Predict the winning pollster

On November 3, 2008, the day before the election, a message board had a discussion about predicting the best pollster.  I wrote:

I was going to start a separate thread, but I'll just post this here.

There will likely be more polls later today and even tomorrow, but I just wanted to post a last overview so we can see how accurate the polling was.

The liberal http://www.electoral-vote.com/ is predicting 353-185. He was wrong in 2004.

The conservative http://www.electionprojection.com/ has the same numbers. He was correct in 2004.

Rove is predicting 338-200: http://www.rove.com/

RCP shows Obama ahead by 7.4%.

Here are the separate polls:

Marist 11/02 - 11/02 Obama +9
FOX News 11/01 - 11/02 Obama +7
NBC News/Wall St. Journal 11/01 - 11/02 Obama +8
Rasmussen Reports 10/31 - 11/02 Obama +6
Gallup 10/31 - 11/02 Obama +11
Diageo/Hotline 10/31 - 11/02 Obama +5
Reuters/C-SPAN/Zogby 10/31 - 11/02 Obama +7
IBD/TIPP 10/30 - 11/02 Obama +5
CBS News 10/31 - 11/02 Obama +9
CNN/Opinion Research 10/30 - 11/01 Obama +7
GWU/Battleground 10/29 - 11/02 Obama +6
Pew Research 10/29 - 11/01 Obama +6
ABC News/Wash Post 10/29 - 11/01 Obama +11

Here are some of the key state polls, also from RCP:

Colorado Obama 5.5
Florida Obama 1.8
Indiana McCain 1.4
Missouri McCain 0.5 (This is sort of the bellwether state and has only been wrong once since 1904: http://en.wikipedia.org/wiki/Bellwether#United_States for a list of similar states)
New Hampshire Obama 10.6 (Kerry won in 2004, Bush in 2000)
Nevada Obama 6.2
North Carolina McCain 0.6
Ohio Obama 3.2
Pennsylvania Obama 7.6
Virginia Obama 4.3

The odds are really bad and the 2004 predictions were much closer. I am still hopeful that as in 2004 the state polls are exaggerating the Democrats lead and also I hope there will be people who get into the booth and realize they cannot possibly trust Obama. Dick Morris thinks a high proportion of those who claim to be undecided will go for McCain, because if you haven't jumped on the bandwagon then you're not going to.

Someone posted a link to a article called "How accurate is the 7-Eleven Presidential Coffee Cup Poll?":

As Election Day fast approaches, pundits, pollsters and politicians are weighing in with their statistics. Yet 7-Eleven, that little store on the corner catering to everyday Americans, may have the hottest poll in the land.

Someone else responded with:

That reminds me -- I don't remember seeing/hearing anything about the Halloween mask "poll."

I wrote:

I'd say it's far off this year:

http://www.7-election.com/

As far as Halloween, I heard that Palin was more popular than Obama. I just did a quick search on it and it appears Obama masks were more popular than McCain.

Someone responded with:

Yeah, President Obama is a scary thought, ain't it?

I shall vote, and stand athwart the tracks of history, shouting "STOP!"

Then I wrote:

I found an interesting post from a blogger who thinks McCain will win.

Someone posted a link to a poll:

About a third of Obama voters want change. Two percent are voting for Obama because he is qualified/experienced.

And said:

Two Percent?, Two Percent!?

I responded with:

And yet apparently that's a negative for Palin.

Someone wrote:

Just for the record, Final Rasmussen, 52 to 46, a 6 point win for Obama. 1% minor party and 1% remain undecided.

So I'll go with a 52/46 win for the O-man -- God save us all.

I responded with:

Good job! What do we owe ya? So why did you expect this outcome?

Then he said:

He was very accurate in 2004 and he had a very steady result for the last month.

Then he posted a link on the most accurate polls:

1T. Rasmussen (11/1-3)**
1T. Pew (10/29-11/1)**
3. YouGov/Polimetrix (10/18-11/1)
4. Harris Interactive (10/20-27)
5. GWU (Lake/Tarrance) (11/2-3)*

In the final result Obama won the popular vote 52.9% to 45.7% and the electoral vote by 365-173.