Monday, March 19, 2007

Fantasy Articles without Fantasy Analysis

This article over at used Baseball Prospectus' PECOTA system to project disappointments at every position on the diamond.

But, looking at this list, do you see disappointments?

Plug Lyle Overbay's 20/82/3/.291 line into Alex Patton's 2006 formula, and you get an $18 player. No, it isn't the $24 player Overbay was last year. But do you find that "disappointing"?

I don't. I'm not expecting Overbay to repeat last year. More importantly, I'm not going to pay full sticker price for what Overbay did in 2006.

Eric Chavez's projection might seem a little low, and is a stronger recommendation...his 24/83/4/.263 line translates into a $16 season. Given that his 2003-2005 earnings were $27, $20 and $22, that tells me that Baseball Prospectus is making a strong call...don't touch this guy, he won't bounce back.

But the article's comment about Chavez sheds little light on what to pay for him this year:

After an awful 2006 the 29-year-old's window of opportunity for another big year has come and gone. There are too many things preventing Chavez from being a fantasy stud, from his disposition to injuries to his inability to hit lefthanded pitching.
OK...but what should I pay for him? A $27 hitter isn't a stud in my book. But will he bounce back into the $20s?

I really like the great analysis that's done at Baseball Prospectus. But I hate seeing an article in a fantasy section of a web site that doesn't truly give me a full analysis of what to expect from a player and what to pay. Without that information, I feel like I'm groping in the dark, and can't compare the information with anything.

Two interesting footnotes:

First, the article's lead said:

Fantasy leagues are no different than the stock market at the turn of the millennium: plenty of overvalued choices that can singlehandedly destroy your portfolio. (Jhonny Peralta, anyone?)

PECOTA's 2006 projection for Peralta was 20/78/3/.273, which translated to $17 in earnings based on Patton's 2005 formulas. A $3 drop from 2005, but hardly a great call by PECOTA. An odd inclusion at the beginning of the article.

Second, Ichiro Suzuki's 2007 PECOTA projection is 6/57/23/.312. His comment says:

There are people in King County, Wash., who claim to have seen him walk on water, but most hitters who rely on speed and batting average encounter trouble once they pass the age of 32. Drafting the 33-year-old Ichiro already means punting on HRs and RBIs.
His 2006 PECOTA projection was even worse at 7/64/25/.307. His actual 2006 line was 9/49/45/.322. It looks to me like PECOTA's going to keep running Ichiro down until, ultimately, they are right.

Beware PECOTA's SB projections, by the way. PECOTA hates the stolen base, or at the very least likes to severely underestimate how many will occur. Here are the PECOTA Top 10 projected 2006 American League SB totals:

Top 10 PECOTA Basestealers Projected
RankPlayerSB Projected
1Scott Podsednik45
2Carl Crawford44
3Chone Figgins39
4Joey Gathright30
5Ichiro Suzuki25
6Corey Patterson24
7tTorii Hunter23
7tJulio Lugo23
9Brian Roberts22
10Coco Crisp18

And here are the actual American League Top 10 last year:

Top 10 A.L. Basestealers Actual
1Carl Crawford58
2Chone Figgins52
3tIchiro Suzuki45
3tCorey Patterson 45
5Scott Podsednik 40
6Brian Roberts 36
7Derek Jeter34
8Orlando Cabrera27
9tJohnny Damon25
9tLuis Castillo25

PECOTA grinds basestealers to dust. 14 steals below on Crawford, 14 below on Roberts, 13 below on Figgins, and a whopping 20 off on Suzuki. Geez, even Podsednik only fell five short of his projection despite the fact that he was playing hurt for two months.

Something to keep in mind if you're using PECOTA to track how many SB are available in your league. You might want to bump up the numbers to make them closer to reality.

No comments: