Wednesday, December 27, 2006

Johnson > Zito

Rumors are swirling that the Yankees are looking to trade Randy Johnson to an NL West team, and possibly use the money they free up to sign Barry Zito. Unless they're getting a big package of talent in return, this will not be a good exchange.

I can certainly understand the Yankees' motivation to move the Big Unit. They're under intense pressure from fans and ownership to get back to the World Series, and Johnson has been a massive disappointment since arriving in the Bronx. Furthermore, there's nothing new about this story: the Yankees acquire a "big-name" pitcher, he disappoints in pinstripes, then he's sent packing. Javier Vazquez, Jeff Weaver, Hideki Irabu...the list goes on and on.

The big thing these pitchers have in common is that the Yankees expected too much out of them. Vazquez had always given up a high number of home runs; he allowed 33 in his one year with the Yankees and was subsequently dealt as part of the package for the Unit. Weaver had been touted as a future star, but at the time of the trade owned worse peripherals than Ted Lilly, the centerpiece of the package the Bombers gave up to get Weaver and now the owner of a much bigger contract than Weaver will receive. Irabu was a good pitcher in Japan, but not the superstar he was made out to be, and the major leagues are a lot tougher than NPB.

In Johnson's case, the Yankees were right to expect a star pitcher, though they should have anticipated the strong possibilities of injury and age-related decline. Johnson still had the peripherals of a star in 2005, with 47 walks and 211 strikeouts in 225.2 IP in a tough AL East. He pitched much better than Kevin Millwood, who posted a 2.86 ERA to Johnson's 3.79. He wasn't Pedro Martinez, but he had a fine season.

Johnson entered 2006 with one of the most favorable PECOTA projections for any pitcher, featuring a 3.24 ERA and 1.08 WHIP. Johnson's actual stats were far worse, but he still had 2.6 BB/9 and 7.5 K/9, measurably better than pitchers who are getting 4 years and $40 million plus this year. As Nate Silver demonstrates, Johnson's stats last year pointed to a much lower ERA, but he was very unlucky, allowing a high batting average on balls in play (BABIP) with runners on base, when it meant the most. His PECOTA ERA projection for next year, after two down seasons and at age 43, still comes out at 3.52. Put simply, Johnson is still a damn good pitcher. (Also, he should be a big bargain in your fantasy draft next year.)

What about Barry Zito? Zito has had the reputation of an ace for years, even though he hasn't really been one since 2003. After his Cy Young season in 2002, Zito's strikeout rate has been in consistent decline, while his walks are trending upward. Zito is extremely durable--he hasn't missed a start in his career--but he doesn't have the peripherals of a number 1 starter, or even a number 2.

How does he continue to produce a decent ERA every year? Zito has a great defense behind him that converts an above-average number of balls in play into outs. Even beyond this, he has had more than his share of luck in BABIP and in stranding the runners he does allow on base.

Most sabermetrics sites have their own statistic designed to separate the pitcher's skill from the variance inherent in baseball. Baseball Prospectus has PERA, The Hardball Times has xFIP, and Baseball HQ has xERA, to name a few. All are on the same scale as ERA, so a 4.50 xFIP projects a 4.50 ERA in a luck-neutral universe. These numbers have been shown to be a better predictor of a pitcher's future success than ERA. For example:

Jason Marquis

2004 3.71 4.73
2005 4.13 5.29
2006 6.02

Al Leiter

2004 3.21 5.30
2005 6.13

The three measurements, though usually similar, are independently calculated, so they don't always agree. All three do concur, however, that Barry Zito is likely to have an ERA considerably higher than 4.00 going forward. His cumulative xFIP from 2004-06 is over 5.00.

What happens to Zito when you put him in front of a bad defense, and he doesn't get so lucky at stranding runners and converting balls in play into outs? His ERA balloons, probably not as profoundly as Marquis' or Leiter's, but still significantly. This is why the brilliant Keith Law ranked Zito 15th among this year's free agent crop--behind the much-maligned Ted Lilly and Gil Meche--even though he will receive the second- or third-biggest contract of the lot.

Johnson is likely to be a better pitcher than Zito in 2007, and probably in 2008 as well. I'd rather have Johnson for the next three (perhaps four) years than Zito for that same span, yet the Yankees are reportedly interested in exchanging one year of Johnson for six years of Zito at a similar average annual value. If anything, they should be considering extending Johnson while his perceived value is low. Brian Cashman is a bright individual, so I'm sure he's at least considering this option, Steinbrenner be damned.

The contracts given to Meche and Alfonso Soriano will draw the most criticism, but if Zito gets six years and anywhere near $100 million, it will be the worst signing of the offseason.

No comments: