NationStates Jolt Archive

We have a lot of political discussion on this forum, but I want to make this thread specifically about the POLLS that are coming out, on a day to day basis, both nationally and internationally. My plan is to create a post a day, each day, before the election. I'd like to, if possible, keep the chatter about candidates themselves down a notch, and make this about the actual polling data. I've been posting this information in the general thread, but it's gotten so unwieldy, this info might be lost, and I thought people might like to see it as it comes out.

The posts for days since I began, in descending order:

October 24 (http://forums.jolt.co.uk/showpost.php?p=14131871&postcount=44) (11 days of campaigning before the election)
October 23 (http://forums.jolt.co.uk/showpost.php?p=14128784&postcount=24) (12 days of campaigning before the election)
October 22 (http://forums.jolt.co.uk/showpost.php?p=14125600&postcount=3) (13 days of campaigning before the election)
October 21 (http://forums.jolt.co.uk/showpost.php?p=14125588&postcount=1) (14 days of campaigning before the election)

I include as way of sample, the post I made YESTERDAY, so this post is for YESTERDAY'S numbers:

By and large a good day for Obama. McCain is up in one poll (Battlegrounds), Obama is up in four (Zogby, Gallup, Pew, and TIPP) and 3 more (Rasmussen, Reasearch 2000, and ABC/Post) have remained unchanged.

However, looking at a few of them, he's leading as high as +14 in the Pew poll, but as low as +1 in Battlegrounds (the hell?). Presuming these are both extreme ends of the outlier spectrum, and we can ignore them, means Obama raised his standing in 3 polls today, and stayed the same in 3.

Among Rasmussen, Research 2000, and ABC/Post, he is unchanged at +4, +9 and +9 respectively (Research 2000 and ABC/Post were at the high end of the Obama spectrum last night, the fact that they're unchanged today, while still calling their numbers questionable, at least suggests the poll is reliable enough to not suffer rapid changes

In the polls he has raised in, Zogby, Gallup, and TIPP (ignoring the outlandish +14 of Pew) he is now at +8, +10 and +6. Thus other than the somewhat conservative Rasmussen (+4) and the outlandish Battleground (+1), no poll has him below a 6 point lead, and that seems to be about where he's settled.

Another reason not ot trust battleground is they're a 5 day polling methodology that skips weekends. In yesterday's poll Obama was at +4. Yesterday's poll results covered October 13, 14, 15, 16, 19. Todays covered 14, 15, 16, 19, 20. Between yesterday and today the poll changed 3 points.

Both polls utilized the exact same data on the 14, 15, 16, and 19. Today's lost 13 and added 20. Which suggests that either Obama polled HUGELY well on the 13th, or McCain polled HUGELY well on the 20th. In fact, to effectuate a 3 point change over 1 day, the difference between the 13th and the 20th would have to be on the order of McCain being about 15 points higher on the 20th than on the 13th. Since NOBODY has seen ANYTHING like this happening, and since, prior to this, Battlegrounds was fairly accurate (suggesting that the 13th was not an unusually pro Obama day) it suggests that their polling on the 20th was a statistical abnormality of unusually high McCain support. It will, unfortunately skew Battlegrounds for the rest of the week, and shoot their reliability to hell.

As for state polls, nothing too shocking, and generally good for Obama. McCain is leading in pretty much all the plaes you'd expect him to lead. But Obama has narrow leads in important places. Colorado is at +5, and Florida (!) at +1. Indiana is at a razor thin +2 (seriously, what's with Obama tying up these red states? if you watch the electoral map states start going from red paler and paler, and when one starts to flip into blue, the one NEXT to it starts going pale. It's like a damned virus). Where three polls out of North Carolina (fucking hell, north carolina of all places?) show him at +3, +1, and tied, respectively.

Poll also shows nevada a tie. Booo.

