Page 1 of 1

ZiPS projections

Posted: Wed Oct 28, 2009 7:43 pm
by Rangers
We are not close, to my knowledge, to settling on a system for 2010, so while those who are signing guys based on their ZiPS projections may well be wasting roster spots, a few of us figure that the entire league should at least be aware that several teams' projections have already been posted, in case anyone was not aware of this.

Here are the ones I see so far:

http://www.baseballthinkfactory.org/fil ... ota_twins/
http://www.baseballthinkfactory.org/fil ... n_red_sox/
http://www.baseballthinkfactory.org/fil ... k_yankees/
http://www.baseballthinkfactory.org/fil ... white_sox/
http://www.baseballthinkfactory.org/fil ... it_tigers/
http://www.baseballthinkfactory.org/fil ... o_rockies/
http://www.baseballthinkfactory.org/fil ... nati_reds/
http://www.baseballthinkfactory.org/fil ... ty_royals/
http://www.baseballthinkfactory.org/fil ... _bay_rays/

Posted: Wed Oct 28, 2009 8:11 pm
by Guardians
The same thing was said last offseason too and we ended up using Zips for the 2009 season. Considering we are 12 months into this problem and no solution yet, I really don't see how there are any options besides Zips at this point for 2010. If an alternate to Zips can be decided upon before the start of the 2010 season, then why hasn't it been done already?

Posted: Wed Oct 28, 2009 8:15 pm
by Padres
Astros wrote:If an alternate to Zips can be decided upon before the start of the 2010 season, then why hasn't it been done already?
Still exploring our (albeit limited) options ...

Posted: Wed Oct 28, 2009 8:30 pm
by Royals
I'm still working on a projection system but it still has quite a ways to go. Although now that I have a job I feel far more comfortable dedicating time to it, so we'll see how it does.

Posted: Thu Oct 29, 2009 8:43 am
by Marlins
Am I assuming correctly that this discussion is just to decide what to do if DMB again doesn't do a proj disk? I know it's not a large chance of that happening, but that is the first choice right? To keep consistent with our history?

Posted: Thu Oct 29, 2009 9:55 am
by Rangers
Giants wrote:Am I assuming correctly that this discussion is just to decide what to do if DMB again doesn't do a proj disk? I know it's not a large chance of that happening, but that is the first choice right? To keep consistent with our history?
I think it's definitely first choice.

Posted: Thu Oct 29, 2009 10:13 am
by Phillies
i wonder why Zips hates Gavin Floyd so much.

Posted: Thu Oct 29, 2009 10:51 am
by Yankees
I will take time out of my life to load PECOTA into DMB if it means not using these horseshit projections. I'm no big fan of Rick Porcello - but a 4.89 era?

It's like ZiPS can't stand to think they may have screwed up last year, and has to keep punishing the people they hated who might actually be good.

Posted: Thu Oct 29, 2009 11:00 am
by Mets
I finally got a useful ZIPS for Hellickson. Honestly wasn't hopeful that complete and utter minor league dominance would lead to anything more than the 2009 Tommy Hanson projection.

Posted: Thu Oct 29, 2009 11:33 am
by RedSox
Royals wrote:I will take time out of my life to load PECOTA into DMB if it means not using these horseshit projections.
I was thinking about that. If there's 1800 players projected, divided by 30 owners that's only 60 players per owner to do before next season. Range numbers could come from UZR on Fangraphs for the MLB guys and CHONE's work on MiLB for the minor leaguers once we agree on where the cutoffs for the different ratings should begin.

Posted: Thu Oct 29, 2009 11:48 am
by Mets
don't know if we can use CHONE.....Figgans always projects too high.

Posted: Thu Oct 29, 2009 11:49 am
by Tigers
Royals wrote:I will take time out of my life to load PECOTA into DMB if it means not using these horseshit projections. I'm no big fan of Rick Porcello - but a 4.89 era?
It's like ZiPS can't stand to think they may have screwed up last year, and has to keep punishing the people they hated who might actually be good.

Looks like their formula pretty much put his ZIPS projection for next season right at his FIP from this season (4.87 FIP or 4.77 FIP depending on which site you look at).

Posted: Thu Oct 29, 2009 2:44 pm
by Guardians
Mariners wrote:
Royals wrote:I will take time out of my life to load PECOTA into DMB if it means not using these horseshit projections. I'm no big fan of Rick Porcello - but a 4.89 era?
It's like ZiPS can't stand to think they may have screwed up last year, and has to keep punishing the people they hated who might actually be good.

Looks like their formula pretty much put his ZIPS projection for next season right at his FIP from this season (4.87 FIP or 4.77 FIP depending on which site you look at).
ERA is such a bad statistic to look at. You are correct that when you look at FIP, it appears Porcello would be about the same as 2009 or a slight improvement. When you look at it that way, why is Zips so bad?? Isn't FIP one of the better statistics for determining the effectiveness of a pitcher?

Posted: Thu Oct 29, 2009 3:36 pm
by Cardinals
I haven't looked it lately, but I think at first glance Porcello's WHIP was going to go up by .12 or something in 2010 based on the ZiPS projections. He seems to predict regression instead of progression especially for young players.

Posted: Thu Oct 29, 2009 7:37 pm
by Rangers
Astros wrote:
Mariners wrote:
Royals wrote:I will take time out of my life to load PECOTA into DMB if it means not using these horseshit projections. I'm no big fan of Rick Porcello - but a 4.89 era?
It's like ZiPS can't stand to think they may have screwed up last year, and has to keep punishing the people they hated who might actually be good.

Looks like their formula pretty much put his ZIPS projection for next season right at his FIP from this season (4.87 FIP or 4.77 FIP depending on which site you look at).
ERA is such a bad statistic to look at. You are correct that when you look at FIP, it appears Porcello would be about the same as 2009 or a slight improvement. When you look at it that way, why is Zips so bad?? Isn't FIP one of the better statistics for determining the effectiveness of a pitcher?
I think it depends on how you look at it. FIP can be handy in seeing whether a pitcher's more obvious statistics are misleading, but I'm not sure that when you only consider stats that determine FIP, if they will give you the best idea of what the guy will do in DMB. For example, I have a feeling that hit rates, which FIP obviously ignores altogether, have a pretty large impact on how a guy's projection plays out. Obviously environment will impact them, but I have to think that there is a good deal of weight of the pitcher's own "in play" rates being applied to a play or a game. The popular conjecture is that FIP is a pretty accurate portrayal of future projections, but I'm not sure that it will apply that way in the DMB engine.

Posted: Fri Oct 30, 2009 9:05 pm
by Royals
The problem with sticking PECOTA into the d-base, is that it's not Park-neutral. i.e. you take a guy like pedroia and play him in boston and he's gonna be double dipping at the Fenway Park offensive bonus fountain. The first thing I'm doing with developing projections is making them park-neutral. Number 2 is making sure every player has lefty/righty splits.