PECOTAing the Yanks and Sox

Baseball Prospectus has posted its PECOTA projections (subscription only). Here are a few Yanks- and Sox-related highlights:

• PECOTA doesn't see A-Rod having a bounce-back year. He'll stay around the rather miserable 34 home run mark, knocking in 104 runs and batting .288.
• It likes Big Papi a lot, although, again, he will not match his spectacular 2006. Ortiz is predicted to hit 41 dingers (I imagine 20 of those will be game winners).
• Melky!!! His numbers don't look great in his second year but 3 of his 4 comps do: Carlos Beltran, Roberto Alomar and Rickie Ashburn. Not too shabby.
• Coco Crisp and Jason Varitek will both have bounce-back years. PECOTA has the Sox center fielder batting over .300, with an OBP over .360. His numbers do not match Johnny "Heart of the Yankees" Damon, but they're in the same territory. Damon is projected to hit 18 homers and get on base at a .362 clip.
• J.D. Drew and Bobby Abreu are very similar players with PECOTA giving the slight advantage in VORP to the man who most clearly resembles the love-child of Jose Canseco and Sammy Sosa (minus the steroids stigma). Abreu's VORP: 21.2; Drew's VORP: 19.6
• Philip Hughes is projected to have a nice ERA and WHIP in limited innings: a 3.91 ERA, 1.29 WHIP and 108 k's in 130 innings. Of all the pitchers on both teams predicted to be starters next year, Hughes has the lowested projected ERA.
• PECOTA's thoughts on the Beckett v. Wang debate: Beckett bounces back (a little) with a 4.47 ERA and 23.7  VORP; Wang regresses  to a 4.35 ERA and a 21.5 VORP. They pitch almost exacty the same number of innings. Essentially, PECOTA has them having very similar years.
• PECOTA thinks Dice-K will be good, along the lines of what Paul has predicted on this site: 162 strike outs in 182 innings and a 4.01 ERA. Schilling is predicted to put up very similar numbers, with the Japanese import having slightly more value.
• PECOTA likes Robinson Cano's upside. The first two comps on his list are Carlos Baerga and Rod Carew. He is projected to have another nice year, although PECOTA is somewhat conservative with his BA (.308).
• From the "Might We Regret The Day We Traded Him" Category: PECOTA likes Hanley Ramirez, predicting a high VORP for the former Sock. I assumed PECOTA was high on Randy Johnson and was going to include him in this category, but it only sees the Big Unit as making 19 starts with an ERA of 3.97. In Arizona, I ain't too impressed by those numbers.

There's a lot more fun stuff which I don't have time to cover. You might choose to put zero stock in these numbers. Anyway, I find them nice to look at during these boring empty months.

Comments

[Rob (Middletown, CT)]

I saw the Melky comps - they do catch one's eye.

For the most part, the predictions seem reasonable. Much more reasonable than what I'd originally heard for RJ (3.5 ERA as a Yankee!), anyway. A ~4 era in limited innings for AZ sounds a lot more like it.

[Rob (Middletown, CT)]

PECOTA is probably smoking crack with regard to Hughes, though. Rookies usually have some growing pains.

[dc]

thanks for pulling all that info together nick...i do have to plead ignorant about pecota...i've seen all you guys make reference to it in the past [it is an "it", and not a human, right?]...does it involve reading palms, tea leaves, or a crystal ball?...is there any process that determines it's accuracy after the fact...i'm not big on predictive models especially since there are so many variables that can blow the thing out of the water...but, it sounds like fun...would someone indulge me with a brief overview on what pecota is?...

[Lar]

Rob - probably because the "limited innings" where they don't have to adapt yet. Still surprised that he "beat" everybody in ERA, but rookies can get hot over a shorter stretch.. (the younger Weaver, perhaps even Liriano)

[mattymatty]

PECOTA is usually pretty conservative in its predictions. I'd be surprised if Matsuzaka posted those numbers. I'm expecting more from him this season, especially in the beginning before the league starts to get a 'book' on him and his stuff.

Rob - Absolutely everything I've read about Hughes is glowing. I have yet to find anything negative, and trust me, I've tried. The only slight chink in his considerable armor is his age, which isn't a negative by any stretch, but it hasn't yet been established if he can handle a major league workload without injury.