Encouraging is a posted +10 lead in PA. If that's accurate and McCain wants to flip PA, assuming a 3 point Bradly factor in his favor, that's still half a point a day, every day, from now until the election. Not likely.

i thought tracks on poles were called monorails, or at least elevateds

(also, i try to avoid what the op seems to suggest, because i see this as a great way to jinx whatever one might happen to be for)

(do football pools jinx whatever team you bet on? my superstition is that they do)

Today is Wednesday, October 22nd. There are 13 days of campaigning left before election day:

As for national polls:

Zogby: Obama +9.6 (up 1.7 from yesterday 51.6/42 yesterday was 50.3/42.4)
Research 2000: Obama +10 (up 1 from yesterday. 51/41 yesterday was 51/42)
Rasmussen: Obama +6 (up 2 from yesterday. 51/45 yesterday 50/46)
Battleground: Obama +2 (up 1 from yesterday. 49/47 yesterday 48/47)
Diego/Hotline: Obama +5 (down 1 from yesterday. 47/42, yesterday 47/41)
Gallup: Obama +8 (down 2 from yesterday. 52/44, yesterday 52/42)
TIPP: Obama +4 (down 2 from yesterday)
Fox News: Obama +9 (can't find previous data)
ABC/Washington Post: Obama +11 (54% to 43%)

Last updated 3:30 PM. Please see yesterday's post about my concern with Battleground polls.

Of the polls published thus far, every one shows Barack Obama in the lead, and five out of eight show a gain for Obama, With the exception of Battleground, the lowest lead for Obama is +4 and three of the eight polls show him leading +8 or greater.

STATE POLLS RELEASED TODAY:

CNN has a series of battleground polls:
North Carolina (really? North Carolina?): Obama + 4 51% to 47%
Nevada: Obama +5 (51% to 46%)
Ohio: Obama +4 (50% to 46%)
Virginia: Obama +10 (54% to 44%)
West Virginia: McCain +12 (41% to 53%)

Those are some serious numbers for Obama, especially the +10% in Virginia, considering the CNN electoral map shows that if he keeps all the states he's projected to win in, plus Virginia, he wins.

Last update 4:40

Today is Wednesday, October 22nd. There are 13 days of campaigning left before election day


Yay, thirteen more days until the overtly long campaigns are over!

update 10/22 1:15 PM : Gallup and Diageo/hotline added.

Today is Wednesday, October 22nd. There are 13 days of campaigning left before election day:

As for national polls:

Zogby: Obama +9.6 (up 1.7 from yesterday 51.6/42 yesterday was 50.3/42.4)
Research 2000: Obama +10 (up 1 from yesterday. 51/41 yesterday was 51/42)
Rasmussen: Obama +6 (up 2 from yesterday. 51/45 yesterday 50/46)
Battleground: Obama +2 (up 1 from yesterday. 49/47 yesterday 48/47)
Diego/Hotline: Obama +5 (down 1 from yesterday. 47/42, yesterday 47/41)
Gallup: Obama +8 (down 2 from yesterday. 52/44, yesterday 52/42)

That's it so far at 1 PM. Please see yesterday's post about my concern with Battleground polls.


If you go bac (http://www.tarrance.com/files/TTG-Vote-Projections-and-Results.pdf)k and look at the history of the Battleground poll, they are amazingly accurate. The worst was in 2000, when they predicted Bush and Gore edged him out, but otherwise, they're within a couple tenths, or so. It's not easy to track them against the other polls at the time, but their end results seem to be pretty good.

Y'all do realize that none of these polls are just numbers. They're weighted and massaged enough to make anyone have a doubt or two.

If you go bac (http://www.tarrance.com/files/TTG-Vote-Projections-and-Results.pdf)k and look at the history of the Battleground poll, they are amazingly accurate. The worst was in 2000, when they predicted Bush and Gore edged him out, but otherwise, they're within a couple tenths, or so. It's not easy to track them against the other polls at the time, but their end results seem to be pretty good.

I think Battleground's numbers are being skewed by one, single, very heavy McCain day that is a statistical anomaly. It's the only way to explain a 4 point shift in one day in a 5 day tracking poll.

Y'all do realize that none of these polls are just numbers. They're weighted and massaged enough to make anyone have a doubt or two.

They are what they are. I make no claims that one is any more accurate than the other, I merely report the numbers, and why I think they are the way they are.

