>>>This was likely driven by the fact that Obama has expanded his lead among women voters

Amy, do you honestly believe that the female vote is so fluid as to effect the national tracking polls by as much as 5-7% over the course of 24 hrs?

How about maybe, just maybe, your poll yesterday that showed the race tightening was just statistical error or polling error on part of the pollsters?

It bothers me that there is NO mention of possible statistical anomaly or human error. This is a poll with a ton of moving parts. You -- as a journalist -- MUST be open and honest about reasons for such huge swings and not try to pin every statistical change to changes in voters' opinions.

CLEARLY Barack Obama has a 5-7% lead nationally. And nothing has changed that trend since mid-September. Not the debates, not the McCain Muck, not the media, nothing. It is a testament to Barack's campaign that he has held this lead under the most incredible of circumstances -- from the biggest economic crisis in generations (not bad leadership from someone with "no experience") to personal and salacious attacks from a prospective first lady (absolutely reprehensible and totally unheard of).

After the outlier polls of the last couple of days, it looks like Hotline/Diageo is back with a more realistic poll consistent with what the other credible pollsters have found.

In addition to the statistical noise and possible errors that Ethan mentions above, it is also worth pointing out that the partisan composition of the sample has shifted again -- previously, Hotline's sample was 40% Democratic, 38% Republican -- a close partisan divide that no one realistically thinks is a reflection of where the electorate really is. Today the poll shows a more reasonable 41% Democratic, 36% Republican. That difference explains the poll variations as much as thinking that voter preferences are swinging wildly.

Ethan you are a tool of the Un-American left who are trying to destroy the values that have made America great. I know that you will be very sorry on that judgement day when you finally understand the truth and the difference between good and evil.

Ethan you are a tool of the Un-American left who are trying to destroy the values that have made America great. I know that you will be very sorry on that judgement day when you finally understand the truth and the difference between good and evil.

Do all of these people really go back and forth, changing their minds daily on this question? I find it more than little hard to believe.

Look at the statistics on this poll or any poll. Margin of error is 3.4%. Most of these polls are biased. They always give the advantage to Obama. Switch the margin of error around and the numbers are 43.6% Obama and 44.4% McCain. Now McCain is ahead. A good article to read at Americanthinker.com called "Polls: Obama or McCain Winning" by Steven Warshawsky. Very eye opening and educational concerning all of these polls.

Look at the statistics on this poll or any poll. Margin of error is 3.4%. Most of these polls are biased. They always give the advantage to Obama. Switch the margin of error around and the numbers are 43.6% Obama and 44.4% McCain. Now McCain is ahead. A good article to read at Americanthinker.com called "Polls: Obama or McCain Winning" by Steven Warshawsky. Very eye opening and educational concerning all of these polls.

I don't understand how this 3 day tracking poll can go from plus 1 to plus 6 in one day. Mathematically, that is a cumulative swing of 15 points over the 3 days studied. This means, by definition, that yesterday's polling had to be 15 points more in favor of Obama than the polling done 4 days ago, the most recent day to be dropped from the 3 day tracking. Is this possible?? I know Obama won the debate on Tuesday night, but a swing of this magnitude is NOT SUPPOSED to happen in a tracking poll which smooths the variations by averaging 3 successive days. Any thoughts??

The answer is simple Dave, diageo was using a +2 Dem split in party preference the past few days and now has returned to their normal +5 Dem split (the correct party split in a year like this, although it could very well be a larger split). So you are right, while the other polls have started to show slight movement to Obama after the debate in polls from today, it hasn't moved their 3 day averages yet, and certainly hasn't moved individual groups the way this poll has. I would be more inclined to believe today's results then i would the last few days since the split's were wrong earlier in the week.

The answer is simple Dave, diageo was using a +2 Dem split in party preference the past few days and now has returned to their normal +5 Dem split (the correct party split in a year like this, although it could very well be a larger split). So you are right, while the other polls have started to show slight movement to Obama after the debate in polls from today, it hasn't moved their 3 day averages yet, and certainly hasn't moved individual groups the way this poll has. I would be more inclined to believe today's results then i would the last few days since the split's were wrong earlier in the week.

The answer is simple Dave, diageo was using a +2 Dem split in party preference the past few days and now has returned to their normal +5 Dem split (the correct party split in a year like this, although it could very well be a larger split). So you are right, while the other polls have started to show slight movement to Obama after the debate in polls from today, it hasn't moved their 3 day averages yet, and certainly hasn't moved individual groups the way this poll has. I would be more inclined to believe today's results then i would the last few days since the split's were wrong earlier in the week.

Thanks Connor. That would explain 9 of the 15 point swing (3 points each day for adjusting % of Democrats up by 3 percent daily = 9 points). The remaining 6 point swing must be due to dropping a bad day for Obama 4 days ago, and adding a good day for Obama yesterday. Is there any reason why they were using a +2 for Democrats, and when did they start doing that representation?

By the way, Obama up by 8% in WEST VIRGINIA, and up by 3% in Ohio according to state polls released today by ARG! Could West Virginia be in play? No recent state polls to compare with for that state. Any thoughts>

TPM's tracker composite for today:

Obama is ahead 49.9%-42.6%, compared to a 49.7%-43.2% Obama lead yesterday.

• Gallup: Obama 52%, McCain 41%, with a ±2% margin of error, unchanged from yesterday. At 11 points, this is Obama's widest lead in the Gallup poll for this whole campaign so far.

• Rasmussen: Obama 50%, McCain 45%, with a ±2% margin of error, compared to a 51%-45% Obama lead yesterday.

• Hotline/Diageo: Obama 47%, McCain 41%, with a ±3.3% margin of error, compared to a 45%-44% Obama lead yesterday.

• Research 2000: Obama 51%, McCain 41%, with a ±3% margin of error, unchanged from yesterday.

• Zogby: Obama 48%, McCain 44%, with a ±2.9% margin of error, compared to a 47%-45% Obama lead yesterday.

So going into the debate -- that most agree Obama won handily -- Barack was still edging up and McCain still slumping.

Lee Ann,
You want us to read an article about bias polls at Americanthinker.com? You really should come out of your right wing bubble more. If all the polls are showing an Obama lead just inside or outside the margin of error, you can bet your conservative soul on that Obama is ahead.