Anyone know the prediction for Schilling?

[Paul SF]

Those vary pretty widely from Bill James' projections. When I get my BP 2007 (on pre-order from Amazon), I'll run the comparison.

Nevertheless, the best starter among both teams, according to the James Handbook, will be Papelbon with an ERA below 3. It also is less high on Coco but thinks Beckett will bounce back more. Interesting. Let the cherry-picking begin!

[Nick-YF]

MM, I'm at school but I think it he had a 4.02 era in a few more innings than Dice-K.

[mattymatty]

Lets see, I'll take PECOTA's Coco and Beckett predictions, the Bill James Papelbon prediction and my own off-the-top-of-my-head prediction for Matsuzka where he goes 30-0 with an ERA of what-does-it-matter-because-he-won-30-games.

Hmm... can't seem to find my objectivity... I know I left it somewhere...

Thanks Nick.

[Paul SF]

Oh, James also says Papi will hit 46 home runs.

[Lar]

No knock on Papelbon, but an ERA below 3 in AL East is slightly outrageous..

[Nick-YF]

As MattyMatty says, PECOTA tends to be conservative. Ortiz's homer total is one of the highest I've seen for a PECOTA projection.

Papelbon's projection is something like a 4.27 ERA in 135 innings.

[Andrew]

I don't think you can predict Pap's numbers as a starter.

A blazing fastball is fine when you're pitching an inning or two at a time, but you're going to have to dial that down a lot when pacing yourself over 6 or 7 innings. And I haven't heard much of his secondary pitches, but when you're throwing low-to-mid 90s fastballs...well, you saw what happened to Randy.

[mattymatty]

Schilling throws low to mid 90s fastballs, and Mussina and Pettitte throw slower than that. Pitching isn't about pure speed, though that is one aspect. Its about location and changing speeds.

Paps numbers as a starter can be predicted, but probably not as acurately as someone who has thrown 200 more innings in the majors. There is clearly a difference between starting and closing, but before he was closing for the Sox he was starting, and he did fine, though admittedly not as well as when he closed.

[Andrew]

That's not what I was saying. Schilling has more control over his pitches than anybody. Andy throws a great cutter, plus he has a pretty good changeup and curveball. Mussina has about a million different pitches, and he can throw them all for strikes on his best days. Papelbon relied on his blazing fastball for the majority of his games last year, and has not really yet proven his secondary pitches to be major-league quality. That's why I doubt his ability to be an good starter more than others, and why maybe he's more suited to a strictly bullpen role.

[Brad]

and has not really yet proven his secondary pitches to be major-league quality.
________________

nearly every strikeout he records is on the splitter, so I'm not sure what you're talking about here. The splitter is obviously done better when it has a fastball setup, but that's every pitcher who throws it. Clearly he can't live with just those two and will need to develop some other pitch, but he absolutely does have a second pitch to rely on.

[dc]

...so what i'm hearing is that all of these prognostications are fun to kick around, but they're basically pointless crap, and this pecota dude is just wild-ass guessing [and hoping] like the rest of us...anyone know what his fan orientation is?...

[Paul SF]

PECOTA is an acronym (I forget what it means) for what is essentially Baseball Prospectus' annual predictions. PECOTA is fairly well respected as such prognostications go. I'm not sure how seriously BP takes their predictions; I for one think it's a lot of fun to see people who know a lot more about baseball stats and trends than I do give their best guess as to how the season will turn out. But it definitely is a mistake to take PECOTA or the Handbook or ZiPS or any of the other projections out there too seriously.

[Nick-YF]

For what it's worth (and it's not worth much) PECOTA was the most accurate published forecaster for the past season.

[dc]

thanks paul and nick...i appreciate the clarification...

[Rob (Middletown, CT)]

Wiki link:

http://en.wikipedia.org/wiki/PECOTA

"PECOTA, an acronym for Player Empirical Comparison and Optimization Test Algorithm..."

Funny, I had to google it - finding it on the BP site was proving difficult :)

[Paul SF]

What differentiates PECOTA is not only the 2007 prediction, but that it predicts beyond just the next year, as well as listing comparative players and giving percentages on whether a player is likely to have a break out season, collapse, overperform or underperform based on last year's stats.

[Rob (Middletown, CT)]

Re: Papelbon...

