The Stars Aligned for the Nationals

With their sweep of the St. Louis Cardinals, capped off by a 7-4 win last night, the Washington Nationals are bringing the World Series back to the nation’s capital for the first time since 1933.

No team gets to the World Series without their fair share of luck, and the Nationals certainly have seen things go their way so far this October. But at the end of the day, talent reigns supreme. It shouldn’t come as a surprise to you that the Nationals were a superior team to the Cardinals. They produced 48.3 WAR this season, more than 10 wins above the Cardinals’ aggregate of 37.9. Washington’s hitters produced a wRC+ eight points higher than St. Louis’; their pitchers produced a FIP- six points lower. The Nationals were simply better across the board. What’s arguably most exciting for fans in Washington is that their top talent has stepped up when things have mattered most.

One of my favorite statistics to follow during the postseason is Championship Win Probability added, or cWPA, housed on the website The Baseball Gauge. It’s very similar to WPA in that it calculates how each plate appearance during every game has changed each team’s odds of winning the World Series. The Ringer’s Ben Lindbergh has covered cWPA in the past, such as in this piece about players who made late-season debuts and contributed to a postseason run, or in this one when analyzing the relative “mundanity” of the 2018 World Series. As you might expect, Nationals players are dominating in cWPA this postseason. Four of the top five individual cWPA leaders don the Nationals’ red, white, and blue:

2019 cWPA Leaders
Rank Player Team cWPA
1 Gerrit Cole Astros .200
2 Anthony Rendon Nationals .131
3 Juan Soto Nationals .123
4 Max Scherzer Nationals .092
5 Aníbal Sánchez Nationals .086
SOURCE: The Baseball Gauge

It’s not entirely surprising that the Nationals have this volume of players atop this leaderboard. In order to increase your World Series win probability substantially, you need a World Series win probability that is relatively low to start. That’s why we don’t see nearly as many Yankees or Astros on this list. The Nationals — if you remember, because it really does feel so long ago — were not only participants in the Wild Card game, but they seemed destined to lose. The team was down 3-1 going into the bottom of the eighth inning, with their lineup set to face Josh Hader. At that very moment, the Nationals had just a 17.8% chance to win the game, let alone the World Series.

Then Juan Soto stepped in. Two outs, bases loaded, bottom of the eighth — exactly the situation in which you want a superstar hitter at the plate. Soto delivered with a single into right field, and thanks to a Trent Grisham error, all three runs scored and the Nationals took a 4-3 lead. That play remains the most important of the postseason to date. Soto (and, inadvertently, Grisham) improved the Nationals’ odds to win the World Series by 7.3 percentage points.

Soto is also responsible for the third-most important play this postseason, too. His game-tying home run off of Clayton Kershaw in Game 5 of the NLDS was worth 5.7 percentage points of World Series win probability. The larger lesson here is pretty simple: If you want to improve your cWPA substantially, win elimination games in the most dramatic fashion. Three of the six most important hits this postseason came in that Game 5 of the NLDS: Anthony Rendon’s double in the top of the 10th inning (+.063 cWPA), Soto’s homer in the 8th (+.057), and, of course, the Howie Kendrick go-ahead grand slam (+.040).

What’s more fascinating about this Nationals’ run is that, perhaps outside of Kendrick (the NLCS MVP), there hasn’t really been an unsung hero. And I would even argue that while the grand slam was certainly a cool moment, it just confirmed the already-near-certain Nationals victory. The Nationals had the bases loaded with nobody out in the 10th inning. A traditional RE24 matrix would expect 2.29 runs to score in that situation. By our individual game odds, they already had an 84% chance to win before Kendrick even stepped up to the plate. Just to emphasize how significant that is, remember that the game was still tied.

(As an aside, you have to feel for Joe Kelly, whose -.115 cWPA this postseason is by far the worst of any individual player. He appeared in three games and was involved in only 18 plays, yet they were crucial to the Dodgers’ title hopes.)

Kendrick ranks 55th overall in the majors this postseason with just a +.012 cWPA. Outside of his homer off of Kelly, Kendrick has surprisingly been pretty average. He has a .289/.341/.474 line this postseason, summed up by his 97 wRC+. He comparatively hit much better in the NLCS — posting a .333/.412/.600 line in 17 PA — but only three of his hits during the entirety of the series improved the Nationals’ World Series odds by more than one percentage point. In all, while Kendrick was certainly a star during the NLCS, the Nationals’ title odds barely shifted with each subsequent victory. They certainly did increase, but in a marginally smaller manner with each passing game. The Nationals’ odds actually decreased with the Astros going up 2-1, even as Washington officially clinched the pennant:

Nationals’ World Series Odds
Date Win World Series Marginal Increase
Before Game 1 21.8%
Before Game 2 26.3% 4.5%
Before Game 3 30.7% 4.4%
Before Game 4 31.8% 1.1%
Today 31.3% -0.5%

Thus, while Kendrick certainly had a successful NLCS, it’s hard to argue that he’s been the most impactful performer during their title run.

