The 2021 NL Cy Young Race Shows the Great Divide in Pitcher Evaluation

With six weeks left in the regular season, the NL Cy Young race is wide open. All three of Walker Buehler, Zack Wheeler, and Corbin Burnes are building cases that could seem credible or not, depending entirely on your methodological viewpoint. And they are far from the only serious candidates. (Nothing would make me happier than seeing Wade Miley rip off three straight shutouts in September to take home the trophy. Kegger’s at my place if he does it.)

The thing that makes award season fun (or irritating, depending on how quickly you tire of arguing) is that there are multiple ways to approach the question of who is each league’s most outstanding pitcher. Specifically: How should voters balance rate stats versus volume of innings? And how should they prioritize results (ERA, wins, and so on) versus underlying numbers?

From one standpoint, Buehler is having the kind of season that usually gets a pitcher into the Cy Young conversation. He leads all NL starters in ERA (2.09), his 12-2 record gives him the best winning percentage of the award contenders, and he’s on pace to pitch nearly 210 innings. Those results-based stats paint a pretty solid Cy Young case, and they certainly stack up well against the numbers of another high-volume candidate: Wheeler. Here’s where the two rank among qualified NL starters:

Buehler has the edge in most of those categories, and he’s pretty close behind in the two he doesn’t. But one traditional stat where Wheeler is outperforming Buehler by a fair margin is strikeouts. Wheeler has 187 to Buehler’s 162, which invites investigation into another side of the methodological coin. We know Buehler is allowing fewer runs in a similar volume of innings, but how much of that edge is the result of Buehler actually pitching better than Wheeler, and how much is the result of extraneous factors?

Buehler’s results make him look like a clear Cy Young favorite: He’s allowing far fewer runs than Wheeler (both his ERA and RA/9 are about 18 percent lower than the Philly ace’s). But Wheeler has the edge in underlying numbers: He’s missing more bats, and what contact he does allow is softer.

Wheeler plays his home games in a more hitter-friendly park than Buehler, and the defense behind him is worse, so it shouldn’t be surprising that he’s getting worse results from better peripherals. But you’ll notice that while Wheeler has a massive advantage over Buehler a step higher in the causal chain, he doesn’t actually lead the league in any of those categories. Corbin Burnes does. Here are Burnes’s numbers, both results-based and underlying:

Burnes basically fights Buehler to a draw in run prevention, and he has as big a gap in his underlying numbers over Wheeler as Wheeler has over Buehler. FIP is a little simplistic, and it doesn’t reflect all we now know that pitchers can control in a game, but the 1.59 number is still eye-popping. Mariano Rivera never posted a FIP that low in any season as a one-inning reliever, and Burnes is doing it while having to turn the lineup over.

Even so, there’s one dent in Burnes’s Cy Young case, and it’s a big one. Between a brief IL stint and Wheeler generally pitching deeper into games, Burnes has thrown 28 fewer innings than Wheeler. That’s a huge gap, equivalent to a quarter of a season so far or about a month’s worth of starts. And whether Burnes’s better per-inning numbers make up for that depends on who you ask. Baseball Prospectus’s WARP has Wheeler three-tenths of a win ahead of Burnes, while FanGraphs has Burnes ahead by a similar margin.

In some respects, that’s a reflection of the changing role of the starting pitcher. Wheeler has dutifully taken his place on the mound each time through the rotation this year and thrown almost seven innings a start. But that type of workload is now a rarity for MLB starters. A decade ago, 93 pitchers threw at least one inning per team game, the qualification threshold for the ERA title. By 2017, that number was down to 58. So far this year, there are just 45 pitchers who’ve thrown that often.

Now the question is: Should Burnes be rewarded as the apotheosis of the modern starting pitcher? Or should Buehler or Wheeler be rewarded for embodying the only Dylan Thomas poem anyone’s read by raging against the dying of the light?

