It happens at the end of every baseball season: the argument over what “valuable” means in the context of the sport’s Most Valuable Player award. I enjoy a good argument, especially over relatively minor things like awards to multimillionaires (or, in the case of very young players, imminent multimillionaires). But I don’t enjoy reading sports columns in which the writer sets himself up as the One True Arbiter, dismissing as a fool anyone who disagrees with his value judgment. There are plenty of examples, such as this one, by Dave Cameron at Fangraphs (a site that has consistently mismeasured the value of a win, btw).
Another example of an off-puttingly high ratio of self-regard to actual merit has been written by NBC Sports columnist Joe Posnanski. Now I’ve got to stipulate up front that Posnanski is a good writer, especially when the subject is his family. But when it comes to the MVP award, Joe’s writing turns into a caricature of internet commentary. His opinion is not only correct, but it’s the only logical opinion. Anyone who disagrees with him is a fool or a knave. Sometimes that’s true, of course–you’ll find plenty of those arguments in the archives of this very blog–but it’s silly to think that considerations of “value” could possibly be inarguable.
There are two distinct parts to the question of value. Essentially, they are quantity and quality. In the case of baseball, the “quantity” aspect is a player’s level of performance. For hitting, this is pretty measurable, although there’s still a split between those who cling to the measures used in the past (particularly batting average and runs batted in) and those currently in vogue such as “offensive WAR” (don’t ask). The measurement of players’ value as fielders is recognized as imprecise by pretty much everyone.* Evaluating defense still relies heavily on direct observation, which means it’s somewhat subjective.
What’s the “quality” aspect of player value? To answer that question, you’ve first got to answer the question, How does a baseball player create value? To his employer, of course, a player’s purpose is to help sell tickets. What sells tickets? One obvious answer is a winning team. There’s plenty of evidence that that’s true, anecdotal and otherwise. During Boston’s legendary “Impossible Dream” 1967 season, for example, attendance more than doubled relative to the previous year. But there’s also the “Ted Williams” answer. Once Ted went off to fight in the Korean War, the Red Sox entered a period of mediocrity that got even worse after he retired. The main reason people bought tickets to Red Sox games after he returned from Korea was to watch Williams hit. Attendance hit rock bottom after he retired (from over 1.1 million in Ted’s last season to 733 thousand two years later). So there’s obvious value in a player’s pure performance level no matter how good or bad his teammates are. The precise balance between these two aspects of value can be debated by reasonable people, but it’s clear that the value of an additional win to a team’s fans is entirely context-dependent. This is a point that is completely missed by Fangraphs.
The Posnanski/Cameron position is either that the value of a win isn’t context-dependent, or that a player’s “value” should be measured without regard to the value of the wins he contributed to his team. (They never specify which position they’re taking, because they never show any awareness of the distinction at all.) I think of the first view as the Gertrude Stein position: A win is a win is a win. And for the same reason that Stein is wrong, so is Posnanski. The evidence supports anyone’s intuition that an extra win is worth more to a team that’s in contention for a championship than it is to a team that’s out of the running. In fact, the data show that an extra win is worth a lot more to a team that’s already likely to win 90 games than it is to a team that’s likely to win only 80 games.
The second interpretation of Posnanski’s opinion is that a player’s value should be considered independently of his team’s record. It’s not fair, in this view, to give a player credit for what his teammates do. This view is simply a personal judgment, but Posnanski claims that his position is somehow more logical than the other.** That’s not merely completely, utterly, incontrovertibly wrong. It’s also illogical.
Dave Cameron implicitly recognizes this. He doesn’t like the fact that people can argue over “value”, so he wants “MVP” changed to “Best Player”. The counterargument to this is that Major League Baseball already has many “best player” awards: the Cy Young award for best pitcher and the Hank Aaron award for best hitter. In addition, the MLB Players’ Association bestows an Outstanding Player award. There is a Gold Glove award for each fielding position, and recently a Platinum Glove award has been introduced (although it’s not an official MLB award) to honor the best overall fielder regardless of position.
So what’s left to recognize with an award? Value, in all its debatable glory.
Let the debate continue, but in complete awareness of what the fundamental issues are.
*Unnecessary detail: “Errors” have long been understood to be unreliable statistics because a fielder who manages to get close to a relatively small number of balls in play can have a low error total but still be of low value. “Assists” are equally unreliable. Many infielders’ errors are due to errant throws, which means that a really good first baseman will improve the stats of his fellow infielders. The credit for his performance is misallocated by “assists”. For outfielders, “assists” are earned by throwing runners out on the bases, but when an outfielder develops a reputation for strong, accurate throws runners are much less likely to try to advance an extra base. Recent attempts to improve on these statistics are clearly unreliable, since the same player’s rating can fluctuate wildly from year to year.
**The adherents to the Gertrude Stein view of value often claim that the alternate view means that only players from first-place teams can be the “most valuable”. This is complete nonsense. A player’s contribution to team value is the product of his individual performance in terms of additional wins and the value of additional wins to his team. Clearly, a player on a 70-win team won’t have the highest value unless his performance vastly exceeds anyone else’s, but it’s logically possible for that to happen. The Gertrude Stein types who commit this error, and consequently dismiss those they disagree with as dolts, are a particularly annoying subspecies.