Next Thursday night, we’re going to find out who won the biggest individual honors in baseball: the Most Valuable Player awards for both the American and National Leagues, as determined by the august and esteemed voters of the Baseball Writers’ Association of America. (Cue trumpet fanfare.)
MVP awards memorialize great individual performances and bestow immense historical significance upon the players who earn them. This is the kind of thing Hall of Fame cases are built on. So you’d think the entire baseball-watching public would be glued to MLB Network or refreshing the BBWAA website on Thursday evening. But… maybe not. Aaron Judge and Shohei Ohtani are almost certain to win, and I guess it’s worth checking social media after dinner just to make sure.
There’s surprisingly little drama over awards these days; postseason betting odds on the MVP races were a little hard to come by, as most bookies have taken the issue off the board. But consider this as a measure of public sentiment: In early October, BetMGM had Judge as a 1-to-50 favorite in the AL, and Ohtani as a 1-to-100 favorite in the NL. Despite a spirited contrarian push by the pro-Fancisco Lindor camp late in the season, it’s all over but the shouting.
Blowouts have become increasingly common in MVP voting in recent years. In the 21st century, the 48 MVP races have produced just seven unanimous winners. Three of those came over the past three seasons, including two in 2023.
Unanimous MVPs of the 21st Century
SOURCE: Baseball-Reference
That had never happened before — two unanimous MVPs in the same season. For what it’s worth, it should’ve happened in 1967 as well. Orlando Cepeda was the unanimous NL MVP, and Carl Yastrzemski won the Triple Crown and led the AL in OBP, SLG, and runs scored for good measure as he hauled the Red Sox to victory in one of the most thrilling pennant races in baseball history.
Yastrzemski’s 1967 campaign is one of the most famous individual seasons ever, and deservedly so. But it got only 19 of first-place 20 votes, because Max Nichols of the Minneapolis Star chose that moment to cast the funniest ballot in BBWAA history. Nichols voted for Twins utilityman Cesar Tovar, who had hit .267/.325/.365 with six home runs that season.
Nichols became a national pariah, which took some doing in the days before cable TV, let alone daily online dogpiles. But his sin was not casting an oddball vote, it was doing so in an MVP race with such an obvious winner. Those occasions were vanishingly rare even well into the 21st century. You’ll notice that breaking the single-season home run record was not enough to make Bonds a unanimous MVP in 2001. And if you think that’s because of any particular animus toward Bonds specifically, Mark McGwire received only two first place votes out of 32 three years earlier when he’d broken the previous record, set by Roger Maris in 1961. (For what it’s worth, Maris won his second straight MVP in 1961, receiving just seven of the 20 first-place votes.)
As both a BBWAA member (though never an MVP voter; maybe someday) and an employee of a publication that helped lead baseball’s empirical revolution, this conundrum brought me to a self-important, though not necessarily inaccurate, hypothesis. A quarter of the way through the 2000s, we are trending — gradually, and intermittently — toward consensus in MVP voting.
The causal mechanism in this theory is the fact that information, at least in a baseball context, is orders of magnitude more plentiful and useful than it was a generation ago.
Consider Nichols’ infamous vote for Tovar. “I suppose it’s a matter of the meaning of most valuable,” he explained after his identity had been revealed. Nichols cited several reasons for voting the way he did, including that Tovar had, owing to two Twins ties, set a new American League record by playing 164 games. Tovar “won games at six positions,” said Nichols, and played exceptional defense across the field.
The Tovar vote, seen as a disgrace even at the time, actually represented a pretty common issue among MVP selectors well into the 2000s. At the time, there was no data to speak of apart from what’s in the box scores. Anything not represented there was left to the imagination and the writers’ own insight and recollection. Before the internet, it was hard for voters to read each others’ analysis to catch up on games they hadn’t witnessed personally, let alone watch tape on players they might have to vote on after seeing only a handful of their games.
Nichols, a Minnesota writer, battled accusations of homerism for bucking the conventional wisdom and voting for a Twins player. Regardless of whether those charges were accurate, the things Nichols found compelling about Tovar, to invoke a cliché, didn’t show up in the box score. You’d have to watch the Twins day in and day out in order to pick up on them, and that simply wasn’t something you could do in 1967 and watch Yastrzemski at the same time. Nobody got the full picture.
If you’ve ever looked at clouds, or taken an ink-blot test, or done a dot-to-dot drawing, you know what happens when the full picture isn’t available to be seen. Imagination and conjecture fill the gaps.
Nowadays, we have numerous ways to evaluate the defense of a Twins utilityman. And even if you don’t trust the overall numbers, you could go back and look at more granular data or set up a Baseball Savant search to watch every single catch he made across the season. More importantly, we know that playing good defense at six positions is all well and good, but when compared to leading the league by a huge margin in every meaningful offensive category, it’s of trivial importance.
A lot of the voting norms we deride now as antiquated — picking the best player from a winning team, voting for whoever had the most RBI, overvaluing defense in the absence of hard evidence — were simply ways to make the picture make sense.