According to the polls this election is over anyways.

According to the polls this election is over anyways.

over? no. Two weeks left. The polls DO tell us one thing is certain. We will either have the first black president in history, or the biggest come back to win the race in history.

I think Battleground's numbers are being skewed by one, single, very heavy McCain day that is a statistical anomaly. It's the only way to explain a 4 point shift in one day in a 5 day tracking poll.



They are what they are. I make no claims that one is any more accurate than the other, I merely report the numbers, and why I think they are the way they are.
The latest AP-Gfk (http://www.ap-gfkpoll.com/) poll shows a 44-43 lead for Obama. I'm inclined to believe that the race is tightening.

The latest AP-Gfk (http://www.ap-gfkpoll.com/) poll shows a 44-43 lead for Obama. I'm inclined to believe that the race is tightening.

You know what? Good. I would be inclined to argue with you, that you're stretching for the minority of polls when it's far more likely that the majority of polls grouped tightly more resemble the actual numbers (being generally how statistics works), that maybe you're not considering the polls- (http://www.fivethirtyeight.com/2008/10/tracking-poll-primer.html)
Battleground / Tarrance / George Washington U.

When Publishes: Weekdays only -- midmornings around 10:30 AM Eastern time.

Key Specifications: Either 800 or 1,000 likely voters, spanning either 4 or 5 days respectively.

Track Record: Battleground is usually in hibernation, emerging only once every four years to produce national trial heat numbers. They've been around the block, however, and had great years in 1992 and 1996, followed by OK years in 2000 and 2004.

House Effect/Lean: Prior to a methodological change that they made a couple of weeks ago, had a very prounced (3-4 point) Republican lean. Since then, they've been fairly neutral.

Features/Strengths: Co-operated by a Republican (Ed Goaes) and Democratic (Celinda Lake) pollster, so to the extent they have a house effect, it isn't the result of partisan bias. Battleground publishes more cross-tabular detail than any other tracking poll, although the formatting is a bit cumbersome.

Quirks/Concerns: They don't conduct interviews on Fridays or Saturdays, which is a defensible decision, but means that the poll may be slower to react to new information early in the work week, especially when coupled with the relatively large (4-5 day) sample window.

As alluded to above, Battleground previously had significant problems with their sample, substantially undercounting young voters, which was likely causing the strong Republican lean. They have since revisited their turnout assumptions and altered their model, and have had more "normal" looking results since.
(no primer on AP-Gfk)

But instead, I'll say good. First of all, the race is bound to tighten. It happens all the time. Second, I want it to tighten because I don't want Obama's support to get complacent and cruise into the last two week, because if there is a way for him to lose it's for his broad support to think that he doesn't need them any more.

The latest AP-Gfk (http://www.ap-gfkpoll.com/) poll shows a 44-43 lead for Obama. I'm inclined to believe that the race is tightening.

It's very funny you mention that poll. Let's actually look at the full poll. the AP poll shows a +5 lead for Obama among their poll of registered voters. +5, not +1. The +1 for Obama result comes from the poll of likely voters. Which is to say, only those who AP deems likely to vote, by their own determination. That means, if it's true, that even though Obama leads +5% among REGISTERED voters, he only leads +1% among LIKELY voters? That would mean that registered voters, favoring Obama, are less likely (by far) to vote than registered voters favoring McCain.

In fact, funny enough, fivethirtyeight.com has, RIGHT NOW on the front page, an article discussing this very problem (I'll just give the highlights):

Frankly, I find polls showing a 4-6 point gap between likely and registered voters to be utterly ridiculous. Why?

1. Among people who have already voted, Democrats lead overwhelmingly. Zogby pegs Barack Obama's advantage at 27 points among people who have already voted.

. . .

2. Enthusiasm is much higher among Democrats than among Republicans. The latest Diageo/Hotline numbers show that 72 percent of Democrats are enthusiastic about voting for their candidate, as opposed to 55 percent of Republicans.

3. Most likely voter models are unlikely to distinguish newly registered voters from what I would call lapsed registered voters.

