Evaluating the BBWAA MVP voting through the years

A few months ago, I examined how the Baseball Writers Association of America did when voting for the Cy Young. As it turned out, the voting was pretty bad in the latter years but has been improving over the past few years.

Today, we’re going to look at how the BBWAA has voted for the other major seasonal award: MVP.

To many people’s surprise, the 2016 Most Valuable Player of the year award went to Mike Trout. This isn’t sarcasm. Trout had been the best player in baseball for the past few years, and still is the best player in baseball. But the writers, for many years, have come up with a number of justifications for voting against Trout. These excuses range from, “he doesn’t play on a winning team”, to trying to subjectively define the term valuable, to actually thinking (somehow) that Trout isn’t the best player in game.

Many of these arguments are flawed. For example, there’s nothing within the BBWAA guidelines that stipulate that the MVP must come from a winning team.

The fact that Trout hasn’t won more MVP awards since he’s been in the league has gotten a lot of attention, especially from the sabermetric crowd. To many, it’s been rather obvious that Trout should have gotten the award, and the fact that he hasn’t won it has caused sabermetric thinkers to criticize the BBWAA.

I, therefore, was interested in examining how the BBWAA has done in the past and compared to the present when voting for the award.

In order to try and quantify this, I looked at the WARP values for every MVP award winner since 1950, and compared it to the WARP leader that season. The reason I went back to 1950 was because that’s how far back Baseball Prospectus’ WARP values go.

While there has been a lot of debate about whether pitchers deserve to win the MVP award, there have been pitchers who have won the award, which is why I decided to include pitchers in my sample. This can cause some complications when comparing pitchers’ WARP values to hitters’ WARP values, but I believe that this is currently the only way of providing any type of comparison between both positions. And because pitchers have won the award in the past, this problem was unavoidable.

Therefore, these values shouldn’t be taken as absolute. WARP isn’t perfect, and there is uncertainty with these numbers, but this doesn’t mean that they should not be used, or are “inaccurate.” It means that if an MVP winner had a WARP of 8.5 and the leader had a WARP of 9 then the BBWAA wasn’t necessarily “wrong.” That said, if there’s a four WARP difference, then we can probably say the writers were wrong.

First, let’s examine the American League, since that’s where most of the controversy has been at for the past few years.

On average, there has been a 2.15-win difference between the WARP leader and the WARP of the MVP winner.

Since 1950, the WARP of the MVP and the WARP leader have been the same only 15 times (22 percent). That’s lower than the Cy Young award’s rate since 1950, where the WARP of the winner and the WARP of the best pitcher have been the same 33 percent of the time.

Unlike the Cy Young award, there isn’t a trend showing that the BBWAA is getting better at voting. There’s a lot of randomness, and the voting seems to vary greatly from year to year.

Mike Trout also isn’t the only generational player to be robbed of the MVP award. Just look at Mickey Mantle. Mantle, lead the American League in WARP for five straight years, and seven out of six seasons. Yet during that time, he only won the award two times. It took two of the greatest seasons of all time for him to be recognized. Basically, unless it was blatantly obvious, the award did not go to Mantle. But even that may be too kind to the writers, as in 1961, Mantle had a WARP of 11.3 and yet he still wasn’t given the award. Instead, the award went to Roger Maris, who had a good season with a 7.4 WARP but didn’t belong in the same conversation as Mantle.

Mantle actually won the award three times, though. His third time was in 1962, but he wasn’t even the best player that season. It really forces you to ask: what were the writers looking for?

We are currently seeing something similar with Mike Trout. Trout, just as Mantle did, led his league five years in a row in WARP. Since 1950, they are the only two players to have done that. (No one who played in the National League, not even Barry Bonds, can claim that fun fact.)

Many BBWAA voters are also torn as to whether pitchers should win the MVP. That uncertainty, along with plenty of bad judgment, has also caused some questionable votes. For example, the “worst” MVP winner was Rollie Fingers, a reliever who had a WARP of 2.2 the year he won.

But while the BBWAA would vote for the subpar performance of Fingers, they for some reason would not vote for Randy Johnson in 1995 or Pedro Martinez in either 1999 or 2000, who put up some of the greatest seasons ever by pitchers.

Now, let’s examine the National League.

In the National League, the average difference is 1.8 wins. The MVP winner WARP and the WARP leader have only been the same 26 percent of the time.

While in recent years, the voting seems to have gotten better, when we look at the larger trend, it isn’t easily discernible whether the voting has gotten that much better. The reality is that there’s a lot of variance from year to year.

The biggest difference in WARP was between Mike Schmidt and Steve Garvey in 1974. That season, Schmidt put up an 11.6 WARP, and Steve Garvey had a 3.9 WARP. Schmidt was better than Garvey at just about every aspect of the game, apart from base running. But, the Phillies missed the playoffs that year, and the Dodgers lost in the World Series. I don’t know if that was the main reason for Garvey’s MVP award, but any other justification would have been hard to make. And while WARP values shouldn’t be taken as absolutes, this is one of those situations where the voters were flat out wrong.

One where it’s a bit murky is with Barry Bonds’ 2002 MVP. That season, Bonds put up an 11.5 WARP, which in most season, means you should win the MVP by a landslide. But that season, Curt Schilling finished with an equally impressive 11.8 WARP. This is a perfect example of the uncertainty in WARP. While Schillings WARP value was higher, it would be unfair to suggest that the BBWAA got it wrong.

Barry Bonds and Willie Mays are also the only two players since 1950 to lead the league in WARP seven times. The only difference is that Bonds won seven MVP awards, and Mays only won two.

While the BBWAA has been doing better in their voting for the CY Young award, there’s less evidence that this is actually happening with the MVP voting.

The MVP award is more complicated, but that shouldn’t excuse the voters for not recognizing some of the game's greats, such as Mays, Mantle, and now Trout. The BBWAA is also going to have to come to a decision regarding pitchers, because on a number of occasions, pitchers have been the best players.

The problem with arguments over “value” is how much subjectivity is built into them. They’re usually made to support one’s own biases, not to counteract them.

They also lead to all-time greats such as Mantle and Mays not being properly recognized. If this sort of flawed reasoning keeps happening, then Mike Trout might also end up historically under-celebrated.