clock menu more-arrow no yes

Filed under:

We’re About to Witness the Best Collection of World Series Pitching Talent Ever

The Astros boast the two favorites for the Cy Young and another a top-tier starter. The Nationals have a three-time winner and two aces behind him. And it’s virtually impossible not to be excited about it.

Elias Stein/Getty

Justin Verlander and Gerrit Cole will finish first and second, in some order, in AL Cy Young voting this year. Max Scherzer, Stephen Strasburg, and Patrick Corbin will all place in the top 10 in the NL. Zack Greinke would have placed in the top 10, too, had he not been traded from one league to the other midseason. And for six of seven potential games in the 2019 World Series, these six co-aces will compete in the greatest collective World Series starting pitching matchup of all time. It’s easy, in many cases, to overstate the value of starting pitchers heading into a series—historically, rotation quality is not an indicator of postseason success. But this series is an exception; the pitching dynamic is the overwhelming story line of this World Series for good reason.

Washington’s top three starters are the majors’ highest-paid pitching trio. Houston’s top three starters rank second. Five of the top 10 qualified pitchers in strikeouts are pitching in this World Series, as are six of the top 17 in FIP, six of the top 16 in ERA, and six of the top 14 in a blend of FanGraphs’ two versions of WAR. And beyond the top six is Washington’s Aníbal Sánchez, who took a no-hitter into the eighth inning in his last playoff start.

This matchup thus poses a fitting finale to the 2019 playoffs, which has been characterized by a return of the starter, against the trend of increased playoff bullpen usage that predominated recent postseasons. Starters have a 3.20 ERA this month, versus 4.12 for relievers, and they’re pitching deeper into games and throwing more pitches than they have in years.

But do the Astros’ and Nationals’ rotations actually yield the greatest collection of pitching talent in World Series history?

For each World Series, we looked at the regular-season statistics for every pitcher who made a start and then averaged the numbers in question. We looked at four broad, context-adjusted figures to account for changes in run environment and to place every set of starters on even footing. For 2019, given that we can’t yet be 100 percent sure who will make a start, we ran two scenarios. In one, the Nationals use their four expected starters and Houston uses only Verlander, Cole, and Greinke (as they did in the ALDS). In the other scenario, Houston uses four starters: Verlander, Cole, Greinke, and José Urquidy, an above-average starter this season who threw the most innings and pitched for the Astros in their bullpen game in the ALCS, and projects as the most likely fourth starter if they choose that route in the World Series.

With those adjustments in place, let’s see how 2019’s World Series starters stack up against the past.

1. ERA-

This stat is simply ERA adjusted for ballpark and league context. Cole is the best of the 2019 World Series pitchers by this stat with an ERA 44 percent better than average this season; Verlander is close behind at 42 percent better, and Scherzer and Greinke both ranked in the top 10 among qualified pitchers as well.

Historically, the 2019 World Series starters rate quite favorably by this measure. The scenario with a three-man Astros rotation (which we’ll call Scenario A from now on) would rank second all time, while the scenario with a four-man Astros rotation (Scenario B) would place in the top five as well. (To understand this chart, know that for ERA-, a lower number is better; every tick below 100 represents 1 percentage point better than league average, so a 70, for instance, means 30 percent better.)

Best World Series Pitching Matchups by ERA-

Year AL Team NL Team # of Starters Average ERA-
Year AL Team NL Team # of Starters Average ERA-
1906 White Sox Cubs 6 62
2019 (without Urquidy) Astros Nationals 7 68.1
1942 Yankees Cardinals 7 69
1943 Yankees Cardinals 7 69.6
2019 (with Urquidy) Astros Nationals 8 70.8
1915 Red Sox Phillies 6 71.7
1938 Yankees Cubs 6 71.7

Beyond the lofty placement of the 2019 series, two items stand out from this first chart.

First, it’s worth a paragraph to appreciate that 1906 crosstown World Series. The 1906 Cubs have the best winning percentage in MLB history and won 116 games (with three ties) in a 155-game season, but the White Sox sprung the championship upset. More importantly for our purposes, that series’ starting pitchers produced outrageous league-adjusted numbers in the deadest part of the deadball era. Here are the regular-season ERAs for the six starters involved: 1.04, 1.51, 1.52, 1.65, 1.88, 2.06. Utter hilarity.

Second, throughout the metrics we’ll examine in this piece, older World Series tend to have superior high-end numbers than newer ones. Two underlying factors come to mind. Before the advent of four- and five-man rotations, teams used just two or three starting pitchers, and when taking a statistical average, it’s easier to find two or three well-above-average outliers than four or five. Also, the playoff bracket expanded in 1969 and then again in 1995, which placed more obstacles between the best rotations and the World Series. Before 1969, the teams with the best record in each league advanced all the way to the championship round; now, historically great rotations like that of the 2011 Phillies—who entered the postseason as the NL’s no. 1 seed with a rotation of Roy Halladay, Cliff Lee, Cole Hamels, and Roy Oswalt—can falter beforehand. If anything, this dynamic makes the 2019 placement all the more impressive, amid these long-ago competitors.

2. FIP-

FIP, or fielding independent pitching, strips away most contextual effects to try to estimate a pitcher’s true performance level. It ignores batted-ball luck and strand rate, instead homing in on the three true outcomes most in a pitcher’s control: strikeouts, walks, and home runs. Like with ERA-, FIP- then takes one additional step to adjust for park and league context.

By this measure, Scherzer was the best pitcher in 2019, with a FIP 46 percent better than average. Cole was second, and Strasburg, Greinke, and Verlander placed in the top 10 among qualified starters as well. Verlander’s drop from third in ERA- to 10th in FIP- is a result of his extreme home-run tendency—26 percent of his hits allowed were home runs, a record for qualified starters—despite an otherwise excellent season.