. . . .

4. There is an enormous discrepancy in the strength of the Republican and Democratic turnout operations.

. . . .

5. Turnout among 'unlikely' voter blocks was substantially up during the Democratic primaries.

. . . .



Essentially ,quoting this thread is not just cherry picking, it's cherry picking the already cherry picked. It takes one poll out of half a dozen showing that small a margin, and then, as if this wasn't bad enough, it only shows their poll of likely voters, not registered voters, and does nothing to explain away the bizarre implication that registered voters favoring Obama are actually considerably less likely to get out and vote than registered voters favoring McCain.

Todays Gallup poll is 51/42 for registered voters.

It's very funny you mention that poll. Let's actually look at the full poll. the AP poll shows a +5 lead for Obama among their poll of registered voters. +5, not +1. The +1 for Obama result comes from the poll of likely voters. Which is to say, only those who AP deems likely to vote, by their own determination. That means, if it's true, that even though Obama leads +5% among REGISTERED voters, he only leads +1% among LIKELY voters? That would mean that registered voters, favoring Obama, are less likely (by far) to vote than registered voters favoring McCain.

In fact, funny enough, fivethirtyeight.com has, RIGHT NOW on the front page, an article discussing this very problem (I'll just give the highlights):



Essentially ,quoting this thread is not just cherry picking, it's cherry picking the already cherry picked. It takes one poll out of half a dozen showing that small a margin, and then, as if this wasn't bad enough, it only shows their poll of likely voters, not registered voters, and does nothing to explain away the bizarre implication that registered voters favoring Obama are actually considerably less likely to get out and vote than registered voters favoring McCain.
You know what the great part of all this is? I don't have to say anything more for two weeks. At that point, one or the other of the misfits will have won the beauty contest and I'll be able to look back to the polls that used likely voters and see that they were right. Or I'll see that those polls were wrong. Either way, there will be a clear answer.

But aside from all that, using likely voters and determining who they are is part of the special sauce that makes a particular polling organization better or worse than any other. We do know that only a fraction -- about half last time -- of registered voters do actually vote. I suspect that the weighting is a little heavier than that for this election, but still, if 80% of all registered voters actually vote, it will be a miracle.

And the explanation for "...the bizarre implication that registered voters favoring Obama are actually considerably less likely to get out and vote than registered voters favoring McCain" is probably due to demographics and the statistical probability that certain groups vote in higher percentages than others.

Tipp poll posted for 10/13, Obama +4 (down from +6 yesterday)

But aside from all that, using likely voters and determining who they are is part of the special sauce that makes a particular polling organization better or worse than any other. We do know that only a fraction -- about half last time -- of registered voters do actually vote. I suspect that the weighting is a little heavier than that for this election, but still, if 80% of all registered voters actually vote, it will be a miracle.

"only a fraction" of registered voters voting shouldn't have this effect. It's not about that. If you poll registered voters, you get a certain %. But if you factor out the people who probably won't vote, theoretically, you're looking at what the election will more closely look like. The problem with that is, unless there is something pushing more groups towards or away from the polls, you have to think that who don't vote, from both demographisc, should be roughly the same.

To suggest that Obama leads +5% in all voters, but only +1% in those who are likely to vote, requires there to be a very significant factor that's stopping Obama voters from voting, and these polls don't say what that is in any way.

edit: especially in a survey that lists 44% of the surveyed as "born again christian"

That means, if it's true, that even though Obama leads +5% among REGISTERED voters, he only leads +1% among LIKELY voters? That would mean that registered voters, favoring Obama, are less likely (by far) to vote than registered voters favoring McCain.

you know, when i think about who is most likely to turn out in mass for this election, my thoughts immediately go to older white people from the suburbs (white total 75%, likely voter 81%; suburbs total 38%, likely voter 44%)

you know, when i think about who is most likely to turn out in mass for this election, my thoughts immediately go to older white people from the suburbs (white total 75%, likely voter 81%; suburbs total 38%, likely voter 44%)

are you being sarcastic or serious? I can't tell.

edit: 3:30 added fox news poll, Obama +9

"only a fraction" of registered voters voting shouldn't have this effect. It's not about that. If you poll registered voters, you get a certain %. But if you factor out the people who probably won't vote, theoretically, you're looking at what the election will more closely look like. The problem with that is, unless there is something pushing more groups towards or away from the polls, you have to think that who don't vote, from both demographisc, should be roughly the same.

