WASHINGTON — This year’s races for MLB’s Most Valuable Player and Cy Young awards could be some of the most hotly-contested in years. That makes Wednesday and Thursday night exciting for baseball fans … and terrifying for me.
I developed this model for predicting the results of each league’s races two years ago, and went 4-for-4. Last year, things got a bit messier. This year will be a real test whether or not my methodology is worth anything in 2017, or if changing ideas about which numbers are important in the game are creating new norms when it comes to awards voting.
Here’s a quick reminder (or crash course, for the newbies) on how this all works. These predictions come from a model developed to try to anticipate how awards voters make up their minds, using a wide-ranging combination of statistical rankings. They are not personal opinions — they are purely the interpretations of data. The model has no biases, other than the ones intentionally programmed into it to replicate human bias. But enough … on to the predictions.
(Photo by Ezra Shaw/Getty Images)
Photo by Ezra Shaw/Getty Images
American League Cy Young: Chris Sale
Both American League awards were tight this year, and the Cy Young was a neck-and-neck, two-horse race. Corey Kluber led the AL in wins, ERA, WHIP and bWAR, the kind of statistical dominance that might mean an easy win in most years. But Chris Sale led the league in strikeouts, K/9, FIP and fWAR, all while throwing the most innings. At the end of all the model’s basic calculations … the race was a tie.
Yep, really. In these cases, I look for something historic to tip the scales. The last time an American League pitcher struck out more than 300 batters in a season and didn’t win the Cy Young was Randy Johnson’s 1993 season, in which he (and Kevin Appier!) was robbed by Jack McDowell. Sale has finished in the top six in the AL Cy Young each of the past five years. He posted career bests in strikeouts, FIP, K/9, and K/BB, in his first year pitching in the daunting AL East. This one should be incredibly close, but the model says Sale.
(Photo by Maddie Meyer/Getty Images)
Photo by Maddie Meyer/Getty Images
American League MVP: Aaron Judge
Likewise, the AL MVP race was a classic, and the process of trying to determine a winner was compounded by a few things that we have no good historical accounting to square up, insofar as the model is concerned. Jose Altuve led the American League in WAR, per Baseball Reference, with Aaron Judge second — Judge was tops in FanGraphs WAR (with Altuve second). Judge led the league in home runs and runs scored, while Altuve won the batting title. The Astros won 10 more games, but both teams made the playoffs, for those voters that interpret “value” through more than just an individual’s isolated numbers.
Plus, Judge is a rookie. The only rookies to ever double up on Rookie of the Year and the MVP are Fred Lynn and Ichiro. In Ichiro’s case, he was the Altuve-type player, a high-average and speed player with just enough pop, edging out statistically superior veteran slugger Jason Giambi (9.2 bWAR to 7.7 for Ichiro) in one of the closest MVP votes ever. The heart says Altuve wins. The model says Judge.
(Photo by Mike Ehrmann/Getty Images)
Photo by Mike Ehrmann/Getty Images
National League Cy Young: Max Scherzer
Clayton Kershaw certainly has a case for NL Cy Young. He led the league in ERA and wins and still topped 200 strikeouts, despite making just 27 starts. His 6.73 K/BB rate was actually a tick worse than the last few years, but still better than anyone else in the league. If he won, nobody would complain.
But here’s the thing — those missed starts really may be the difference. Max Scherzer, who ran away with the award last year, actually pitched better this season. He set career marks in K/9 (12.0) and WHIP (0.902), allowing a miniscule 5.7 hits per nine innings. His 268 strikeouts led the NL. And, most significantly, his WAR is best in the league no matter which measure you use, and by a wide margin (7.3 bWAR to 4.6 for Kershaw, 6.0 to 4.6 fWAR). This almost feels like it might come down to the predominant type of voter — old school or new school — assigned to choose. The model aims to find the correct balance, but picked Justin Verlander over Rick Porcello last year. The model says Scherzer.
(Photo by Mike McGinnis/Getty Images)
Photo by Mike McGinnis/Getty Images
National League MVP: Giancarlo Stanton
Plenty of Nationals had good years, but none even cracked the top five of the model (Bryce Harper was in good position until his injury). In late September, SI’s Tom Verducci picked Nolan Arenado to win … and he didn’t even finish in the top three. Our model had Charlie Blackmon above him, though it also has a Coors Field buffer built in, as Rockies candidates are often discounted for their home field. Ultimately, it won’t matter — neither is among the finalists for MVP.
Which leaves a fairly close race, but one led by Giancarlo Stanton, according to the model. Paul Goldschmidt put together another fine year, but missed out on ranking in the top three in all but one major offensive category (runs, third). He helped lead the D-backs to the playoffs, but Stanton clearly had the superior year, pacing the NL in home runs, RBI, extra-base hits, slugging, and bWAR. It’s good to see Joey Votto rewarded in spite of the Reds’ awfulness, but he should finish third, with our Colorado friends close on his heels.
(Photo by Mike Ehrmann/Getty Images)