Today, there’s no need to compensate for lack of information, or to let heuristics swamp the historical record. We know how runs are scored and wins are conjured, in exacting detail. We know how to account for defense, and differences in position, and we know that one-inning relievers aren’t as valuable as people thought they were in the 1980s. (The fact that three closers won MVP awards between 1981 and 1992 is overpowering evidence of how good recreational drugs were back then.)
In 2007, the BBWAA moved for the first time to include writers from online publications. Most of those admitted were newspaper veterans who’d moved to new jobs, but a year later, a number of Baseball Prospectus writers got the call as well. This was part of the process of mainstreaming writers, and eventually voters, who came up not as reporters but as specific experts in empirical baseball analysis.
It’s indisputable that baseball writers know more about the sport now than they did 20 years ago. And it’s not really because the old heads let the bloggers and nerds into the clubhouse. Everyone knows more about baseball now than they did 20 years ago. There are MVP voters who can’t do basic arithmetic, but they all talk to sources with teams or friends in the press box who think in spreadsheets. More to the point, their audience has access to all that information too and won’t settle for “César Tovar plays six positions” when wRC+ shows Yastrzemski ahead 194 to 102, and ahead in WAR 11.1 to 2.1.
Numbers don’t make baseball harder to understand. They make it harder to pass off bunk as legitimate analysis. Regardless of background, the modern baseball writer has to be au fait with numbers now.
As a result, every voter comes to awards season with similar information and a similar conception of what the award means. Small wonder that they’re coming to similar conclusions.
But how would you measure that? Well, in addition to the historical spike in unanimous MVP choices, we’re seeing winners take home a greater proportion of first-place votes even in normal years. (As an aside: The BBWAA electorate comprises two writers for each chapter represented in the league. Until the Astros switched from the NL to AL in 2013, that meant there were 32 NL MVP voters and only 28 AL MVP voters each year. All historical vote comparisons in this article have therefore been adjusted for a 30-member electorate.)
In the last century, there was rarely consensus about which player deserved the MVP; in fact, there was frequently a great deal of confusion over who was even in the conversation. MLB has taken to announcing the three “finalists” for each award — to no small amount of derision, as the votes have already been tabulated, so this is in fact the award podium and not a list of “finalists” in the traditional sense. But it wasn’t that long ago that such an announcement would’ve been genuinely revelatory.
Since 2012, there’s only been one occasion on which more than six players across both leagues received a first-place MVP vote. But as recently as 2003, there were 11 first-place vote-getters. In 1999, Ivan Rodriguez was named AL MVP in a squeaker over Pedro Martinez, who actually got eight first-place votes to Pudge’s seven. Even with only 28 voters, six players got at least one first-place vote and five got at least four votes. The gap in total points between Rodriguez and seventh-place Nomar Garciaparra was just 115 points, or less than the difference between first and second place in the National League that season.
Now that was a debate worthy of the name. We still have MVP debates today, with defensible cases for multiple players. But it’s usually two or three candidates, rather than six or seven, and when the votes are tallied, the results usually don’t end up being that close.
A contentious Judge-Ohtani race in 2022 — the “62 home runs” vs. “come on, it’s Shohei Ohtani” year — ended with 28 first-place votes for Judge and just two for Ohtani, and a 130-point winning margin. In 2017, Judge ended up on the other end of a blowout that I thought was going to be a squeaker, as Jose Altuve boat raced him by 27 first-place votes to two. A contentious NL MVP race in 2021, featuring five first-place vote recipients but no slam-dunk candidate, ended with Bryce Harper beating Juan Soto 348 points to 274, with a 17-6 margin in first-place votes.
Since I joined the BBWAA in 2018, the only truly nail-biting MVP race was the 2019 AL MVP contest, in which an injured Mike Trout held off a late charge from Alex Bregman, 355 to 335 on points and 17 to 13 in first-place votes.
The membership and ideological outlook of the electorate might change over time, but the voters are also beholden to the events of the season before them. Actually, that’s backward. Voters have to consider the season before them, regardless of their individual or collective opinions on the concept of “most valuable” and how to measure and express it.
Sometimes, there’s an obvious runaway winner who puts up bonkers numbers on a good team. Sometimes not, and the electorate is left to choose from a handful of guys who seem like good downballot options — but are we sure this was the best player in the league this year? In those cases, there’s obviously going to be some difference of opinion.
In the American League, even with variance from year to year, the winner tends to end up with more points now than at the turn of the century.
But in the National League, there’s no obvious trend toward consensus at the top of the ballot.
That probably has something to do with Bonds and Pujols dominating the National League in the aughts, while chaos reigned in the American League. Voters cast their ballots based on what they see in front of them, and sometimes there’s just an obvious answer.
So maybe it’s not that everyone thinks the same, but that outlier candidates beget outlier vote totals, and any additional consolidation is merely marginal.
Let’s consider another hypothesis, then: Sabermetrics didn’t ruin the MVP debate — Ohtani and Judge did.
