New DailyKos/Research 2000 tracking poll

[Brittain33]

Bleh. I remember reading polls with wishful thinking in 2004 and don't have the stomach or desire to do it again this year.

[Brittain33]

Apparently, they over represent Democrats.  So even after that, McCain's tied.

They have McCain up 7 among independents. If that is true, they HAVE to be oversampling Dems in order to get a tie.

If MAC wins Indies by 7 on election day, he wins comfortably.

"Oversampling" and "overrepresenting" are two very different things. Oversampling can mean you're overrepresenting, but it doesn't require it to be so.

[Brittain33]

Bump. Not sure why this isn't stickied yet...

[Brittain33]

have they released their LV model?

I bet they are factoring in higher than usual African American and Youth voting (which may or may not be a good bet). Regardless, they should show their RV numbers and/or reveal their LV model

They seem to be factoring in significantly higher Latino voting, which is an even shakier bet.

[Brittain33]

I find it quite amusing that people care so much about this.  What do you have against it being listed as a DailyKos poll?

The labeling is a petty issue. However, so is the refusal to sticky it. Not sure what that accomplishes...

[Brittain33]

I find it quite amusing that people care so much about this.  What do you have against it being listed as a DailyKos poll?

The labeling is a petty issue. However, so is the refusal to sticky it. Not sure what that accomplishes...

Personally, I think it merits being stickied.  But labeling it "DailyKos" is just worthwhile, basic information that will prevent duplication and reduce confusion.  (i.e., more people know it as the DailyKos tracking poll than the Research 2000 tracking poll.)

I agree, it just doesn't matter.

[Brittain33]

BTTT

[Brittain33]

09/19/2008

Obama   49%
McCain   42%

WTF ever. I'd love to believe it, but this balloon has floated out of the zone of plausibility for me.

[Brittain33]

Now that this has been stickied, can I please ask the Daily Kos kids again to tell their mentor:

DEM 389/1100 = 35.36%
GOP 281/1100 = 25.45%
IND 331/1100 = 30.09%
OTH 100/1100 = 9.09%

Which means the sample is now technically 35D-25R-30I-9O.  Not 35D-26R-30I-9O.

Thank you. 

You know, you're more likely to get a response from e-mailing him directly than I'd get from being the 357th post on the thread, pointing it out. He reads his e-mail, but who knows if he reads the comments. He won't know who you are if you e-mail and mention the change concisely; I really doubt he cross-checks neutral e-mail against the banned e-mail list.

[Brittain33]

Seriously, did you get a chance to e-mail him? He keeps trumpeting how great he is for revealing the internals...

[Brittain33]

This poll's partisan breakdowns are the same as Gallup's, according to this post:

Pollster        D     R     I     (Dem-Rep)
Research 2000:  35    26    30    +9

Yeah, this is the post I alluded to above.