Anthony Rendon and Ketel Marte have both had very good seasons, but all year long they’ve been a few beats behind Christian Yelich and Cody Bellinger. With only a few days left in the regular season, Marte has already been shut down and Rendon would need a historic week to match the two favorites. While Yelich’s season is unfortunately over, his work already done clearly puts him, along with Bellinger, in the top two for the MVP. With the numbers on offense creating a substantial edge for Yelich, who deserves MVP comes down to a question nobody really likes answering: How much better is Cody Bellinger on defense than Christian Yelich?
It’s possible some voters will ding Yelich for his injuries and playing time lost, though it didn’t stop Mookie Betts and Mike Trout from finishing 1-2 in the AL MVP last year, or keep Josh Hamilton from winning in 2010 when he missed almost all of September. Historically, there haven’t been very many MVPs with around 130 games and 600 PAs, though that likely has less to do with voters being unwilling to vote for players who have missed some of the season and more to do with it just being incredibly difficult to be the best player in the league when other candidates have an extra three weeks to compile numbers.
Since 1931, there have only been 22 position player seasons of at least 7.5 WAR and fewer than 600 plate appearances. Of those 22 seasons, only 10 topped the league in WAR. Six of those 10 players won the MVP that season. In 2009, the under-appreciated Ben Zobrist topped the league in WAR, but Joe Mauer, behind by 0.3 WAR, won the trophy. In 1989, Lonnie Smith played in 134 games and tied Will Clark for the league lead in WAR at 8.1, but it was Clark’s teammate Kevin Mitchell and his 49 homers that took home the MVP. In 1985, Pedro Guerrero topped the NL in WAR at 7.8, but Willie McGee’s 7.1 WAR won him the award thanks in no small part to a .353 batting average. Finally, way back in 1954, Ted Williams topped the AL in WAR, but the Yankees’ Yogi Berra won the vote. Yelich’s situation isn’t unprecedented, but it is fairly rare; players in his position have won MVP around half the time.
As for a tale of the tape, which has been done in the last week at Baseball Prospectus, there’s a seeming consensus that Yelich has been the better offensive player this season, with a small edge at the plate. That gap widens once baserunning is included; here at FanGraphs, Yelich’s edge is 12 runs. At Baseball-Reference, he’s is also up by 12 runs. At Baseball Prospectus, the gap is seven runs with their error bars ranging roughly from Yelich ahead by 1 to 13 runs.
At Baseball Savant, there is no gap between the two players using xwOBA, but that’s prettily easily explainable, as xwOBA doesn’t account for shifts. Bellinger pulls the ball nearly 50% of the time while Yelich is just under 40% on batted balls. Bellinger pulls 70% of his ground balls and nearly 60% of his line drives. The result is that Bellinger fares worse than xwOBA believes he should due to the standard defense against him. Bellinger pulls the ball more to get more hits and home runs overall, and willingly makes the tradeoff to take more outs into the shift. Bellinger’s 75 wRC+ in more than 300 plate appearances against the shift isn’t bad luck on hard hit balls. Yelich has about half the plate appearances, but has a 140 wRC+ against the shift, an indication that shifting against him might not be that useful, and the relative number of plate appearances against the shift opens up more holes for Yelich.
If we chalk up only the difference between xwOBA and wOBA on pulled ground balls and line drives to a more realistic situation, we are back at 12 runs difference between Yelich and Bellinger and in agreement with FanGraphs and Baseball-Reference. Bellinger’s extra playing time nets him two more runs above replacement than Yelich, cutting the deficit to 10 runs. On defense, both players have a positional adjustment of about six runs. So without factoring intangibles or clutch or how much a player meant to a team or where their playoff position would be without them, Bellinger needs to be 10 runs better than Yelich on defense to even the score.
Here at FanGraphs, using UZR, Yelich is plus two on defense compared to Bellinger’s plus 9.5, meaning we have Yelich still up by 2.5 runs, which explains how Yelich’s 7.8 WAR is three-tenths better than Bellinger’s 7.5 mark. That’s not true at Baseball-Reference, where Bellinger is plus 24 based on DRS, while Yelich is neutral. Those 24 runs put Bellinger 14 runs ahead of Yelich and explain his 8.6 Baseball-Reference WAR versus Yelich’s 7.1 WAR. So what are we to do? Splitting the difference puts Bellinger up by a handful of runs, but weights DRS more heavily due to its greater spread of results. Taking three-year averages of the pair of metrics would make the players almost even. We could break the tie with Baseball Prospectus, where the 16-run difference between the two players puts Bellinger less than a handful of runs ahead overall. We can consult Statcast data as well.
Both UZR and DRS have Bellinger’s arm worth about five more runs, so if Bellinger’s range and errors are worth five more runs than Yelich, that would put the two players even. Any larger a gap, and Bellinger will have produced more runs than Yelich this season. Jake Mailhot discussed Bellinger’s defensive game earlier this month. According Outs Above Average, Bellinger is plus-seven and Yelich is minus-four. That’s not a lot of help. Even dialing down and noting that Bellinger was two outs better on balls going back and eight outs better on balls going up, assuming two doubles and eight singles, provides a run value of right around eight runs. Using the traditional defensive metrics for his arm and Statcast for range, Bellinger ends up three runs better overall. Over the course of a six-month season, three runs isn’t a lot to hang your hat on when trying to differentiate between players. One season of defensive metrics isn’t a lot to rely on, either.
When weighing a potential choice for MVP, the decision in the NL essentially comes down to whether Cody Bellinger is more than 10 runs better on defense than Yelich this season. Depending on which defensive metrics you’re using and how you’re using them, relying only on this year or taking a three year average, the difference ranges all the way from six runs (three-year UZR average/1000 innings), to eight runs (single-season UZR), to 11 runs (three-year Statcast or three-year DRS/UZR average), to 13 runs (single-season Statcast), to 16 runs (three-year DRS or single-season UZR/DRS average), to 24 runs (single season DRS). Average all of those together, and we end up at 13 runs. We can manipulate our brains and these numbers to shake out any number of ways, but none of that wrangling seems to provide us with confidence in a definitive answer.
We can do the same mental gymnastics with clutch hitting. Yelich came up to the plate 48 times in high leverage situations, and his 287 wRC+ is the best in the game in those situations and much higher than Bellinger’s 133 wRC+ in 57 PA. In medium-leverage and low-leverage situations, the two players were nearly identical (164 wRC+ for Yelich versus 163 wRC+ for Bellinger), but those high-leverage situations helped Yelich to a two-win gap in WPA that would essentially double the 12-run offensive gap between the two players if you look only at results and how much they impacted individual games. While Yelich likely isn’t an inherently more-clutch baseball player than Bellinger, those are the results on the field. If we believe that single-season DRS might not necessarily represent the true defensive talent of the individual players, but does represent what happened on the field this season, wouldn’t we also consider WPA? I’m not sure we should be doing either.
Any number of narratives apply to both players. Bellinger is the best player on the best team and he’s played all year long. His defense makes him the better all-around player. Yelich’s performance was more integral to whether or not the Brewers make the playoffs. He’s the better offensive player and came through more often in the clutch. Ultimately, these are just narratives to help people feel better about their vote or the candidate they support. The race is too close to call. A vote for either candidate is simultaneously justifiable and difficult to justify.