These days, New York Mets closer Jeurys Familia is saving everything in sight. On Tuesday night, Familia closed out the back end of New York’s doubleheader split against the St. Louis Cardinals for his 36th save of the 2016 season, the most in the majors. Toss out those pesky postseason stats, and Familia has successfully preserved Mets victories in 52 consecutive save opportunities, stretching back nearly a full calendar year — the third-longest streak in major league history.
But as any Mets fan can tell you, the sailing is not always smooth with Familia. Underneath his unblemished saves mark, Familia’s control can be erratic; the 3.4 walks he’s allowing per nine innings ranks 11th-worst among regular closers this season. As a result, he frequently pitches his way into varying degrees of peril before bearing down and escaping with a save.
To put Familia’s — and other closers’ — habit of pitching dangerously into perspective, I compiled what I’m calling the “heart-attack index” for relief pitchers. One of the core concepts is leverage index: how important any moment of a game is (relative to average) based on how much it could potentially swing each team’s probability of winning. On average this season, Familia has entered the game when the leverage index is 1.71,1 which means these moments are about 70 percent more important than a typical at-bat. In the average Familia outing, that number balloons to 3.08 at its peak, meaning the pressure ratchets up more per appearance for Familia than it does for any other qualified2 closer in the game:
RELIEF PITCHER INDICATORS | AVG. LEVERAGE | ||||
---|---|---|---|---|---|
PITCHER | SHUTDOWNS | MELTDOWNS | INITIAL | MAXIMUM | INCREASE |
Jeurys Familia | 30 | 2 | 1.71 | 3.08 | +1.37 |
Alex Colome | 16 | 5 | 1.62 | 2.95 | +1.32 |
Santiago Casilla | 18 | 10 | 2.27 | 3.51 | +1.24 |
Brad Ziegler | 21 | 3 | 1.68 | 2.88 | +1.21 |
Jeremy Jeffress | 19 | 4 | 1.59 | 2.79 | +1.20 |
Steve Cishek | 19 | 7 | 2.00 | 3.19 | +1.19 |
David Robertson | 23 | 6 | 1.76 | 2.86 | +1.10 |
Trevor Rosenthal | 10 | 11 | 1.52 | 2.57 | +1.06 |
Jonathan Papelbon | 17 | 5 | 1.69 | 2.68 | +0.98 |
Ryan Madson | 19 | 8 | 1.76 | 2.73 | +0.97 |
Jeanmar Gomez | 21 | 5 | 1.85 | 2.82 | +0.96 |
Mark Melancon | 20 | 4 | 1.54 | 2.49 | +0.95 |
Arodys Vizcaino | 17 | 6 | 1.48 | 2.43 | +0.95 |
Sam Dyson | 23 | 3 | 1.83 | 2.78 | +0.95 |
Will Harris | 26 | 2 | 1.60 | 2.44 | +0.84 |
A.J. Ramos | 20 | 2 | 1.60 | 2.41 | +0.82 |
Tony Cingrani | 17 | 9 | 1.58 | 2.35 | +0.77 |
Wade Davis | 16 | 2 | 1.51 | 2.28 | +0.76 |
Luke Gregerson | 23 | 6 | 1.70 | 2.46 | +0.75 |
Zach Britton | 26 | 1 | 1.63 | 2.37 | +0.74 |
Andrew Miller | 24 | 4 | 1.72 | 2.46 | +0.74 |
Francisco Rodriguez | 18 | 3 | 1.69 | 2.41 | +0.72 |
Shawn Tolleson | 9 | 8 | 1.38 | 2.08 | +0.71 |
Roberto Osuna | 19 | 3 | 1.39 | 2.09 | +0.71 |
Cody Allen | 17 | 5 | 1.67 | 2.36 | +0.69 |
Craig Kimbrel | 16 | 6 | 1.74 | 2.42 | +0.69 |
Kevin Jepsen | 11 | 9 | 1.55 | 2.19 | +0.63 |
Aroldis Chapman | 17 | 3 | 1.90 | 2.53 | +0.63 |
Hector Rondon | 14 | 6 | 1.50 | 2.13 | +0.62 |
Fernando Rodney | 17 | 4 | 1.52 | 2.12 | +0.60 |
Carlos Estevez | 12 | 8 | 1.51 | 2.08 | +0.56 |
Kenley Jansen | 23 | 5 | 1.82 | 2.35 | +0.53 |
But pitching into trouble3 is only half the Familia formula — he’s also been masterful at pitching his way out of the jams he creates. To get a sense both for the holes that Familia digs and for his ability to climb out of them, I plotted the amount of win probability every reliever loses per at-bat before hitting the maximum leverage index of each appearance against his win probability added per at-bat from that moment onward:

Unsurprisingly, there’s Familia again on the top left, losing more win probability than any other qualified reliever before the pressure-valve alarms start to buzz, only to also gain the most win probability afterward as he brings the team back from the brink of a meltdown. (To some extent this is circular — Andrew Miller of the Yankees, for instance, seldom pitches himself into enough trouble to rack up huge win probability credit on the other side of max leverage — but that’s also kind of the point.)
Other relievers, such as deposed Chicago Cubs closer Hector Rondon, pitch into nearly as much trouble as Familia but aren’t as skilled at extricating themselves from danger. (Which might help explain why the Cubs on Monday dealt for Aroldis Chapman, who gets into considerably less trouble on the mound than Familia but pitches well when in a jam.)
In Familia’s case, it’s fair to ask how much of this high-wire act is repeatable: Surely his save streak will eventually end if he keeps pitching so dangerously, right? Well, maybe. WPA isn’t a very predictive statistic because of the large weight it places on high-leverage situations; if a player performs above his usual level in important moments, it’s unlikely to hold up for very long. But by the same token, Familia isn’t a bad pitcher who just happens to have pitched great under pressure — putting leverage-based metrics aside, he also ranks eighth among qualified closers in fielding independent pitching, despite the shaky walk rate. (It’s also worth noting that Familia’s signature pitch, the sinker, tends to either miss low in the zone — hence the walks — or induce ground balls, which often double up base runners who walked to first.)
Familia may not be the best closer in baseball, and his save streak is destined to end sooner or later. But a pulse-pumping pitching style isn’t automatically a bad thing — provided fans’ hearts can take the drama while they watch from the couch.