Award arguments aren’t really about the players. They’re about values. From blinkered partisans to quantitative hard-liners to narrative-obsessed clubhouse insiders, everyone who advocates for an award candidate is trying to sell you a way of looking at or thinking about the sport. The actual player on the ballot is merely an avatar for a system of evaluation.

This became clearer than ever in the late 2000s, when sabermetrics upset the established quantitative balance of the sport. WAR, K%, WPA, FIP, and a collection of initialisms worthy of the federal government came in and displaced the back-of-the-baseball-card stats in award voting primacy. But while the war to get to that place was long and waged on many fronts, the 2010 AL Cy Young vote was its Saratoga.

That year, Félix Hernández—who threw 249 stellar innings for an absolutely hopeless 101-loss Mariners team—won the AL Cy Young Award with a 13-12 record. And he did so decisively, receiving 21 of 28 first-place votes. The vote killed the idea that wins are an important indicator of a pitcher’s quality—but wins wouldn’t have persisted for so long as an indicator of value if the two things weren’t tied so closely together.

In fact, in an amusing historical footnote, the correlation between wins and success in Cy Young voting is actually greater now than it was before Hernández won his watershed award. From the start of the two–Cy Young era in 1967 until 2009, 57 percent of awards went to pitchers who at least shared the league lead in wins. From 2010 to 2020, 68 percent of Cy Young Awards went to wins leaders, and at least one of the two trophies has gone to a league leader in wins 11 years running.

Still, the practical fineness of that distinction hasn’t mitigated the ferocity of the fight; as recently as September 2018, Ken Rosenthal wrote a column about that year’s AL Cy Young race and said: “Those who want to keep calling the writers Neanderthals and worse for their voting patterns, feel free. But when it comes to the value of wins, let me make this very clear: YOU’VE WON.”

In that same column, however, Rosenthal also laid out a definitive treatise on one of the two new Cy Young voting fault lines: quantity versus quality. Rosenthal quoted extensively from one of that year’s two front-runners, Justin Verlander, who believed he deserved the Cy Young because he’d thrown substantially more innings than his rival, Tampa Bay’s Blake Snell. Snell, who led the AL in ERA and—hilariously—wins, would throw some 33 1/3 fewer innings than Verlander by season’s end and become the first starter to win the Cy Young in a full season without at least coming close to 200 innings pitched. Verlander struck out more batters per inning than Snell and walked half as many batters per inning, but had an ERA six-tenths of a run higher.

That battle speaks to the second fault line in modern Cy Young voting: results versus peripherals. Getting more granular than wins and ERA was only the first step in baseball’s tumble into empiricism. Wins are the product of run prevention (runs allowed and ERA), which is the product of keeping runners off base. That in turn is a function of missing bats and inducing weak contact, which comes from command, sequencing, velocity, and movement. All this has been common knowledge in baseball coaching circles for more than 100 years, but in the past decade and a half, we’ve gained the ability to quantify those aspects of the game.

So which is more important, results or peripherals? Do we reward pitchers for excelling in the areas they can control, or for achieving what they set out to accomplish—recording outs and, yes, winning games? This divide has no clear empirically based choice, and no consensus among the sabermetric community. Of the three major computations of WAR, for example, Baseball-Reference uses runs allowed as the baseline for pitchers, while FanGraphs and Baseball Prospectus use ERA estimators.

What makes this NL Cy Young race particularly interesting, then, is the clarity of the fault lines between quality and quantity and results and underlying stats. There’s a candidate for every worldview—plus a Goldilocks contender in Brandon Woodruff, who’s close to Burnes on the underlying numbers front and isn’t scraping quite as close to the minimum threshold for qualifying for the ERA title.

But Woodruff isn’t leading the league in any significant statistical category; he’s just in the proverbial ballpark, like a human dialectical synthesis. Without a number or a narrative to hang his candidacy on, he’s unlikely to attract enough first-place votes to win even if Buehler, Burnes, and Wheeler divide opinion along various methodological and ideological lines.

At this point, it probably would take something like a September reign of terror from Miley to clarify the Cy Young race. Because right now, there are at least three front-runners for the award, depending on who you ask.