Comments on: What’s in a projection system?

By: Gardimentary

Gardimentary — Sun, 21 Mar 2010 13:27:39 +0000

Pecora is useless trash.

Instead of this geek crap, just use what you know of the game and it should be pretty easy to determine that Mariano will save 40 games.

By: Gardimentary

Gardimentary — Sun, 21 Mar 2010 13:26:01 +0000

Pecora useless trash.

By: WIlliam

WIlliam — Sat, 20 Mar 2010 03:53:24 +0000

wow, spelling fail today

By: WIlliam

WIlliam — Sat, 20 Mar 2010 03:53:06 +0000

noot 2ho, 250

By: WIlliam

WIlliam — Sat, 20 Mar 2010 03:52:50 +0000

Posasa didn’t decline that much that a sib-2ho avg and 6 homers were expected.

By: Accent Shallow

Accent Shallow — Fri, 19 Mar 2010 22:39:46 +0000

But those look like reasonable projections after their weak 2008s.

Well, “reasonable” in that “if further decline occurs.” Since both players were playing through injury, it’s not unreasonable to expect them to bounce back after either surgery (Posada) or rest (Jeter).

Since PECOTA doesn’t know that the poor seasons are due to injury, it’ll just think that this is the end of the line.

By: Accent Shallow

Accent Shallow — Fri, 19 Mar 2010 22:33:23 +0000

Actually, I was looking at some neural net software awhile back as a new way to do projection systems, and one of the major application examples was handicapping horse racing.

So maybe it is, and you’re right

By: WIlliam

WIlliam — Fri, 19 Mar 2010 22:20:11 +0000

But still, that does show that while Pecota can be good for 3oish players, anyone above 35 does not collapse, which brings Pecota down.

By: WIlliam

WIlliam — Fri, 19 Mar 2010 22:19:17 +0000

FOund this touring the interweb.

PECOTA projected the following for Jeter and Posada in *2009*:
Jeter: .288/.353/.383
Posada: .249/.336/.406 (7HR and 33RBIs)

‘Nuff Said

/WIldly boversimplified and selective examples

By: Geek

Geek — Fri, 19 Mar 2010 21:43:08 +0000

If the projections systems had validity then handicapping horse races would be a snap.