OK, WTF is with "Battleground Tracking."

[A-Schwarzenegger]

http://www.realclearpolitics.com/RCP_PDF/BG_2-way-ballot-trender.pdf

This doesnt say anything about who they are or how they do what they do.

[A-Schwarzenegger]

1. Hmmm ...

Looks like they did purty good in recent ones...

http://www.tarrance.com/files/Tarrance-Group-Projection.pdf

[DemzRock]

2. I seem to remember Battleground being right wing in previous years...

The above is probably referencing a final result, right? That might be close, but their polls leading up to the election are probably like Gallup's. They are trying to manipulate opinion. Then on the final day, they will give us a more accurate poll so they don't look like complete fools. And they were a little off with 2004. The final was 50.7 vs 48.3. Reminds me how Zogby used to be pretty accurate too. He was one of the few that predicted Gore's popular vote victory. Now Zogby seems like a whore, sadly.

[DemocratSinceBirth]

3. It's Supposed To Be A Bipartisan Poll

It's supposed to be a bipartisan poll with Ed Goeas representing the Republicants and Celinda Lake, a highly respected pollster, representing the Democrats...The poll seems to have a significant Republicant bias...They were wrong in 00 but right in 04...

Anyway, I am going to e-mail Ms. Lake to ask about their methodology and how it is not helpful to have the impramtur of bipartisanship given to a poll with a significant Republican bias...

[DemzRock]

4. Well, again, this is a poll over a month before the election so...

they don't have to be anywhere near right yet. A day or two before the election, they can publish something close and claim they are oh so accurate. Look at Gallup which manipulates internals and swings wildly.

[DemocratSinceBirth]

5. My E-Mail

Dear Ms Lake:

I have been watching the Battle Ground Poll for at least the past three election cycles and have noticed a sampling bias that invariably favors the Republicans. If I remember correctly it was wrong in 00 but right in 04.

I have seen panels with you on it, only days before the presidential elections, and you predict a Democratic win while "your" poll predicts a Republican win.

Right now The Battle Ground Poll has McCain winning 48% -46% while every other poll in the field including three other tracking polls shows Senator Obama winning, ergo:

http://www.realclearpolitics.com/epolls/2008/president/us/general_election_mccain_vs_obama-225.html


http://www.dailykos.com/

I realize some polls are of likely voters and some polls are of registered voters but it's hard to believe that explains all the difference.

Thank you.

Brian

[DemzRock]

6. Should be interesting to see a response from that...

I also notice that realclearpolitics only references the pdf file for battleground hosted locally on realclearpolitic's site, while all the others it references link to an external website. Just seems odd that they would pump out a pdf and not have a website to give information about themselves and track changes in the poll. Seems too drive-by.

http://www.realclearpolitics.com/epolls/2008/president/us/general_election_mccain_vs_obama-225.html


I googled for them... and

http://www.tarrance.com/2-way-ballot-trender-9-22.pdf

doesn't even work. 404 dude.

That is a bit suspicious!

[DemocratSinceBirth]

7. The Battle Ground Poll Is Legit

They just oversample Republicans.

Here's their website:


http://www.tarrance.com/bg.cfm

[DemzRock]

9. Yeah, they are "legit". Not saying that they don't exist...

Still, right now their daily tracking document is not even on their OWN site. Kinda odd, eh?

[Clio the Leo]

8. they also do McCain/Palin crowd counting on the side. NT

[DemzRock]

10. LOL! Good one!