With this stat, there is no contest when comparing this series to historical precedent. In both Scenario A and Scenario B, the 2019 World Series is no. 1 by a wide margin.

Best World Series Pitching Matchups by FIP-

Year AL Team NL Team # of Starters Average FIP-
Year AL Team NL Team # of Starters Average FIP-
2019 (without Urquidy) Astros Nationals 7 72.3
2019 (with Urquidy) Astros Nationals 8 73.5
1996 Yankees Braves 8 80
2017 Astros Dodgers 8 81
2003 Yankees Marlins 8 81.5
1981 Yankees Dodgers 8 82

3. RA9-WAR

FanGraphs displays two wins above replacement calculations for pitchers: one based on actual runs allowed (essentially using ERA- as a base) and one based on the elements more in a pitcher’s control (essentially using FIP- as a base). We’ll look at the former first.

WAR is a counting stat, meaning it measures both performance and opportunity, so Verlander takes the top spot among 2019 pitchers in RA9-WAR because he pitched so well and led the majors in innings pitched. Here, all three Astros aces beat all three Nationals aces, as Cole ranked second in the majors and Greinke sixth, while the three Nationals placed in the 8-15 range.

When comparing the 2019 group to historical forebears, an important caveat arises: Because WAR is a counting stat, it’s a greater challenge for pitchers to accrue top-end totals now that starters throw fewer innings than they did in previous decades. The average World Series starter in the 2010s threw 23 percent fewer regular-season innings than he did a half-century ago, meaning 23 percent fewer innings in which to add value (Urquidy isn’t included in this calculation or graph).

That innings reduction wreaks a huge effect on the WAR ceiling for pitchers. Nobody has reached 10 RA9-WAR in a season since Pedro Martínez in 2000, and no World Series starter has reached 10 RA9-WAR in a season since Greg Maddux in 1995—the first season of the wild-card era.

So with that background in mind, it’s astounding how well the 2019 group performs in RA9-WAR relative to past World Series starters. Scenario A places fifth while Scenario B falls just outside the top 10, thanks in large part to Urquidy, who collected 0.8 RA9-WAR in just 41 innings and drags down the average by himself.

Best World Series Pitching Matchups by RA9-WAR

Year AL Team NL Team # of Starters Average RA9-WAR
Year AL Team NL Team # of Starters Average RA9-WAR
1906 White Sox Cubs 6 7.2
1912 Red Sox Giants 7 6.9
1969 Orioles Mets 6 6.8
1910 Athletics Cubs 6 6.7
2019 (without Urquidy) Astros Nationals 7 6.4

The 1906 Cubs–White Sox clash takes the top spot once again, and the 2019 series looks extra special given the broader context in WAR compression: Every other series ahead of 2019’s Scenario B occurred in 1973 or earlier.

4. FIP-WAR

FanGraphs’s other flavor of pitcher WAR uses FIP- as its base, and led by Cole, five of the top nine 2019 pitchers in this stat pitch for either Houston or Washington. It’s no surprise that 2019’s World Series, with its best-ever rotational FIP-, fares better by this measure than RA9-WAR. Still, given the difficulty of amassing large WAR totals with modern innings distributions, this final leaderboard is still something of a surprise.

Best World Series Pitching Matchups by FIP-WAR

Year AL Team NL Team # of Starters Average FIP-WAR
Year AL Team NL Team # of Starters Average FIP-WAR
2019 (without Urquidy) Astros Nationals 7 5.5
1963 Yankees Dodgers 6 5.1
2019 (with Urquidy) Astros Nationals 8 5
1974 Athletics Dodgers 6 4.8
1968 Tigers Cardinals 6 4.8
2001 Yankees Diamondbacks 8 4.7

Even with Urquidy, the 2019 series ranks second, narrowly behind the 1963 World Series that starred Sandy Koufax, Don Drysdale, and Whitey Ford. And without Urquidy, the 2019 series ranks first overall.

Put another way, based on regular-season value, the seven-headed starting monster of Cole, Verlander, Greinke, Scherzer, Strasburg, Corbin, and Sánchez would be the best group of starters in World Series history.


What’s unclear about these numbers is whether the dynamite pitching matchups will yield a competitive World Series. According to Vegas odds, the Astros are the strongest favorites in any Fall Classic since 2007.

It’s also possible that one group of starters just outperforms the other, one game at a time. Other past top pitching matchups produced lopsided World Series. The 1963 Dodgers swept the Yankees as they allowed four total runs in four games. The 1969 Mets beat the Orioles in five games and the 1974 Athletics beat the Dodgers in five as well, allowing nine and 11 runs, respectively.

Nor is a star-studded set of starters necessary for a memorable World Series. Looking at the 2019 series’ ranks across all four stats in question, it comes out as no. 1 overall, and no other series is close. By this holistic measure, the worst matchup of starting pitchers in World Series history came in 2014, when the Royals and Giants gave starts to post-prime Tim Hudson, post-prime Jake Peavy, Ryan Vogelsong, Jeremy Guthrie, and Jason Vargas. Everyone remembers Madison Bumgarner’s heroic relief outing in Game 7; fewer folks remember that Hudson and Guthrie started that game and combined to allow five runs in five innings before the bullpens combined for 13 scoreless frames.

But it’s also nigh impossible to look at these pitching probables and not grow immensely excited for the week of games to come. Cole vs. Scherzer. Verlander vs. Strasburg. Greinke vs. Corbin. And, if the series goes long, repeat. This is the best collection of World Series pitchers ever. They begin their showdown Tuesday night.