The estimable Craig Calcaterra took John Heyman to task yesterday for conditioning his Most Valuable Player vote on whether a player’s team was in contention. Quoth Heyman:

I am not strictly opposed to a player on a non-contender winning the award, which has happened on occasion (think Alex Rodriguez of the last-place Rangers in 2003) although I admit that's a tougher one for me since the word valuable suggests that the players' achievements did not go for naught and actually helped a team play into October…

…[S]ince the award is for most valuable player, and not most outstanding, the effect a player had on the pennant race should be vital. If someone else wants to interpret most valuable as synonymous to best, they can. And if someone else wants to interpret it as being valuable to a particular team, they can, too. But there is plenty of precedent to suggest it means valuable in the league.

The ultimate goal of any player is to win, so the value of the individual accomplishments that lead to a pennant should be viewed in that context.

So while [Jose] Bautista has been the most outstanding player in the league whether you use WAR or OPS or or any other key stat, it’s a tough case to make for him as MVP in a year when so many stars are ushering their team into the playoffs.

This is pure silliness, though it didn’t upset me nearly as much as it did Craig—what got me was, “In addition to A-Rod, in 1987 the Cubs' Andre Dawson won the award after hitting 49 home runs (equaling the second-highest total in a quarter-century), a rare show of support for a player on an also-ran team, and that may happen when such a player laps the field statistically.” The idea that Dawson lapped anything more important than an ice cream cone that season still chaps me nearly 25 years after the event; Dawson and his minuscule on-base percentage wasn’t one of the 15 most valuable National Leaguers that year, and I will never, ever get bored of saying so. (My vote, had they been giving them out to high school students, would have been for Ozzie Smith.)

But I digress. In our pennant race book, It Ain’t Over ‘Til It’s Over, Jay Jaffe wrote a chapter called “The Summer of Loving Carl Yastrzemski,” where he looked at the idea of a singular player carrying a team to the postseason. Yastrzemski certainly did that in the short term in 1967, batting .523/.604/.955 over the final two weeks of the season, but no one maintains that kind of pace over the full season, so the winning effort is inevitably the result of a team effort. As Jay wrote, “The best hitter can bat only once every nine times, the most durable pitcher needs a few days of rest between starts, and even the best fielder (beyond catchers) handles the ball only a handful of times a game, making it extremely unlikely that a team could rely on the same player over and over again for that extra boost.” Jay found that the greater the gap between a team’s best player and its second-best player, the less likely the team was to win, and that was true no matter how good the best player was. The correlation between a team’s best-ranked player in WARP and winning was actually much lower than that of its second-best player and winning. As the song goes, it takes two—at least two.

Thing is, you don’t need to do a major sabermetric study to know that winning in baseball requires a team effort. When Babe Ruth set the record for slugging percentage in 1920, personally out-homering every team in the league, the Yankees finished a close third. When Steve Carlton went 27-10 with a 1.97 ERA in 1972, the Phillies still lost 97 games. The Orioles lost 95 games in 1991 despite Cal Ripken playing every game, hitting .323/.374/.566, and having his best year with the glove (for 10.5 WARP as we figure it). And however illicitly Barry Bonds might have achieved his great seasons from 2000 to 2004, the Giants went to the postseason in only three of those five years. Bonds’ bat could affect many things, but it couldn’t give the team better pitching in 2001, or a deeper lineup in 2004. Of all the things one could blame Bonds for, why blame him for that?

In truth, the voters didn’t blame him; they gave him the MVP every year from 2001 to 2004 (though the Giants went to the postseason in 2000, they chose his teammate Jeff Kent, who was almost as good as he was—there again, it takes two). Of course, even the Giants teams that didn’t win were good and Bonds’ seasons statistically outstanding, so his awards would no doubt fall under Heyman’s generalized “contenders“ rule or “Dawson laps things” exception, but the bigger point is that even in their best years, not even the game’s supermen, Ruth and Bonds, could always haul their teams to a pennant.

“Valuable” is an interesting term in that most of the time its meaning is relative, not absolute. A snorkel is valuable if you’re hanging around the Marianas Trench with Jacques Cousteau, less so if you’re wandering through the desert with Lawrence of Arabia. Fortunately, the universe of baseball is far more limited, and a valuable player in New York would be a valuable player in Los Angeles. Babe Ruth in a St. Louis Browns uniform would have been no less Babe Ruth. To paraphrase Gertrude Stein, value is value is value. Had Derek Jeter made it to the majors with the 1996 Kansas City Royals, he would still have 3,000 hits, would still be an all-time great shortstop—he just wouldn’t have any rings. Once he’s on the field, a player creates his own value; the rest is up to the general manager.

I am not necessarily endorsing Jose Bautista for the AL MVP award; there are other, more useful arguments that can be made against him than his team’s place in the standings, particularly a relatively quiet (.254/.412/.483, six home runs through Sunday) second half. However, I would rather see him win it than have this naïve confusion of baseball with a sport that can be rassled down by one determined player—perhaps Clint Eastwood’s Man with No Name in flannel—continue instead of the game being understood as the team sport that it is. As Calcaterra writes, “I don’t understand what good an award is if it’s premised on completely and utterly divorcing it from the essence of the game itself.”

No player can win as an island, a reality that would seem to do major damage to the psycholotgical component of the "team in contention" argument. None of us on the outside of Baustista's head can compare how he processes pressure compared to Curtis Granderson (or vice-versa), but writers sure like to try, claiming that the player on the contender is under greater stress than the one on an also-ran. I'm not so sure. Who is under greater pressure, the good player on a bad team, knowing that he must live up to being the club's sole drawing card and offensive support, or Curtis Granderson, subject to a different kind of expectations but surrounded by a strong supporting cast that will pick up the club whenever he fails? The truth is probably that the reaction will vary with the individual, and therefore generalizations of any kind are useless. The only thing we can know for sure is that pressure comes in all kinds of flavors, and that of being on a contender is far from the only one.

In the article cited here, Heyman says that wins above replacement is a flawed statistic because the weighting of its components can be arbitrary. The different definitions of WAR or WARP, as opposed to, say, the singular definition of batting average, no doubt makes the statistic harder to rely on for some, though I see (pardon the expression) value in having more than one formula: Unlike batting average, which is a simple mathematical fact, wins above replacement is an estimate, and it’s not a bad thing to have a range of competent estimates. When those estimates reach a consensus, you can be more secure that you’re on the right track.

As such, I propose this to Mr. Heyman: Baseball Prospectus and Baseball-Reference, which figure wins above replacement in slightly different ways, agree that a season of 10.0 or more WARP is a rare and special thing. We list 28 such seasons since 1950, they list 36. Whichever figure you prefer, it’s a tiny fraction of the numerous full seasons that have been recorded by players over the last 60 years. They also agree on the upper limit of a player's wins above replacement. BP ranks Barry Bonds’ 2001 season as the best since 1950 with 12.2 WARP. Baseball-Ref agrees, scoring the season as worth 12.5 wins.

If that’s the most a player can give on the road to the 90 or 95 or 100 wins it takes to compete in and win a division, then it is clear that no one player can make the difference unless he is added to a roster that is already nearly complete. Twelve wins isn’t even near half what a team needs to contend, so doesn’t that admit the possibility that judging a player against his team’s accomplishments is completely unfair?

Heyman indirectly acknowledges this possibility in his own MVP selections: He lists two Yankees, two Tigers, and four Red Sox in the American League, two Brewers, two Braves, two Cardinals, and two Phillies for the National League. He missed the forest for his own trees.