To suggest that Obama leads +5% in all voters, but only +1% in those who are likely to vote, requires there to be a very significant factor that's stopping Obama voters from voting, and these polls don't say what that is in any way.

edit: especially in a survey that lists 44% of the surveyed as "born again christian"
Like I said, we shall see... It would be interesting to look at some Zogby or Gallup polls from the same period before election day, 2004, 2000, 1996, or 1992 and just see what the split between registered voters and likely voters was. Not interesting enough for ME to do it, but certainly enough for someone else to put forth the effort.

Or you can look at the differences between registered and likely voters in the 2004 elections at this site
http://www.pollingreport2.com/wh2004a.htm I see about 3-4 points difference in similar polls on similar days.

are you being sarcastic or serious? I can't tell.

when i look back on this election season, it seems to me that those who have been disproportionately enthusiastic have been the mcmansion set

Update: 4:45 October 22nd. CNN has released poll of crucial swing states, details updated here (http://forums.jolt.co.uk/showpost.php?p=14125600&postcount=3)

holy fucking shit, now that's what i call organizing:
http://www.abcnews.go.com/PollingUnit/Politics/story?id=6079714&page=1

Four years ago first-timers backed Democratic nominee John Kerry by 7 points.
Today they favor Obama over John McCain – by 47.

Indeed it’s first-timers who give Obama his clear advantage in presidential preference.
Among people who’ve voted previously, Obama and McCain are at 50-47 percent in this
ABC News/Washington Post poll, not a statistically significant gap. But first-time voters
favor Obama by a lopsided 73-26 percent, lifting him to an overall advantage.

if 80% of all registered voters actually vote, it will be a miracle.
A 70% turnout this time would not be surprising. Suppose that 50 of that 70 is old voters split 50-47 for Obama, and 20 out of 70 are first-time voters split 73-26 for Obama, like FreeSoviets said: that would make it 56-41, a runaway landslide with McCain losing states he never thought were even threatened.

Today is October 23rd, there are 12 days left in the campaign.

NATIONAL POLLS

Zogby: Obama +11.9 (up 2.3 from yesterday. Now 52.2%/40.3%, was 51.6%/42%)
Research 2000: Obama +10 (unchanged from yesterday's 51%/41%)
Battleground: Obama +4 (up 2 from yesterday)
Rasmussen: Obama +7 (up 1 from yesterday. Now 52%/45% was 51%/45%)
Hotline: Obama +5 (unchanged)
Gallup: Obama +6 (down 2 from yesterday. 51%/45% from yesterday's 52/44)
ABC: Obama +11 (54/43 unchanged from yesterday)


Of the 6 polls, Obama is up in 3, down in 1, and 4 remain changed. Research 2000 shows no change from its 10 point prediction, nor ABC from its 11. Battleground remains low, however from previous posts you may have noted some oddness about this poll, including a particularly large, 5 day sampling spread. 2 point gain in a 5 day spread means that the day coming on had to have been about 10 points higher for Obama than the day going off. I also think Battlegrounds will see a big jump early next week as what I get the feeling is a statistically abnormally high day for McCain drops off the charts.

This is also the highest margin Obama has gotten in a Rasmussen poll I believe.

That's all as of 6:00 PM

What's interesting today is how the national polls all seem to be closing in on a clump. A range from 4-12 all just about fits into a 3.5 margin of error. and all really seems to support an Obama +7ish field out there.

I got bugged by a poll last night. I answered it though. heh. Obama, Obama, Obama. lol :p

12:40 PM 10/23: Added Battlegrounds

12:40 PM 10/23: Added Battlegrounds

I think Murtha is fucking up Pennsylvania for Obama.

Obama still wins the election, but PA dumps Murtha, and in the process hands PA to McCain.

