• Security incident: ISF was recently accessed by intruders. Please change your password, and change it anywhere else you used it. Read more

Poll Vault

Status
Not open for further replies.
not_so_new said:
I know it's an outlier but...



I find that hard to believe.

:confused:

I guess on the other hand we should look at Gore / Bush, it was pretty close there in Florida and I don't think Gore was as good a candidate as Obama (which explains why I didn't vote for Gore but will for Obama). On the other hand Gore had the advantage of the sitting president.

I don't know, I just didn't really see Obama making it close in Florida. I guess it's closer than I had in my minds eye.
Click to expand...
Quinnipiac has been out of step with other agencies, usually in Obama's favor, pretty much across the board. Look at PA and OH.
 
chipmunk stew said:
Quinnipiac has been out of step with other agencies, usually in Obama's favor, pretty much across the board. Look at PA and OH.
Click to expand...


They do seem to be showing higher numbers in general for Obama. However, they've been quite accurate in the past. Maybe their methodology is picking up something others aren't. Time will tell.
 
Pookster said:
They do seem to be showing higher numbers in general for Obama. However, they've been quite accurate in the past. Maybe their methodology is picking up something others aren't. Time will tell.
Click to expand...

Cell phone only voters?

I hear conflicting reports on this topic. Some people say that the pollsters are taking non-landline voters into their projections, others say they are not. Even some of the polling organizations say they take this into account and others don't.

I know it was talked about earlier in the thread but is there a definitive answer on the cell phone issue and how it relates to the polling data?

In short are we heading for a Dewey / Truman outcome where the polling data was so off the mark based on bad polling practices?
 
not_so_new said:
Cell phone only voters?

I hear conflicting reports on this topic. Some people say that the pollsters are taking non-landline voters into their projections, others say they are not. Even some of the polling organizations say they take this into account and others don't.

I know it was talked about earlier in the thread but is there a definitive answer on the cell phone issue and how it relates to the polling data?

In short are we heading for a Dewey / Truman outcome where the polling data was so off the mark based on bad polling practices?
Click to expand...


I don't know Quinnipiac's methodology. I do know they're respected and usually quite accurate. Rasmussen doesn't do cell phones and has been one of the most accurate pollsters for many years. Their methodology seems to have overcome the cell phone issue ... up till now at least.
 
not_so_new said:
Cell phone only voters?

I hear conflicting reports on this topic. Some people say that the pollsters are taking non-landline voters into their projections, others say they are not. Even some of the polling organizations say they take this into account and others don't.
Click to expand...
Some pollsters call cell phones, some do not, yes.

I know it was talked about earlier in the thread but is there a definitive answer on the cell phone issue and how it relates to the polling data?
Click to expand...
I believe the issue remains unresolved if omitting cell-phone voters is negatively skewing results. Most pollsters will adjust oversample or otherwise balance demographics in their result set to balance that and other factors, and it seems likely at this point that demographic adjustments will weigh more heavily on pollster accuracy in this election than cell phone polling.

For example, Selzer has been leaning blue because they have higher estimates of youth and black turnout than other pollsters, while Battlegrounds is leaning red and (apparently wildly) underestimating youth turnout.

In short are we heading for a Dewey / Truman outcome where the polling data was so off the mark based on bad polling practices?
Click to expand...
I doubt it - polling practices are much improved in 60 years - pollsters accommodate the voters they know exist but can't get on the phone.
 
David Wong said:
I wish we had some more recent polling in Michigan. Surprised nobody is doing it there.
Click to expand...



McCain is giving up on Michigan, according to AP.

In a major concession, Republican John McCain has abandoned efforts to win Michigan, a Democratic-leaning battleground state the GOP presidential candidate had hoped to capture.

Republican officials with knowledge of the strategy said the GOP nominee is removing staff, curtailing advertising and canceling visits to the Midwestern state, which offers 17 electoral votes. Resources will be sent to Ohio, Wisconsin, Florida and other competitive states.
Click to expand...
 
Last edited:
If that's true, what's McCain thinking in terms of the map? Flipping Pennsylvania? Wisconsin? Minnesota?

Or does he just retreat and defend all the Bush states, since obviously that's enough to win on its own?
 
David Wong said:
If that's true, what's McCain thinking in terms of the map? Flipping Pennsylvania? Wisconsin? Minnesota?

Or does he just retreat and defend all the Bush states, since obviously that's enough to win on its own?
Click to expand...

Just about every pollster has him losing Iowa. There go seven electoral votes from Bush's total, giving McCain 271. Now McCain has to hold every other state that Bush won, but -- according to the polls -- he's trailing in New Mexico, Colorado, Nevada and Virginia, and he's at best close in Florida, Ohio and North Carolina.

He'd better be darn good at defense!
 
The latest from the prediction markets...

IEM-WTA
Obama 70.7
McCain 31.4

IEM-Vote Share
Obama 54.9
McCain 45.4

NewsFutures
Obama 64
McCain 36

Intrade
Obama 66.6
McCain 32.8

McCain is losing ground and looks to be squarely on the defensive. This can not only be evidenced in the prediction markets, but headlines like this are not going to help him at all either...

