November 20, 2007
Who is the National League Most Valuable Player?
I have no idea. I'm going to research this article as I write it, in the hopes of having an answer by the end of it, in advance of today's announcement of the BBWAA's choice. I don't expect that press release to end the discussion by any means, given the nature of this year's field and the tendency of the BBWAA to confuse "best story" with "most valuable player."
Here are the stat lines for 10 players, a group that includes the five or six players most often mentioned as candidates, as well as a handful of others I'm including because I think they belong in the discussion. This should not be construed as my ballot.
Player PA AVG OBP SLG SB CS VORP Hanley Ramirez 706 .332 .386 .562 51 14 89.5 David Wright 711 .325 .416 .546 34 5 81.1 Chipper Jones 600 .337 .425 .604 5 1 76.0 Matt Holliday 713 .340 .405 .607 11 4 75.0 Albert Pujols 679 .327 .429 .568 2 6 72.1 Miguel Cabrera 680 .320 .401 .565 2 1 71.4 Prince Fielder 681 .288 .395 .618 2 2 69.1 Chase Utley 613 .332 .410 .566 9 1 68.8 Jimmy Rollins 778 .296 .344 .531 41 6 66.1 IP ERA VORP Jake Peavy 223.1 2.54 77.0
Hanley Ramirez is the National League MVP. No, wait a minute. This list only considers offense, save for the work of Jake Peavy. Offense comprises most of a player's value, but not all of it, and in 2007 there was a wide range of defensive performance by the players in the MVP discussion.
Just based on this list, here are some early observations:
Let's run another list, this time including WARP (which integrates Clay Davenport's defensive statistics) as well as some other fielding metrics:
Player WARP FRAA +/- PMR Albert Pujols 11.3 22 +37 113 David Wright 10.6 5 +13 106 Jake Peavy 10.6 0 +5 -- Matt Holliday 9.7 10 +9 97 Chase Utley 9.3 5 +22 106 Jimmy Rollins 9.2 8 +7 103 Miguel Cabrera 9.2 5 -24 91 Chipper Jones 9.1 1 -3 100 Hanley Ramirez 8.9 -8 -37 99 Prince Fielder 6.8 -15 -15 98
First, off, "+/-" is from Baseball Info Solutions, and almost entirely pulled from the 2008 Bill James Handbook, one of the great resources available today. The system is simply plays made or not made relative to an average fielder, and based on observation by BIS' staff of analysts. PMR is David Pinto's statistic, based on play-by-play data. I could not locate data for pitchers. Average is 100. I could also not locate Ultimate Zone Rating for the 2007 season.
I include a number of defensive metrics because there is no one magic bullet for defense. Where these numbers agree-and in most cases, the arrows are pointing in the same direction-I think they serve to provide a good evaluation of a player's defensive value. Where they don't, you have to use some judgment. I think the range metrics have a better read on Miguel Cabrera's defense than Clay's system does, and the agreement between FRAA and +/- seems to put the PMR figure to shame on Hanley Ramirez. Overall, the supporting evidence on defensive performance shows that WARP is a good measure of actual value, because FRAA and the range-based metrics line up well.
Does this mean that Albert Pujols, who isn't anywhere near the popular discussion, was the NL MVP in 2007? It's very hard to build a case against him. He was a Gold Glove-caliber defender in all three systems, and the fifth-best hitter in the league by VORP (third-best by EqA and EqR, fourth in RARP). We're already adjusting for positional value and park there, which makes it hard to start working up cases for Utley, Rollins, or Wright.
That Pujols has never been talked about as this year's NL MVP has more to do with the Cardinals' lousy season, which had nothing to do with Pujols and everything to do with the injuries to and collapses of everyone around him. There's also very little recognition of what a good defensive first baseman can bring to the table. Look at those numbers again: Pujols isn't just a good defensive first baseman; he's one of the best defensive players in baseball, by any system. He's saving, and I'll use a wide range here, somewhere between 15 and 30 runs a season with his glove.
I'm comfortable, when an MVP race is close, with the idea that you can look at soft factors in picking a winner. On the margins, you can give credit for things like clutch performance, or whether a player was playing in a race, or perhaps late-season heroics. The problem here is that this race doesn't look particularly close; Pujols has a lead close to a full win in value on the next two candidates, and is close to two wins clear on Matt Holliday. How much credit can you give a player for his teammates' performance before you BBWAA the whole process?
I'm not sure what the answer to that is, but I know it isn't "enough to make Matt Holliday or Jimmy Rollins the MVP." The NL MVP was one of Pujols, David Wright, or Jake Peavy, the three best players in the league who happened to play on three teams that didn't make the postseason. The one-game, or half-game, difference between the Mets and Padres, and Rockies and Phillies, is so small that it doesn't belong in this discussion. The excessive weight that the actual voters will put on that difference skews things in a way that makes it impossible to have a real discussion about value.
In choosing among Pujols, Wright, and Peavy, the soft factors come into play a bit. Although Wright gets lost in the collapse of the Mets' bullpen, his play in September was terrific (.352/.432/.602; .397/.451/.575 during the season-ending 5-12 stretch). He wasn't the reason the Mets lost the division by a game; he was a reason they didn't lose it by more. Pujols, like Wright, played well as the Cards fell down around him, going .386/.486/.625 in September, and .372/.453/.628 during the team's cripping 1-13 stretch in early September. Peavy had his highest monthly ERA in September (counting the one-game playoff on October 1 in those stats).
Team performance has come to mean too much in the determination of the Most Valuable Player awards. In baseball, we can measure individual contributions to winning so well that there's little need to make contextual arguments in ascribing value to a player's season. To the extent that it's done, it's generally to make a post facto case, rather than to actually evaluate the players involved. No one is talking about the great Septembers that Pujols and Wright had because their teammates performed poorly, but those performances would be held up as evidence in their favor had those teammates been just a bit better. These two players were, with Peavy, the best in the league this season, and choosing between them isn't easy, but what we can say with certainty is that they were better than the most popular candidates, and should be on ballots above them.
As much as I want to vote for Wright, who was my off-the-cuff choice for NL MVP over the past five weeks, he was outplayed by Pujols, and there aren't enough soft factors to overcome that. Defense matters, and Pujols was a tremendous defensive player, so good that he leapfrogs the few players in the league who out-hit him. Albert Pujols is my choice for NL MVP.