Does Marcel Undervalue Roy Oswalt?
It was a tale of two halves for Roy Oswalt in 2008. Before the All-Star break, Oswalt suffered from injuries (strained abductor muscle, hip injury) and a bad case of gopheritis. But in the second half he was one of the top pitchers in baseball. Oswalt was 7-8 with a 4.56 ERA with 18 HRs in 116.1 IP in the first half. But he was 10-2 with a 2.24 ERA with five HRs in 92.1 IP after the break.
Oswalt reversed a three-year trend of declining K/9 rates, posting a 7.12 rate. Additionally, his BB/9 and WHIP rates were nearly identical to his career rates. On the flip side, Oswalt posted the second-highest HR/FB rate of his career at 12.7 percent. The only thing that saved him was that he allowed the fewest fly balls of his career. Oswalt’s 29.2 percent FB% ranked 14th in the majors last season.
What can we expect in 2009? Oswalt’s Marcel projection, which regresses to the mean and uses an age factor, has him failing to reach 200 innings for the first time since his injury-plagued 2003, a 3.65 ERA, which would be the highest mark of his career, and the fewest strikeouts of the last five seasons. This seems a tad pessimistic.
Once Oswalt got over his injury problems, he was a completely different pitcher. The only way his Marcel projection makes sense is if Oswalt battles injuries again throughout the year, certainly a possibility given his injury history. He has endured wrist, groin, foot and hip injuries (among others) throughout his career. But he has also topped 30 starts in six of his seven full seasons in the majors.
His velocity has been unchanged the past few years and he is one of the top groundball pitchers in the majors. Oswalt’s 2008 HR rate seems like an injury-related thing, especially since he allowed 0.49 HR/9 when he was healthy after the break. Oswalt’s upside is as a top-10 pitcher. Because of his injury history, one should downgrade him somewhat. But nowhere near where his Marcel projection rates him.
Marcel seems to undervalue every single player in the majors. Marcel has 1 hitter eclipsing 34 homeruns this year. Just for reference, 14 players hit 34+ homeruns in 2008, a down year for Fantasy stats overall.
I’m sorry….Marcel has 1 hitter hitting more than 32 homeruns, something 19 hitters did in 2008. 13 hitters did it in 2007. 29 hitters did it in 2006.
Marcel’s playing time estimates are generally conservative, and even considering that, it’s very very difficult to project playing time anyway. If you feel that Oswalt should be getting more IP, or power hitters should be getting more PAs, you can adjust it as such. Marcel is much better at predicting rates than totals.
Within the context of the Marcel projections, here are the pitchers with a projected ERA lower (and in order) than Oswalt:
Rich Harden,
C.C. Sabathia,
Jake Peavy,
Johan Santana,
Tim Lincecum,
Brandon Webb,
Justin Duchscherer,
Roy Halladay,
Chris Hamels,
Chris Young,
Ben Sheets,
Danny Haren,
Chris Volstad,
Chad Billingsly,
Eric Bedard,
Derek Lowe,
Roy Oswalt,
I think the top 6 pitchers you’re going to take over Oswalt hands down, same with Halladay and Hamels. Duchscherer you can probably get later than Oswalt. Chris Young, Volstad, Bedard (given injury history), and Lowe I wouldn’t take before Oswalt. I like Billingsly a lot, I’d take him before Oswalt. Haren you can make a case for.
Anyway. My point is… where Oswalt is ranked in the projections (at least by ERA), he’s not too far off where I think he should be.
I’d also like to say that the projections on FanGraphs are not to be blindly used. In reality, they are all quite similar and you could probably win your Fantasy league using any of them as a guide.
The reason we’ll eventually provide at least 4 different sets of projections is so you can choose for yourself which one you’d rather use.
If you look at Oswalt’s rate stats in 2008 and compare them to the Marcel projection, nearly every single one of them is worse. ERA, K/9, BB/9, K/BB, AVG, WHIP, BABIP are all worse in the projection. The thing that I find interesting is that Oswalt struggled with his HR rate last year and the projection has him with a better HR/9 but worse overall numbers.
Obviously, I know that projection systems are conservative by nature and that Marcel is a very simple one, on top of that. I just find this particular one interesting. YMMV
I just read that article posted on the Marcel article’s comments and I better understand now why it underestimates many players.