Just so you know, new Newsweek poll oversampled repukes (R36%, D33%, I31%)

[jefferson_dem]

The raw partisan breakout is as follows :

345 Republicans (plus or minus 6)
316 Democrats (plus or minus 6)
304 Independents (plus or minus 6)

http://www.prnewswire.com/cgi-bin/stories.pl?ACCT=104&STORY=/www/story/10-30-2004/0002323285&EDATE=

I'm betting, if we correct to 2000 turnout #s, the race is about tied.

[flowomo]

1. deleted by poster

[Sean Reynolds]

3. 4% margin of error.

Which isn't great either.

[flowomo]

5. right, my mistake

[GreenPoet64]

2. How can they do that ???? It's just not factual. eom.

[jefferson_dem]

4. They just include whoever answers the phone. Repuke were more available,

apparantly. Some pollsters weight respondents to reflect the likely partisan reality within the voting public.

This speaks to a much bigger issue about WHO is participating in these polls. There has been much discussion that those with cell phones only are NOT being polled. And, i would argue that they are more likely to be Democrats. We may be in for a big surprise on Tuesday!

[flowomo]

6. Bloomberg spinning as consequence of OBL tape.... bullshit

the poll was Oct. 27-29, tape out on 29th. But that didn't stop Bloomberg from this headline: Bush Leads by 6 Points in Newsweek Poll After Bin Laden Tape

http://quote.bloomberg.com/apps/news?pid=10000087&sid=aWbqxr7axfwc&refer=top_world_news

[UL_Approved]

7. This is probably everywhere

I'll bet that most polls have not been equally biased amongst participants. No poll done can be perfectly sampled anyway. You need more than (almost) 1000 people to have an accurate result anyway. The real poll is on November 2nd.