I think Murtha is fucking up Pennsylvania for Obama.

Obama still wins the election, but PA dumps Murtha, and in the process hands PA to McCain.

Doubt it, polls are still coming in strong for Obama, state polling published today has him +10%. It will narrow, but I think obama keeps PA

Doubt it, polls are still coming in strong for Obama, state polling published today has him +10%. It will narrow, but I think obama keeps PA

can't say the same for Murtha... his numbers have slid considerably...

4:20 - added hotline

Doubt it, polls are still coming in strong for Obama, state polling published today has him +10%. It will narrow, but I think obama keeps PA

I HOPE he does.

4:20 - added hotline
what you update you might want to include a link to the post that its updating.... some days it might get so busy in this thread that the daily stats are hard to find.

4:20 - added hotline
Heheh...I add 'hotlines' at 4:20, too...

what you update you might want to include a link to the post that its updating.... some days it might get so busy in this thread that the daily stats are hard to find.

well the link to the day's post are all in the first post of the thread, but fair enough, I'll try to do that from now on.

well the link to the day's post are all in the first post of the thread, but fair enough, I'll try to do that from now on.
ohhhh

i never look at the OP again so i didnt notice that you were doing that.

thanks.

6:00 PM - October 23rd post (http://forums.jolt.co.uk/showpost.php?p=14128784&postcount=24) updated: ABC/Washington Post added, Obama +11 (unchanged)

I see Obama is sinking like a rock in the TIPP (http://www.ibdeditorials.com/Polls.aspx?id=309635713550536) poll. 44.8% to 43.7%. It may be worth watching the results come in on election day, after all.

I see Obama is sinking like a rock in the TIPP (http://www.ibdeditorials.com/Polls.aspx?id=309635713550536) poll. 44.8% to 43.7%. It may be worth watching the results come in on election day, after all.
After looking at all the past polls on that site, I can safely say you have no idea what the fuck you are talking about.

Comparing todays EV prediction
http://www.electoral-vote.com/evp2008/Pres/Pngs/Oct24.png

With yesterdays
http://www.electoral-vote.com/evp2008/Pres/Pngs/Oct23.png

We see that McCain's lead in Montana is now zero, Florida went back to Obama, and Indiana is looking very blue, from barely red.

If we compare this with some random day in September (22nd)
http://www.electoral-vote.com/evp2008/Pres/Pngs/Sep22.png

We see a grim looking future for McCain. Over the course of a month, he has lost Florida, Montana, Indiana, North Dakota, Missouri, Nevada and North Carolina. While states like Michigan, Pennsylvania, Maine, Minnesota and Wisconsin, which were once within McCain's reach, are now well and truely Obama's. Meanwhile, North Dakota and Montana, once strong republican states, have been drawn down to a tie.

Where McCain was once predicted to lose by only 8 electoral votes, he is now predicted to lose by 218 electoral votes (with 6 ties, swinging strongly to Obama) - more than the number of electoral votes he is expected to win!

Anyone saying McCain has even a chance of hope is delusional. By every good poll available, and by an average accross those polls, Obama is winning, by a lot.

After looking at all the past polls on that site, I can safely say you have no idea what the fuck you are talking about.

They show 74% of young voters voting McCain. Yeeahh....

http://www.fivethirtyeight.com/2008/10/whats-wrong-with-this-picture-aka-nate.html

I see Obama is sinking like a rock in the TIPP (http://www.ibdeditorials.com/Polls.aspx?id=309635713550536) poll. 44.8% to 43.7%. It may be worth watching the results come in on election day, after all.

indeed. make sure to tell all your republican friends. have a big party, even.

I do not ,and will not, include the TIPP poll for the serious errors it seems to have. The youth vote alone is competely backwards, 74% for McCain is obscene and fundamentally against every piece of common sense.

McCain must have won the youth over with his dancing skills.

