The Mets found themselves in a dicey situation last Saturday night at Citi Field. Jacob deGrom tossed seven quality innings against the Marlins and left his bullpen with a five-run lead to protect. Sean Gilmartin then allowed a run in the eighth and Carlos Torres coughed up three more in ninth. With two outs in the top of the ninth, Dee Gordon on first, Giancarlo Stanton looming on deck, and their lead down to just one, the Mets brought in Alex Torres to face Christian Yelich.
Had Yelich reached base, the Marlins’ most dangerous slugger would have come up in a crucial spot. And who were the Mets warming up to face him? Not Jeurys Familia, their best reliever. Instead, it was Erik Goeddel—Goeddel, of the then-eight major league appearances and a 5.37 ERA in Triple-A last year.
That's because Familia had the night off after the Mets used him in five of their previous six games. Given a chance to rest his closer the day before, manager Terry Collins brought him in to protect a comfortable 4-1 lead in the ninth and thereby earn a save. As a result, Familia was unavailable to pitch in a much tougher and more important spot the following night.
Luckily, it worked out for the Mets. Torres struck out Yelich, the Mets won the game, and Familia came back the next day fully rested to close out the four-game sweep.
Still, it’s obvious that the Mets did not use their bullpen efficiently in that series. The Mets—like virtually every other team in baseball—are prisoners of an antiquated set of rules governing bullpen usage. According to these unwritten rules, a team’s best reliever should be anointed its "closer" and pitch primarily in "save situations." Teams therefore try to save their best relievers for the ninth inning of games that they’re leading by three runs or fewer. Should a crucial situation arise in the seventh or eighth innings, they will call on lesser pitchers to put out the fire, while their better pitchers watch the action from the pen.
This is not the best strategy for managing a bullpen. Instead, teams should simply use their best relievers in the highest-leverage situations. As such, bullpen usage remains one of the few areas of the game in which teams—despite being armed with an unprecedented amount of information—continue to shoot themselves in the foot.
Take the Friday night Mets/Marlins game as an example. When Familia entered that game in the top of the ninth, the Mets were winning 4-1 and had a 97.6% chance of winning the game. This was a low-leverage situation, evidenced by its 0.61 leverage index. Granted, that is probably somewhat understated, as leverage index doesn’t account for quality of hitter, and the Marlins had the heart of their order—Stanton, Martin Prado, and Michael Morse—coming up. Regardless, a three-run lead in the ninth is a very comfortable spot; using your best reliever there is not optimal.
Photo: Rob Foldy/Getty Images
Compare that to the following night, when the Mets rested Familia. After Gordon’s two-run single in the ninth made it a one-run ballgame, the Mets’ win expectancy fell to 93.2% and the leverage index rose to a significant 2.59. With Yelich and then the heart of the order coming up, it was probably an even higher-leverage situation than the index suggests; had Yelich reached base against Torres, the stakes would have only increased. That would clearly have been a better spot for Familia than was the previous night’s game, when the leverage index upon his entering was literally a fraction of what it was the following night.
We can find an equally striking example of bullpen misuse in the first game of that series. With the Mets winning 7-5 after seven innings, the Marlins had the heart of their order—Stanton, Prado, and Morse—coming up in the top of the eighth. The Mets’ win expectancy at that point was 89.3% and the leverage index was 1.29. With Familia sitting in the pen, the Mets brought in Carlos Torres to face three of the Marlins’ best hitters. Keep in mind that Familia’s ERA last year bested Torres’s by 85 points, and his FIP by 79 points.
Don't forget the benefits of being a Mets fan
From a Mets fan deep in Rays territory, everything Flushing and Citi Field have to offer isn't as obvious as you may think.
That night, Torres gave up a leadoff double to Stanton, but then retired the next three batters without allowing a run. By the time Familia took the mound in the ninth, the Mets’ win expectancy had increased to 94.3%. The leverage index was an identical 1.29, although, with the bottom of the Marlins' lineup—J.T. Realmuto, Adeiny Hechavarria, and pinch-hitter Reid Brignac—coming up, this was clearly a less crucial spot than the previous inning was.
Again, it worked out for the Mets. Torres pitched out of the eighth, Familia nailed down the save, and the Mets won the game. But just because the outcome was good in a handful of games doesn’t mean that the process made sense, or that the outcome will be equally good in the future.
To be fair, the Mets (like most teams) do tend to use their better relievers in high-leverage situations. Closers are usually among teams' better relievers, and the save situations in which they pitch tend to have higher-than-average leverage indices.
Last year, for example, Mets closer Jenrry Mejia led the team with an average leverage index of 1.47 upon entering the game. Collins also leveraged his other relievers so that the better ones—like Familia, Carlos Torres, and Vic Black—came into bigger spots than did weaker relievers like Daisuke Matsuzaka and Gonzalez Germen.
It will probably work out that way again this year: only Jerry Blevins and Rafael Montero in very small sample sizes have higher average leverage indices than does Familia. Buddy Carlyle and Carlos Torres are next, with Alex Torres in the middle of the pack, and weaker relievers Gilmartin and Goeddel rounding out the group.
While the Mets do a pretty good job of leveraging their relievers, their series with the Marlins last weekend shows that they can do even better. The absence of Mejia, Black, and Bobby Parnell makes Familia by far the Mets’ best relief option. Using him in a low-leverage "save situation" simply makes no sense if it renders him unavailable to pitch in a potentially bigger spot the following day. Furthermore, if an opposing team’s best hitters happen to come up in the eighth inning, it’s hard to justify saving Familia for the ninth to face the bottom of the order.
Improperly leveraging relievers is not an issue unique to the Mets. The "closer role" has become an institution in baseball, and teams treat it as its own position that only certain players on the roster can occupy. This is no doubt fueled by a popular myth that just a handful of pitchers—specifically, those with a "closer’s mentality"—have what it takes to finish games. However, the fact that so many good relievers with zero closing experience have seamlessly transitioned into the closer role really undermines that theory. Parnell, Mejia, and Familia, for example, had saved a combined 20 major league games and just seven minor league games before becoming the Mets’ closer with little fanfare and no rocky transition.
Finally, it’s odd that, once a reliever establishes himself as good enough to close, he is usually relegated to fewer innings. Last year, the 30 relievers with the most innings pitched averaged 74 innings, and few if any of them were longmen; the 30 relievers with the most saves, on the other hand, averaged just 62 innings. Of those 30 closers, only five reached the 70-inning plateau.
Moreover, while the 30 closers had an average ERA of about twenty-five points higher than the innings eaters’ (2.71 to 2.45), their FIP was about twenty points lower (2.80 to 3.01). This suggests that managers were riding relievers with low ERAs (a less reliable indicator of reliever performance), while limiting their more effective pitchers to save situations. So not only should managers use their best pitchers in bigger spots, but they should also use them more frequently.
Changing the closer role to a more flexible "relief ace" type of role probably can’t be done overnight. First, pitchers will need to be trained differently, as relievers who currently close would almost surely have to enter games with more runners on base than they do now. It will also take communication on the part of organizations to convey to their best pitchers that putting them in more important spots against other teams’ best hitters is not a demotion; in fact, it’s quite the opposite.
In short, bullpen misuse is a pretty glaring inefficiency that has yet to be fully corrected. It would be nice to see the Mets lead the way into a new era in which teams use their best relievers in the highest-leverage situations. If they don’t, some team surely will; the Rays already seem to be on their way.
In today’s information age, it’s simply inexcusable to continue a practice that makes no sense. The sooner the Mets transition away from the antiquated closer mentality the better, as doing so would put them at a competitive advantage against a league that’s been slow to adapt.