PECOTA is out

yeah rlly, 92-98 wins for all three teams is not a shocking call. Injuries and other unforeseeables will determine the order.

J, I am familiar w/ early 21st century vernacular.

rlyw described pecota as assuming that the bullpen would be terrible and the nyy wouldnt do anything to fix it and so their 93 wins proj was basically bullshit. they also said one of their multipliers was inexplicably high compared to previous years and that = bullshit

"The CAIRO IP are based on the last set of Diamond Mind simulations I ran, averaged over 100 seasons so they don't line up with what the CAIRO spreadsheet, and they need to be tweaked further. I didn't realize Burnett was only pitching 167 innings, for example. However, if you look at where things differ, it's primarily with Burnett and Gaudin in terms of rate of performance, and then in how many innings they're allocating to people like Boone Logan and Jonathan Albaladejo.

Digging a little deeper into what these numbers say, PECOTA isn't displaying total runs allowed so I divided the total 789 runs allowed in their standings by the 673 projected earned runs on the team depth chart to reverse engineer it. Right off the bat this looks off, because the ratio is 1.172. The typical ratio is generally 1.08. In fact, this is really, really, really screwed up. Since 1946, there have been 1507 team seasons. I'm using that cutoff to remove the small glove/ deadball era and the WWII era where you had players who didn't belong in the majors to get closer to the modern error rate. In those 1507 seasons, there have been 59 teams that had a RA/ER ratio that high or worse. So they're either predicting the 2010 Yankees to allow unearned runs at a rate that has happened at a rate of 0.03915063 in the MLB team seasons since 1946, or they screwed something up.

They have some smart guys over there, so my guess is it's the latter. As someone who screws up his spreadsheets constantly, I can understand that. I left a comment over there asking for them to elaborate. So don't worry about these yet."

Neyer is going over the CAIRO projs (which I'd never heard of):

http://espn.go.com/blog/sweetspot

"I'm waiting to be convinced. BP didn't have a good 2009, and there are some squirrelly things people have noticed about the 2010 numbers. "

Steven Strasburg: 1.40 WHIP, 4.07 ERA
Matt Wieters: .289/.363/.464

Best Defenders by Position by League:
1B: Pujols, David Ortiz (lol wut!!!)
2B: Utley, Mark Ellis
3B: Rolen, Longoria
C: Yadier, Dioner
CF: Carlos Gomez, Franklin Gutierrez
LF: Nyjer Morgan, Crawford
RF: Jayson Werth, Ichiro
SS: Tulo, Vizquel

Worst Defenders by Position by League:
1B: Fielder, Miguel Cabrera
2B: Luis Castillo, Alberto Callaspo (?KC)
3B: Mat Gamel (?MIL), Edwin Encarnacion
C: John Baker (?FLO), Josh Bard
CF: Nate McClouth, Torii Hunter
LF: Jason Bay, Travis Snider (?TOR)
RF: Hawpe, Abreu
SS: Yuniesky, Jeter

Eric Seidman, BP:

Reading through the comments of yesterday’s announcement that the PECOTA projections have been released, it is evident that there is a lot of concern over several aspects of the data, ranging from the projected standings to individual quirks. We understand and appreciate that this reflects a lot of passion for what we do here at Baseball Prospectus. To be blunt: we messed up, and are working to fix the issues.

One issue involves the run environment: individual player projections do not match up with the run totals on the projected standings.

Another problem revolves around BABIP, as defense was being double-counted (double-counted).

These and other issues are being worked on and we hope to have an update provided by the end of the day, to unmess up. Please stay tuned for further updates.

and they're unmessed:

http://www.baseballprospectus.com/fantasy/dc/

angels last in the west? umm

C: John Baker (?FLO)

he's an original moneyballer

PECOTA predicts that 22/30 teams will finish within five games of .500 = PECOTA thinks MLB is the NFL.

There was a season recently where this just about happened... or, I think, everyone was between .400 and .600.

Obviously PECOTA not apt to predict flameouts and all-goes-right years.

Yeah, but this is a lot more extreme -- they're predicting that more than 70% of MLB teams will finish between .470 and .530!

Not predicting flameouts and all-goes-right years != widespread parity in baseball

I have no idea if there's been a season where 70% of MLB teams have finished between .470 and .530... seems possible. (sounds like the entire NL East in '73)

nats will be a .500 team? really pecota?

