Tuesday, November 22, 2011
NL MVP vote reassuring or exasperating?
By Christina Kahrl
The expectation was that the NL MVP would be a two-man race, and the Baseball Writers' Association of America delivered exactly that, with left fielder Ryan Braun of the Milwaukee Brewers coming out decisively ahead of Dodgers center fielder Matt Kemp. Braun got 20 first-place votes and finished first or second on all 32 ballots. Kemp received half as many first-place votes, earned 16 second-place votes and wound up 56 points behind.
We’ll get into how much this is simultaneously good and bad news in a minute, but first things first. This is great for Ryan Braun, and congratulations to him.
Even if Kemp had received first- or second-place votes instead of the six third-place votes, the only bragging right he’d have achieved would have been losing by 40. So there isn’t much to rail about as far as Prince Fielder's snagging a lone first-place vote and four seconds, or the one first- or second-place vote that wasn’t among those three going to Justin Upton. Voting for Upton (or Fielder, to a lesser extent) over Kemp and Braun was weird, but it wasn’t decisive.
The good news, if you’re of an analytical bent, is that the right two guys showed up first or second on all but six of 64 potential ballot positions. The narrow split between Braun and Kemp in terms of their unexamined hitting statistics was a key factor in the equally narrow result. Braun led by just eight points of OPS, as they finished 1-2 in slugging and total bases and 4-5 in OBP (to Kemp’s narrow advantage, .399 to .397).
If there’s a surprise, it’s that Kemp led in some stats you’d normally associate with a victory via BBWAA voters, topping the circuit in home runs and RBIs and even tying for second in stolen bases. Kemp played the more important defensive position, and whatever fielding metric you want to turn to, he was at least adequate, where Braun’s play might be most favorably described as strong-armed or dramatic. From the analysis side of things, Kemp came out handily ahead, winning WAR and adjusted OPS+ and batting wins and batting runs and equivalent runs and true runs. (Not to mention just plain old runs.)
For all those category wins, Kemp may as well have won top honors at Cannes, Sundance and the People’s Choice Awards. It still isn't an Oscar, and the frustration among statheads is that this reflects a systemic problem with the voters and with the very thing being voted upon.
As Braun already has said generously, the biggest reason (he felt) he won was that he was the guy on a winning team. That might seem like a problem, with the award or with the voters, but the stated criteria for what goes into an MVP award have been broad enough in defining the value in “valuable” that the electorate has been able to go where it will. For more than a few writers, as with a lot of fans, there’s a readiness to associate team success with individual value.
That’s unfortunate, but the muddle on the sabermetric side of the debate over what constitutes value probably doesn’t help matters any. But in this instance there wasn’t any disagreement: Kemp and Braun were the two most valuable players in the league in terms of wins above replacement, whichever brand you’re using, Baseball-Reference or FanGraphs. (Baseball Prospectus’ BWARP had Braun third behind Joey Votto, that bit of statistical iconoclasm had zero traction with the voters -- he didn’t get a vote higher than third place.)
So, although a natural reaction among my fellow statheads is that this is further proof that my other colleagues within the BBWAA don’t get metrics, how fair is that? Collective guilt’s a bit of a slippery subject, though, and 10 voters gave Kemp their first-place votes.
Perhaps some of the voters don’t get the relative importance of a hitter who plays a decent center field over a broadly similar hitter who plays an adequate left field. Perhaps a few others are reliably intent in associating the “valuable” part of being MVP with team success. Without a standard definition, you can sort of see where people voting that way are coming from. For better and for worse, this isn’t an award handed out for statistical achievement alone.
If anything, the fact that Kemp and Braun finished so close in the balloting reflects that there was at least consensus that these were the two most valuable players in the league, and that from among the 32 electors who had this year’s vote, a decisive majority favored Braun. However, if you wound up with another 32 electors from across the 16 National League markets, I think it’s fairly likely you’d find the exact reverse of this outcome and wind up with 20 first-place votes for Kemp. If the BBWAA’s franchise were a full instead of a selective democracy, we might have seen a different outcome, which is significant in terms of who’s got hardware, if perhaps less so in terms of confirming a consensus that Braun and Kemp were the league’s two most valuable players. That’s admittedly speculative, and it isn’t like expanding the franchise, even within the BBWAA, is even remotely in the cards.
Down-ballot, there are always the interesting tidbits to pick out. Which two voters dropped Albert Pujols from their ballots? Which four forgot about Votto? Why did Miguel Montero show up on only one ballot? Who was putting John Axford on any ballot anywhere, as three people did? It’s fun to kick around a little bit right now and perhaps for the next 15 minutes, but it’s another way of putting the trivial in MVP voting trivia.
The other phenomenon that I find fascinating is that on the real ballots as from the SweetSpot network’s caucus, the Phillies’ Roy Halladay drew more support from the voters than his rival for this year’s NL Cy Young, Clayton Kershaw. Kershaw won the hardware, but given Halladay’s superiority over Kershaw in a whole passel of interpretive performance metrics, my quick inference would be that there are more than a few statheads among the electorate.
In the end, the result is dissatisfying to those of us with an investment in seeing these outcomes reflect what the best analysis already suggests. The process might be slowly trending in that direction, and that might reflect the changing demographics of the baseball writers and their ages. However, recent votes and revealed ballots have made it clear that there are plenty of “old guys” in the organization who’ve been receptive to modern performance-evaluation tools and have used them to help inform their votes. As much as Kemp’s case was strongest, the best thing those of us who use sabermetrics can do is deal with the result by continuing to try to elevate the conversation with facts and data. Giving up on the process just guarantees that you won’t win an argument you can’t even have.