UPDATE 8:40 PM: Bumped - there's newer stuff below.
The National League Rookie of the Year award will be announced Monday, and Major League Baseball is already saying that Zach Duke is out of it.
Anyone who reads this site regularly knows I'm not a rah-rah Pirates guy at all, and I think that non-Pirates fans' perceptions of the Pirates are generally closer to the truth than Pirates fans' perceptions. But here, the non-fans' Sports Illustrated-addled perceptions are wrong. Duke should be the Rookie of the Year.
VORP tells you a lot of what you need to know:
Zach Duke 32.7
Ryan Howard 28.6
Jeff Francoeur 21.3
Garrett Atkins 21.0
Ryan Church 19.5
Gary Majewski 19.4
Freddy Sanchez 18.6
....
Willy Taveras 11.9
VORP measures a player's value in runs (on offense only for hitters, on pitching for pitchers) above or below that of a bench player or AAA replacement at his position. It is cumulative, like home runs or RBIs, which means that guys who play more get more time to accumulate it.
MLB lists Francoeur and Howard as top candidates, followed by Taveras, then lists Duke and Atkins at the bottom of the page.
Taveras should not be even a candidate for Rookie of the Year, unless you happen to feel that the Rookie of the Year should be someone who played the whole year instead of someone, who, you know, contributed. This is like saying that anyone who gets perfect attendance in school deserves straight A's. Otherwise, the only case that can be made for Taveras' value involves appeals to his batting average and stolen bases. But he didn't walk enough to have a good OBP or hit for power at all. Batting average is irrelevant if you can't get on base, and stolen bases can almost never be the basis of a good offensive game.
Atkins shouldn't be a candidate either. Consider that Atkins' VORP was two runs better than Freddy Sanchez', even though Atkins' took nearly 100 more plate appearances to accumulate. Was Freddy Sanchez more valuable than Duke to the Pirates this season? I didn't think so. Then consider that Atkins doesn't have a plus defensive reputation.
That leaves Francoeur, Howard and Duke. MLB dismisses Duke for not playing enough, but neither Francoeur nor Howard played the whole year, either. Francoeur had 274 plate appearances; Howard had 348.
Many major league regulars had around 700, so Francoeur played about 39% of the time he could have and Howard played about 50%. Among starting pitchers, only a handful pitched as many as 220 innings. Duke pitched 84.7, or 39% of what he might have if he'd pitched the whole season under, say, Dusty Baker. That's the same as Francoeur, even before considering that rookies rarely pitch 220 big league innings in a year anymore. If you're going to dismiss Duke because he didn't play enough, you've got to dismiss Francoeur too.
Then consider that Francoeur didn't even bother to show up for the final month of the season. He posted a .278 OBP and a .700 OPS in September and October. That, more than anything else, explains the difference in VORP between Francoeur and Duke. Francoeur had one phenomenal month, one good one and one downright bad one; Duke had one phenomenal month and then two that were merely great. He posted ERAs below three in each of his first three months. Then, for good measure, he shut down the Brewers in October.
To put this in perspective:
Francoeur:
July: 1326 OPS (2004-era Barry Bonds)
August: 878 OPS (Mike Sweeney)
September and October: 700 OPS (Darin Erstad)
Duke:
July: 0.87 ERA (Walter Johnson coming back from the dead to kill you ALL)
August: 2.83 ERA (Johan Santana)
September and October: 1.80 ERA (Roger Clemens)
It should be pointed out here that Francoeur deserves extra credit for his defense. This is true. It is a joy to watch him throw. But no amount of defense in a year could make up a month's worth of the difference in value between Roger Clemens and Darin Erstad.
It also should be pointed out that Duke's ERAs are unsustainable; he's not going to be Walter Johnson, or even Johan Santana, forever. Since his DIPS ERA is much higher than his actual ERA, one might even say he got lucky (although DIPS ERA doesn't reflect Duke's outstanding groundball to flyball ratio, which should keep his ERAs down). But those criticisms could also be made of Francoeur, whose numbers are just as unsustainable. He'll probably be a good player in the big leagues, but his control of the strike zone is terrible. He'll probably take a step backward next year.
Howard is probably a better candidate than Francoeur - he hit very well in every month he played except for a few at-bats in May, and he was downright great in August and September. But he lacks Francoeur's defensive value, so I see no reason to give him extra credit for the gap in VORP between him and Duke. Howard was merely excellent for most of the season; Duke was downright spectacular. Howard's OPS would have been 19th best in the big leagues if he had qualified; Duke's ERA would have been 1st if he had qualified. I'll take great over good any day. The Rookie of the Year award should go to Duke.