If you are inclined to be more creative, you could argue that the Nationals received a perhaps-unexpected significant contribution from Aníbal Sánchez in Game 1 of the NLCS, and I wouldn’t disagree with you. Neither would cWPA — he’s fifth on the leaderboard, after all. But still, it’s not Sánchez that deserves the majority of the credit on this squad. We all know this. It’s the Nationals’ big three — superstars Juan Soto, Anthony Rendon, and Max Scherzer — that are primarily responsible.

No one is talking about how Rendon has hit .375/.465/.594 this postseason; he’s third in OPS among the 34 batters with at least 20 at-bats (mostly limiting us to those batters playing in the LCS). Rendon unsurprisingly leads the 2019 Nationals in cWPA as a result, not as much due to the big hits, but due to the volume of hits.

On the flip side of the coin is Soto. Despite having two of the three most crucial hits this postseason, Soto’s .237/.326/.421 slash line and .312 wOBA would actually rank as one of the worst 10-game stretches of his season; doing the math, it would be in the bottom 20% of all 141 10-game stretches he had in 2019. But no Nationals fan would trade in a better overall performance for those two major hits. In the postseason, each individual event, those “clutch” plays, matter much more. While clutch hitters don’t exist in the way that we tend to think they do, clutch moments still do. Soto’s clutch moments have mattered so significantly that his general poor play has been outweighed; Soto already ranks sixth in Expos/Nationals history in total cWPA, and he could easily surpass Tim Raines for the top spot the longer this run continues.

Then, on the pitching side, there’s Scherzer. He’s pitched to a 1.80 ERA this postseason and is fourth in cWPA. All of his top four events this postseason came in Game 2 of the NLCS, when he threw seven shutout innings in a 1-0 game to take the series to 2-0. While Scherzer has certainly been great, he hasn’t been the Nationals’ best pitcher this postseason. By ERA, FIP, and xFIP, Strasburg has been better, but Scherzer’s outs have been collectively more important. cWPA doesn’t care about your FIP — it only cares about the context in which you are making your outs. Because Strasburg surrendered a 3-0 lead in the first inning of Game 5 of the NLDS, his cWPA suffered.

There’s a question of objectiveness to be had here. Is cWPA really fair? In short, the answer is no, it’s not. Context-dependent stats never are. If Soto wasn’t up to face Hader with the game on the line, he wouldn’t be the recipient of the cWPA bump. Someone else would get it, whether that’d be Hader or a different Nationals batter. But in the postseason, context is arguably all that matters. Barry Bonds had a 1.559 OPS in the 2002 postseason, hitting .356/.581/.978 over 74 PA. The Giants lost the World Series not because of anything Bonds did (or didn’t do). It was due to a three-run double by Garret Anderson — who posted a 102 wRC+ that postseason, 157 points behind Bonds — in Game 7 that they did. And, to add insult to injury, Bonds still led all hitters in cWPA that year.

The postseason is all about moments. For the Nationals, it’s been their superstars that are stepping up. And that’s exactly what they could have wanted.





Devan Fink is a Contributor at FanGraphs. You can follow him on Twitter @DevanFink.

10 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
dwm8amember
4 years ago

Where is Strasburg? I’m somewhat surprised to see Scherzer on the cWPA leaderboard but not Strasburg.

Dave T
4 years ago
Reply to  dwm8a

He’s 26th this postseason on the list at the Baseball Gauge, with a cWPA very close to 0 (0.033). Fangraphs shows Strasburg with negative WPA in Game 5 of the NLDS, which I assume is why his cWPA this postseason isn’t that impressive.

dwm8amember
4 years ago
Reply to  Dave T

Scherzer and Strasburg each pitched in two winner-take-all games – the WC and G5 of the NLDS. Their starts were close to identical – Scherzer gave up three runs in 5IP in the WC and Strasburg gave up three runs in 6IP in G5. Their relief appearances in each game were good, but I would expect Strasburg’s was better in terms of WPA (3IP, 0 runs vs. 1IP, 0 runs). If the WC and G5 games were the only ones in the sample, then you’d expect Strasburg’s cWPA to be higher – unless cWPA treats the WC and G5 quite differently in terms of adding to the probability of winning a championship.

Each has made two other starts. For Strasburg, he started G2 of the NLDS when he gave up one earned run in 6IP, and G3 of the NLCS when he gave up zero earned runs in 7IP. For Scherzer, he started G4 of the NLDS when he gave up one earned run in 7IP, and G2 of the NLCS when he gave up zero earned runs in 7IP.

These are exceptionally similar results. I can’t believe that G2 and G3 of an NLCS that was ultimately a sweep have meaningfully different effects on cWPA. So the difference has to be that Strasburg’s “good” NLDS start came in G2 whereas Max’s “good” NLDS start came in G4, when the Nats were facing elimination. Yet, in both scenarios, the Nats were down by one game in the series. It still seems curious that cWPA would value Scherzer’s performance in NLDS G4 so much more than Strasburg’s in NLDS G2.

HappyFunBallmember
4 years ago
Reply to  dwm8a

So the difference has to be that Strasburg’s “good” NLDS start came in G2 whereas Max’s “good” NLDS start came in G4, when the Nats were facing elimination.

Facing elimination is an important part of context. In G2 you still have the potential of three more games serving to dampen the swings of probability.

dwm8amember
4 years ago
Reply to  HappyFunBall

Agreed. I guess I’m surprised it makes *that much* difference.