Defense and winning championships, but mostly defense

The Post-Dispatch ran an article yesterday about the Cardinals' two Gold Glove contenders, occasioned by John Dewan releasing his yearly take on his own DRS statistic. Pause: How cool is it that the Post-Dispatch ran an article yesterday about defensive play-by-play metrics? We complain about sportswriting, sometimes—sometimes we even write entire blog entries about it—but guys like our own Derrick Goold and Bernie Miklasz are doing a great job of pushing their huge readership toward what it's becoming increasingly difficult to call "advanced" statistics.

In any case, the article offers another look at the Cardinals' defense, which we seem to have been following, this year, mostly in terms of its impact on the ever-more-popular WAR superstats. DRS is more positive than UZR (Fangraphs) and TZ (Baseball-Reference) on Albert Pujols, Yadier Molina, and especially Brendan Ryan, and similarly bearish on Colby Rasmus's sophomore season. What can we divine from having three vaguely understood stats to look at, instead of one?

From Colby Rasmus and Albert Pujols: One year of defense isn't a trend. In 2009 Rasmus's disappointing offense was masked, incredibly conveniently, by the sudden widespread availability and acceptance of advanced defensive statistics, specifically UZR. If he had had the same rookie season in 2005 we might have framed it, "He's a fine defender, but his offense was a disappointment." In 2009 that became, "He was only so-so on offense, but he was worth 10 runs! on defense." 

But I'm sure we'll look back on 2009 with the same skepticism we do 2005, as far as that evaluation goes—this year he dropped, on UZR, from 10 to -6; on TZ from 15 to -9; on DRS from 6 to 1. WAR looks so expertly packaged that it's easy to forget that the thing that made Colby Rasmus a useful center fielder in 2009 could have been a trick of the defensive statistics, a simple fluke, even a career year. Luckily he found other methods of making himself useful in 2010.

The same goes for Albert Pujols, whose 7.2 WAR (via Baseball-Reference), in addition to leading the National League, is his lowest total since 2002.

He was a little off his highs on offense, but the main difference was that his defense, which has been routinely feted by each stat since he moved to first base, dropped off from best in the league to a little below average; the decline came from the 1.6 WAR he lost on defense, not the 0.4 he lost on offense.

I'm willing to believe Colby Rasmus isn't really a win better than an average center fielder on defense, but I'm not willing to believe that Albert Pujols really lost the ability to play first base this year, after playing it superbly for five seasons. (UZR, incidentally, has seen him as basically average for the last two years—there's a lot to be wary of when it comes to assigning value from these.)

From Yadier Molina and Brendan Ryan, we learn that defensive stars can be "one-dimensional sluggers", too. The difference between Brendan Ryan this year and Brendan Ryan last year is like the difference, maybe, between Matt Holliday and Adam Dunn. Dunn and Holliday are both incredibly dangerous hitters, but Holliday has also proven himself to be a (typically) capable fielder. One of these guys gets a seven year contract and the adulation of Cardinals fans everywhere, so long as he doesn't fail to drive in a baserunner twice in a row; the other wants terribly to stay in Washington but won't be given the opportunity. 

For a long time the trend was to overvalue the guy who can mash, a la Josh Phelps on the cover of Baseball Prospectus, but now I think things have moved in the opposite direction; we'd rather have Brendan Ryan than Jermaine Dye, even if they both manage one win above a replacement player over the course of a season.

(Of course, the variability of defensive metrics come into play here, too—if Ryan was as good as TZ and UZR suggest, 10 or 11 runs above average, he was just barely passable last season; if he was 27 runs better, like DRS suggests, he somehow managed to be an average position player with a .573 OPS, which is a little like Adam Dunn putting together a league-average season at shortstop.)

As defensive specialists begin to reach salary parity with offensive specialists—I don't think it will happen for a while, given the level of certainty with which we can value offense—I think it'll be especially important to remember this last issue. It's intoxicating, even now, to be able to have some run value to ascribe to defense, after doing without so long. But as we move past an era that is, to me, typified in its excess by Dave Cameron writing this jubilant post about Nyjer Morgan (+5 by DRS this year, +22 last year) being as valuable as Adam Dunn (-33 last year, -3 this year), we'll need to learn, again, to view the one-dimensional defender with the same jaded eyes we turn to Ty Wigginton.