No-Win Situation: Jacob deGrom Is Making History by Winning So Few Starts

Jacob deGrom's Cy Young candidacy shouldn't be punished by the Mets' ineptitude, which has significantly repressed deGrom's win-loss record in a dominant season.

I had spent no more than 15 minutes examining Jacob deGrom’s statistics before my head started hurting. The plight of the Mets’ ace is well known—deGrom has a league-leading 1.81 ERA but only seven wins in his 24 starts, although he has now won two in a row, including his start Monday night against the Yankees—and the far-and-wide telling of that plight has dealt an apparent coup de grâce to pitcher wins as a useful statistic. A representative item, from the Wall Street Journal: “The Mets Are So Bad They Killed the Win.” But as I would discover, his season has been even crazier than that.

If the win is dead, it had been on the ventilator for a while. The statistic had gone from a better-than-nothing cross-generational pitcher performance metric (hey, it was the 90s) to a nostalgic one in the recent present (as in, “Not only did Corey Kluber lead AL pitchers in ERA+, WHIP and WAR in 2017, he won 18 games!”). The pivotal moment in this transition came in 2010, when Mariners right-hander Felix Hernandez won the AL Cy Young award despite his 13–12 record; Hernandez had provided 50% more value, by Baseball-Reference WAR, than second-place finisher David Price, who went 19–6. Voters chose not to hold Hernandez responsible for the fact that Seattle’s league-worst offense left him in poor position to keep pace with Price in the win column.

The Dodgers' Bullpen Is an Unfettered Mess Without Closer Kenley Jansen

The win has never been especially suited to measuring a starting pitcher’s performance in his role, and it’s drifting further from its former target. In the days when starters would complete most of their starts, the win measured whether a pitcher had allowed fewer runs than his opponent, which was useful information but not a complete picture. In the days where pitchers threw as many innings per start as they could before fatiguing, the win measured whether a pitcher had, after exhausting himself, exited the game with his team leading, and, if so, whether the reliever or two who followed him had managed to hold that lead. So continued the slippage. And these days, the win measures whether a starter has handed off a lead to his vaunted bullpen and seen that bullpen excel. In 2008, relievers logged 35% of all innings pitched; in 2018, that number is up to 39%, and it stands to climb further as rosters expand in September. In 2008, eight pitchers threw at least 220 innings; last year, for the first time ever in a non-shortened baseball season, no pitcher did.

What is the pitcher’s job? Baseball is a game of run production and run prevention, and although NL pitchers have some hand in run production, run prevention is where the bulk of the pitcher’s value appears. And it figures that since run prevention is a valuable thing, teams want each individual pitcher to provide as much of it as he can. For relievers, that ask might take a number of forms, but for starting pitchers, teams call upon them, every fifth day, to throw as many innings as they can before the onset of ineffectiveness without exceeding 100 pitches by much. (There’s probably some room for tailoring a pitcher’s responsibilities to his particular body and capabilities, but that’s for another column, or for the Tampa Bay Rays.) Some role evolution has occurred here over time, but even as starters’ output has changed, their fundamental responsibility remains pretty similar to what it once was. Throw lots of innings; allow few runs.

Nine Innings: Deciding the NL East, the Rise of Position Players Pitching and October X-Factors

Most nights that deGrom has taken the hill for his Mets, he has done that job very well. To go with his league-low ERA, he is second in the NL in innings pitched per start and in quality-start percentage. But the Mets have gone 10–14 in his starts. Four times he exited the game in position to get a win only to have the bullpen blow it. Other highlights include: an eight-inning, zero-run outing against the Phillies when the Mets didn’t score until the tenth; an eight-inning, one-run outing against the Rays when the Mets didn’t untie the game until the ninth; and three eight-inning starts where he allowed only two earned runs and lost. He can thank the Mets’ offense, which is second-to-last in the NL in runs, and the Mets’ defense, which is third-to-last in the NL in defensive efficiency, and the Mets’ bullpen, which has allowed the third-highest percentage of inherited runners in the NL to score.

