ones. As a result, even the best closers find that around 53% of the time, they aren’t pitching in a high-leverage situation.

Managers also seem unable to decide how to use excellent relievers who aren’t designated closers. Some superb non-closers such as Pat Neshek pitch in high-leverage situations 13.68% of the time, while others like Hector Rondon are pitching in them 37.50% of the time. Unlucky pitchers with low FIPs but higher ERAs shouldn’t expect to pitch any meaningful innings, as their managers don’t want to take the heat for putting in a seemingly worse (though in reality, a better) reliever in an important situation. It’s a dangerous territory for managers to tread, and they appear unsure of how to do so.