Good Poll, Bad Poll

There are a LOT of polls out there, folks, but then you knew that. What's worse, even now they don't agree, and if you try to figure out why, they all tend to sound alike in asserting the value in considering everybody. I am almost at the point, however, where I am willing to cross out some of the less reliable ones. I have put up several posts comparing the accuracy and methods of the polls, but I have tried to keep the door open to everyone. I'm not closing the door exactly, but it's time to separate the group into classes. Over on Polipundit, I have addressed one particularly stinky sample.

Here are the players, and their classes:

FROM WORST TO BEST

Partisans, Suspect Methodology

Ayres McHenry (GOP)

Battleground - Lake, Snell, Perry (Dem)
Battleground – The Tarrance Group (GOP)
Democracy Corps (Dem)
New Democratic Network (Dem)
Investor's Business Daily (IBD)/TIPP (Money)

Strategic Vision (GOP)

Zogby/Wall Street Journal (Ego, Money, Dem)

(Just because a poll is partisan, doesn't invalidate it. But you should know going in, that these guys are getting money from partisan sponsors. Also, their refusal to show their internals makes some of these completely unacceptable)
.
.
.
Hides Details, Suspect Methodology

American Research Group (ARG)
Rasmussen

(No details, no confidence. It's that simple)

.

.

.
Unusual Methodology, Unconfirmed
Harris Interactive
Zogby Interactive

(Internet Polling could be useful, but don't take these results as anywhere near as valid as the established polls)

.

.

.

Good Basic Polls
ABC News/Washington Post
Associated Press/Ipsos

Fox News/Opinion Dynamics
Harris
LA Times

Marist College
National Public Radio (NPR)/POS-GQR

NBC News
Quinnipiac University

(Like the name suggests, these guys are pretty good. I'd like more detail, and they sometimes are not as consistent as I'd like, but they're worth reading)

.

.

.
Very Solid Detail and Methodology

CBS News CBS News/New York Times

CNN/USA Today/Gallup
Gallup (solo)
Newsweek

Pew Research
Survey USA

TIME

(Very, very good work. Read these guys first. This is Polling's Major League)

.

.

.

It would also be useful to take a look at some of my related articles:

.

On the reliability of national polls

.
On the Historical accuracy of polls

.
On poll demographics, and the demographic impact on Presidential Elections