McCain campaign writes off Michigan

As McCain retreats, Obama advances. Based upon the most recent numbers, it seems that Obama has a better-than-even shot at Ohio, Virginia, Nevada, AND Florida. Yahoo.com's Political Dashboard has a great breakdown based upon an Electoral College prediction market.

If things keep going like this, the election won't be a contest - it'll be a route.
 
Last edited:
not_so_new said:
Cell phone only voters?

I hear conflicting reports on this topic. Some people say that the pollsters are taking non-landline voters into their projections, others say they are not. Even some of the polling organizations say they take this into account and others don't.

I know it was talked about earlier in the thread but is there a definitive answer on the cell phone issue and how it relates to the polling data?

In short are we heading for a Dewey / Truman outcome where the polling data was so off the mark based on bad polling practices?
Click to expand...


Here's a great article about the topic of polling and cellphone users...

Estimating the Cellphone Effect: 2.8 points

The bottom line in that article is that when taking into account the cellphone effect (some people don't have a landline, just a cellphone, so in many polls they don't get called), the poll numbers for Obama rise by an average of 2.8 points. That's huge, especially considering the current poll numbers.

This is because cellphone-only users tend to be younger (the under 30 crowd), and that age demographic heavily favors Obama.

Read the link for all the details. A very interesting analysis.
 
GW/Battleground, which was the lone Mccain outlier a few days ago, now has Obama up by 7.

Edit: RCP mixed the numbers up. Rasmussen has Obama up 7.
 
Last edited:
Mark Blumenthal on the polls of the VP debate.

Biden won, but the VP debate probably won't matter anyway (I suppose it could have if one of them had really embarassed themselves, but they didn't).

Palin barely budged the needle on whether she's ready to be president.
Asked before the debate if Sarah Palin is "qualified to be president," 42% said yes and 54% said no. After the debate, 46% said yes and 53% said no.
Click to expand...
More uncommited voters made up their minds to vote for Obama-Biden than for McCain-Palin.
18% of these originally uncommitted voters made up their minds to support Obama, 10% made up their minds to support McCain, 71% were still uncommitted.
Click to expand...
A whopping 98% of people think that Biden is knowledgable. Palin improved her number to 66% from less than half.
After the debate, 66 percent see Palin as knowledgeable about important issues – up from 43 percent before the debate. But Biden still has the advantage on this – 98 percent saw him as knowledgeable after the debate. That figure was 79 percent before the debate.
Click to expand...

So, it probably doesn't matter much, but if it does push the needle at all, it would probably be toward Obama.

With McCain Palin raising "the white flag of surrender" in Michigan (terrible PR if you ask me) I think it's all but over barring a very, very big October Surprise.
 
Some recent "interesting" State polls ...


Kerry States in play:

Michigan (17)
October 2 - PPP (D) Obama by 10 (51%-41%)
September 25 - Strategic Vision (R): Obama by 3 (48%-45%)
September 25 - Detroit Free Press: Obama by 13 (51%-38%)
September 25 - National Journal/FD: Obama by 8 (47%-39%)
September 25 - NBC/Mason-Dixon: Tied (46%-46%)
September 25 - EPIC-MRA: Obama by 10 (48%-38%)

Minnesota (10)
October 3 - SurveyUSA: McCain by 1 (47%-46%)
October 1 - CNN/Time: Obama by 11 (54%-43%)
September 23 - Quinnipiac/WSJ/WP: Obama by 2 (47%-45%)
September 22 - Rasmussen: Obama by 8 (52%-44%)

New Hampshire (4)
October 3 - Rasmussen: Obama by 10 (53%-43%)
October 3 - St. Anselm/SRBI: Obama by 12 (49%-37%)
September 25 - Research 2000: Obama by 4 (48%-44%)
September 25 - 7News/Suffolk: Obama by 1 (46%-45%)
September 25 - Strategic Vision (R): Obama by 1 (46%-45%)
September 25 - National Journal/FD: Obama by 1 (44%-43%)


Bush States in play:

Colorado (9)
October 2 - Ciruli Assoc.: Obama by 1 (44%-43%)
September 29 - FOX News/Rasmussen: Obama by 1 (49%-48%)
September 24 - CNN/Time: Obama by 4 (51%-47%)
September 24 - Rasmussen: Obama by 3 (50%-47%)
September 24 - InAdv/PollPosition: Obama by 9 (51%-42%)

Missouri (11)
October 1 - CNN/Time: Obama by 1 (49%-48%)
September 26 - SurveyUSA: McCain by 2 (48%-46%)
September 26 - Post-Dispatch/R2000: McCain by 1 (47%-46%)

Nevada (5)
October 1 - CNN/Time: Obama by 4 (51%-47%)
October 1 - InAdv/PollPosition: Obama by 1 (48%-47%)
September 22 - Suffolk University: McCain by 1 (46%-45%)

New Mexico (5)
October 2 - Rasmussen: Obama by 5 (49%-44%)
October 2 - SurveyUSA: Obama by 8 (52%-44%)
September 22 - PPP (D): Obama by 11 (53%-42%)
 
Status
Not open for further replies.

Share this page

ISF - Join now!

Every member here is approved by hand. No bots, no spam, just people who care about evidence and honest debate.

Membership is free!

Create your free account

Back
Top Bottom