THT's Fantasy Archives

---

Posted by Matt Hagen at 2:12am (1) Comments

Clone or drone

Jayson Werth has put together another fantastic season, which comes on the heels of his breakout 2008. Some fantasy commentators are now slotting him near Curtis Granderson in their Average Draft Position rankings for next year. Since Granderson went around the 61st pick (in CBS Sportsline rotisserie drafts) whereas Werth went around 157th this year, that is quite a promotion. Certainly Werth has delivered better value this year. Of course, the real question is what we should project him for next year.

For reasons that will become clear later in the article, the best way to value him for next year would probably be to project his stats using a projection system like Chone or Marcels. Another way to show his value is by using heuristics: comparing one player to another player whose value is easier to grasp.

Werth represents a certain type of player that I think is very difficult to project using any method. He's already 30. He's only had two seasons with fantasy-worthy output, so any Marcels-type projection system which takes three years of data is going to penalize him (perhaps rightly) for a very bland 2007. During the past two seasons, he's shown a wide-variety of—but not especially stable—set of skills and propensities.

His batting average, walk and strikeout rates are stable. He strikes out quite a bit, though he does draw he share of walks. He hasn't been as successful stealing bases this year but the decline in stolen bases is primarily due to fewer attempts. He's hit many more home runs this season, resulting in more runs and RBIs.

So Werth is one of those coveted power-speed outfielders. He's never going to be compared to Carlos Beltran or Matt Kemp, but we have to find a home for him somewhere. Obviously no two players are exactly the same (though Troy Patterson finds some couples that are awfully alike), and even if they have been very similar, past performance is no guarantee of future verisimilitude. But the task remains: whose old clothes will better fit Werth next season: Eric Byrnes or Granderson? (Or perhaps Aaron Rowand, Milton Bradley, or Michael Cuddyer?)

Again, none of these players are perfect fits: Byrnes stole anything he could while he was healthy, but then suffered a set of injuries. Granderson's two years younger than Werth. Rowand, who also had one of his best years roaming the Phillies' outfield, has never put together back-to-back solid seasons. Bradley has playing time issues for a variety of reason, but also has hit for a better average (some years). Cuddyer has never been a great source of stolen bases.

When doing casual similarity heuristics, it is very easy to pick the winner of the "Who is Werth most like?" beauty contest by dismissing (or just not considering) other candidates for "easy" reasons. Byrnes' injuries and perhaps chemically aided performance rumors might disqualify him from the contest. Bradley is a headcase and Werth is not, so perhaps Uncle Milton should be out. And so on ... But by doing so, we do lose some potentially valuable cautionary tales.

If pressed, I'd say Werth falls closer to Cuddyer than Granderson. Cuddyer and Werth are the same age. Actually, even though Cuddyer and Werth both have elevated HR/FB rates and could be due for a fall in home runs next season, Cuddyer might be the better bet to maintain his power. Hittracker.com has Cuddyer on the leaderboard for the Golden Sledgehammer Award, a measure of average true home run distance, whereas Werth leads the NL in home runs that were "just enough" to go out (though some of this may come from being in a home park where those kinds of fly balls can go out as home runs). We shouldn't expect a resurgence in stolen base attempts now that Werth is sliding a bit further down "the razor blade of life".

More globally, I hope this also illustrates why using (or making your own) systematic projection system has value. You needn't follow it slavishly. But, by using averages instead of case-studies, it can summarize the useful information that all of the Rowands, Bradleys, Byrnes, Grandersons and Cuddyers have while washing out some of the differences.