Esp. surprising given Strasburg's rather conservative estimates.

I dont see how the Dodgers give up fewest runs in mlb short of legendary pitching performances. The defense isnt that great. The only plus defenders I can think of are Kemp, Blake and Furcal (if you dont count throwing errors). Martins defense has been falling off, and Loney has a good reputation but its not backed up by numbers PECOTA would probably use..

Good point, the only Dodger who showing an above average defensive rating is Martin. NB: Not sure how wonx0red my PECOTA data is, I got it before they admitted it was janky.

Got revised again. Dodgers still have absurdly low runs allowed totals. But slip to 3rd in the NL West behind AZ and Colorado.

They really have to learn not to jump the gun trying to please the panting anticipation of roto maniacs.

seriously. they look like idiots.

I don't look at that stuff in any depth til the last week before the season.

yeah it's crazy to put it out so early given that rosters aren't even really set yet. who cares? but between this and the infamous no index book they've been sort of lame recently.

Kahrl boasted last week that the new annual will have an index.

i saw that; v. nearly started my own slow clap cause that's the least the incredible feat of having an index deserves

PECOTA is down and out

Is anyone other than Goldstein/Kahrl worth reading anymore? Or worth paying for the right to read (more accurately)?

pretty sure i'm gonna cancel when my sub is up in may

yeah, I like Perrotto's weekly reporting.

and the Laurila interviews.

yeah my subscription's up today and i don't think i'll be renewing, i basically only read goldstein at this point, plus the chats (are those subscription only?)

i can get by fine with chone for my projections, in fact i think it's been better than pecota the past couple years!

though admittedly less in-depth

I usually skip Goldstein! I find it hard to care about prospects til I can see em play.

maybe if your team had a real farm system...

Their prospects "play a lot of games."

this is an abomination, i refuse to ask for a ysi of it again next year

I usually skip Goldstein! I find it hard to care about prospects til I can see em play.

Me too.

You don't need to subscribe for the chats (which are normally good, as long as they stick to baseball and don't sidetrack into video games or music, i.e. read Marc Normandin at your own risk), and Steven Goldman's "You Can Look It Up" column is still great (if you're really into baseball history). Speaking of history, I guess Christina Kahrl will be doing the daily analysis column now that Sheehan is gone, but I'm sort of dreading that because when it comes to baseball writing, a bit of creative liscence is definitely a good thing but I can only tolerate so many references to Austrian aristocracy of the 1840's.

now that Sheehan is gone

OK, I've been busy nailing down a job this month. When did this happen?

end of the year--his contract was up i guess? did not know they were under contract. he just said they were going separate ways. i won't really miss him.

CHONE projected standings are out, for comparison...

http://baseballprojection.com/2010/standings2010.htm

looks mostly the same i guess

I dont see how the Dodgers give up fewest runs in mlb short of legendary pitching performances. The defense isnt that great. The only plus defenders I can think of are Kemp, Blake and Furcal (if you dont count throwing errors). Martins defense has been falling off, and Loney has a good reputation but its not backed up by numbers PECOTA would probably use..

― mayor jingleberries, Monday, February 1, 2010 2:17 PM (4 days ago) Bookmark

pecota's pretty bad at understanding defense and projecting rosters, which is why they couldnt anticipate (say) the mariners' run prevention being good enough to have them contending in the west

strange how CHONE predicts the Giants can go from Wild Card Contenders up until the last 8 days of the season to last place in the weak NL West with essentially the same offense (if Randy Winn & Ryan Garko = Mark DeRosa & Aubrey Huff which shouldn't be too difficult to replicate) and an improving pitching staff (Lincecum, Cain, Sanchez, Zito + closer Wilson)...? Not saying they were fluky last year but they'll probably finish mid-pack in the NL, rather than in the bottom only 2 games ahead of the Nats/Pirates/Astros/Mets.

Ha, this has got to be Nate Silver:
http://www.sporcle.com/user/nrsilver/

supposedly pecota is out.. again.

supposedly it hates the shit out of my team. =( I dont subscribe so I cant see, but Im not sure I really want to.

who's your team again? (sorry)

how has pecota's accuracy been for the last couple years?

i remember the last time a blog did an analysis of all the stat forecast systems and PECOTA ranked below average.