Baseball history, though, is littered with hard-luck pitchers. Without getting too granular, WAR attempts to control for all the variables out of the pitcher’s hands and to adjust for the league’s run-scoring environment, the ballpark’s effect on scoring, and the team’s defense—and WAR finds that deGrom is one of four pitchers in baseball history to produce at least 7.1 WAR but win 11 games or fewer. (I picked 11 games because, barring an injury or the Mets easing the throttle, deGrom will likely make eight more starts in 2018, and four wins seems optimistic but also reasonable.)

	Team	Year	bWAR	ERA	IP	W-L
David Cone	KC	1993	7.2	3.33	254	11-14
Roger Clemens	BOS	1996	7.7	3.63	242 2/3	10-13
Chuck Finley	ANA	1998	7.2	3.39	223 1/3	11-9
Jacob deGrom	NYM	2018	7.1	1.81	159	7-7

What about the pitchers who have produced deGrom-esque value but have not posted winning records in the expansion era? (Clemens’s and Cone’s seasons repeat.)

	Team	Year	bWAR	ERA	IP	W-L
Dave Roberts	SD	1971	7.4	2.10	269	14-17
Gaylord Perry	CLE	1973	7.8	3.38	344	19-19
Jon Matlack	NYM	1974	9.1	2.41	265 1/3	13-15
Bert Blyleven	MIN	1974	7.9	2.66	281	17-17
Phil Niekro	ATL	1977	7.2	4.03	330 1/3	16-20
David Cone	KC	1993	7.2	3.33	254	11-14
Roger Clemens	BOS	1996	7.7	3.63	242 2/3	10-13
Ben Sheets	MIL	2004	7.2	2.70	237	12-14
Jacob deGrom	NYM	2018	7.1	1.81	159	7-7

Clearly, though, even with a month left to play, deGrom’s season has little overlap with the others’. He’s been dominant on a per-inning basis in a way none of the other WAR/wins discrepancy cases were.

Perhaps a more illuminating statistic with which to examine deGrom’s performance is WPA, or Win Probability Added. What WPA measures is not the context-neutral value of a player’s contributions, but the value of a player’s contributions within the context in which they occurred—the ninth-inning walk-off home run is recognized as substantially more valuable than the ninth-inning home run when the hitter’s team down by 10, even though both outcomes are the best the batter could produce in any given situation.

In 2018, according to Baseball-Reference, deGrom’s total WPA is 4.1, which means, crudely, he has singlehandedly produced enough value for the Mets to win eight games. (Both teams start each game with a .5, or 50% chance, of winning.) That WPA figure on its own is not an uncommonly high one; Scherzer, who has done just as much to put an underperforming NL East team on his back, has 4.7 WPA this year. But deGrom is—get this—the only starting pitcher in history who has made at least 24 starts in a season to not record at least as many wins as he’s provided per WPA’s accounting. The only one! If you expand the list to starters who logged at least 100 innings, he gets a little bit of company.

	Team	Year	WPA	ERA	IP	W-L
Les Mueller	DET	1945	3.036	3.68	134 2/3	6-8 (26 G, 18 GS)
Ross Baumgarten	CWS	1980	1.294	3.44	136	2-12 (24 G, 23 GS)
Jacob deGrom	NYM	2018	4.083	1.81	159	7-7 (24 GS)

Baumgarten now works in wealth management in Naples, Fla. I tried to reach him on Tuesday, but we weren’t able to connect. Mueller, who was best known for throwing a record 19 2/3 innings in one July 1945 start against the Philadelphia A’s (the game finished a 1–1 tie after 24 innings), died in 2012.

Even if there is no one in deGrom’s immediate league, in expanding the parameters of the search I found him some worthy company. Resetting the goalposts to a minimum of 24 starts, but now including starters who had won just barely more than their WPA alone provided (wins = .425*WPA):

	Team	Year	WPA	ERA	IP	W-L
Craig Swan	NYM	1978	3.871	2.43	207 1/3	9-6
Pedro Martinez	BOS	2000	8.141	1.74	217	18-6
Roger Clemens	HOU	2005	6.032	1.87	211 1/3	13-8
Jacob deGrom	NYM	2018	4.083	1.81	159	7-7

On this list, deGrom can truly feel at home. He’s alongside two all-time greats having elite seasons on teams that just couldn’t hit, and one deeply unlucky Mets starter on a hopeless club that's won barely 40% of its games. I wonder: Who does he think he has more in common with?