He's thrown just a hair over 100 MLB innings, with an aggregate ERA of 1.5. So on the one hand my gut reaction to predicting a sub-3 era in the AL East for *anyone* is "yeah, right!" I can understand why someone (or something) might spit out such a projection.

He's really, really good. The main question is health.

[Rob (Middletown, CT)]

Funny PECOTA question & answer in chat just now:

John Barten (Indianapolis,): Peeking through the PECOTA projections, one player sticks out like a sore thumb, Jeff Bianchi, who has 140 career AB in the Arizona Rookie League and is projected for a .299/.369/.422 ML line. Any thoughts?

Joe Sheehan: Guys who have one partial season in professional baseball give PECOTA heartburn. It's best to take those lines with an entire salt lick.

[SF]

If Phil Hughes throws 130 innings for the Yankees, then that means they missed out on Clemens and/or someone got hurt and/or someone expected to do ok didn't.

[Andrew]

Or Hughes was just that good in triple-A and Pavano was traded.

[Paul SF]

To who? And for what? If Pavano is pitching poorly enough to lose his spot midseason to someone with zero MLB experience, no matter how good, then who would be willing to take him on?

[Andrew]

Uh, Pavano could be traded in Spring Training when he shows he can actually throw a baseball, because Cashman and the players and everyone is absolutely sick of him. A major league pitcher with a $10 million dollar a year contract is a pretty valuable commodity these days. That said, I would not be unpleased if he and his contract are traded for a bag of balls to, say, Washington, whose rotation is worse than the Yankees' lower minor league system.

Pavano has value, St. Louis was even very interested in him earlier, but I think balked when Cashman demanded they take on his entire contract. And this was when he hadn't thrown a major leauge pitch in one and a half years. So yuh, a starting pitcher has value.

[QuoSF]

Valuable commodity or not, I'm not sure the Kasten/Lerner operation in Washington is going to take on a $10M contract (in actuality $20M given the two years remaining) in any situation. It's been pointed out on Nationals blogs that the FO isn't secretive about being in a rebuilding mode.

Also, look at very recent history. RJ, however poor his ERA was, was sent along with money to the D-Backs who re-worked his contract. This is a guy who has posted 200 IP the last two seasons with good win totals, bad back and all. He's also a hero to Arizona, at least as a pitcher, if not neccessarily as a person. RJ is case one of why Pavano is going to be traded along with money, if at all.

[dc]

the main difference may be in the numbers...pavano is about 10 years younger, and while he's been "injured", you'd expect him to be able to return from those types of injuries [unless they're in his head], much more readily than rj's cronic back issues...as for the money, i think the yanks agreed to pick up $2m of the annual $16m rj has left, while pavano's only making $10m annually...in any event, it's no deal to anyone if the yanks can't demonstrate that the guy can pitch again...

[QuoSF]

dc, you're right about most of that, honestly. The issue for me I suppose was the team suggested. The Nationals, especially now that Stan Kasten and not Jim Bowden is in the driver's seat for that franchise, are not going to take on a player like that. They just plain aren't. Be prepared with a humongous plate of crow in case they do, but I'm unconcerned.

[Rob (Middletown, CT)]

"If Phil Hughes throws 130 innings for the Yankees, then that means they missed out on Clemens and/or someone got hurt and/or someone expected to do ok didn't."

All three scenarios are clearly possible. Ergo, one of the three coming to pass is not only possible, but likely.

Right now the rotation is:

Moose (will miss a start or three)
Wang (Big innings jump last year, previous shoulder trouble)
AP (elbow always a concern)
Igawa
?

Pavano is theoretically in line to be a starter, but I'll believe that when I see it. Maybe not even then.

Igawa is an unknown who is used to pitching in a 6-man rotation and is projected by most to be a so-so 5th starter type.

Other options with major league experience include Jeff "Scary Flyball guy" Karstens and Darrell "mystery shoulder injury" Rasner.

I'm frankly fine with this situation, but getting Hughes 130 innings in these circumstances really shouldn't be that hard. Even if they *do* get Clemens.

[Rob (Middletown, CT)]

I should add, however, that getting Hughes 130 ML innings means he really should only throw ~50 innings in the minors. He threw ~150 innings last year, IIRC. Jumping above 180 this year would be risky.