Today is Friday October 24th, there are 11 days left of campaigning

NATIONAL POLLS

Zogby: Obama +10.3% (down 1.6 from yesterday, 51.3%/41% was 52.2%/40.3%)
Research 2000: Obama +12 (up 2 from yesterday. 52%/40% was 51%/41%)
Battlegrounds: Obama +3 (down 1 from yesterday, 49%/46% was 49%/45%)
Rasmussen: Obama +7 (unchanged since Wednesday, 52%/45%)
Hotline: Obama +7 (up 2 from yesterday, 50%/43% was 48%/43%)
Gallup: Obama +7 (up 1 from yesterday, 51%/44% from yesterday's 51%/45%
TIPP: Obama +3.5 (up 2.4 from yesterday)
ABC: Obama +9 (down 2 from yesterday 53%/44% was 54%/43%)


So in 8 polls, Obama up in four, down in 3, unchanged in the 8th. Battlegrounds is still iffy, I still feel that there's a heavy McCain day in there which should be expunged around Tuesday, so we'll see a gain there, I predict. TIPP also had some severe methodological problems exposed on fivethirtyeight.com last night, where they had weighed the youth vote in ways that seem directly opposite current trends. It's interesting that the "low" end of the poll both seem to be sceptical, but I'm willing to let that appear my bias, and just call them the outliers rather than actual bad math.

On the other hand Research 2000 I'm skeptical about, merely because anything in any way affiliated with daily KOS starts SCREAMING partisan in my head, even if it is truly neutral. Zogby going down isn't really that big a deal, the Obama +12 result of yesterday was obviously a bit high, and I'm tending to discount zogby and Research 2k as being too high, just as Battlegrounds seems, to me, to be a bit too low. What's funny is, once you account for a margin of error, Rasmussen seems just about right, at 7%, which is in its 3rd day of not budging.

Meanwhile, at the Hotline poll, in the previous 4 polls, Obama was at +5 for three of them, and +6 at one (22nd) now shows him +7, erasing the incremental gain, while actually lowering undecided voters to the point that, if these polls were accurate, McCain could capture all undecided votes and STILL not make it).

Gallup also saw a slight gain today, almost nullifying the small gains McCain made yesterday. What is exceptionally telling is that of 6 polls, 3 of them are dead tied at 7, and if we take out the slanted Research 2000, we're left with 10 points at Zogby and 3 points at Battlefield, which, with respective margins of error, seem to point to the middle ground being, yup, right around at 7.

The erratic nature of the recent polling seems to be converging on a stable, sizeable Obama lead, despite predictions that a race would be "tightening"

STATE POLLS

Some very freaky movement yesterday, and some that seems to push this direction more in Obama's favor.

Florida: Miami Herald and Quinnipiac have +7 and +5 for Obama respectively

Indiana: +10 from Big Ten; +4 from SurveyUSA

Ohio: +11 and +14 from Big Ten and Quinnipiac

Pennsylvania: +10 from Big Ten, Morning Call, and National Journal, +13 from Qunnipiac and +12 from Survey USA

North Carolina: +2 from WSOC-TV

Florida, Ohio, and Pennsylvania are all MUST wins for McCain, he pretty much needs to take all 3 to win, these polls suggest that this will be an issue for him, and all his campaigning in Pennsylvania doesn't seem like it can close the deal on a 10 point lead, 11 days from the election. Moreover, even if Ohio and florida go his way, Obama holding on to Pennsylvania is likely to be a death sentence for McCain's campaign.


Add to that a surprising lead in what was thought to be reliably red Indiana, and just a narrow Obama lead in north carolina, and it looks like McCain is not only not sealing the deal in the crucial states, he's starting to lose what looked like somewhat reliable red states.

Add to that a mere 5 point lead for McCain in GEORGIA, and Obama is on serious track to win. The problem McCain is having right now is, the election momentum seems to have largely stalled for BOTH candidates, and when the polls stagnate and don't fluctuate, Obama's lead stays the same, and McCain has less and less time to retract it. It now seems he has a 7 point deficit and 11 days to close it. That's almost 2/3 of a point a day every single day from now til the election, including TWO weekends, where news tends to be slower.

McCain needs one of two things. Either something earth shattering to happen next week, or for the vast bulk of the polls to be horribly, horribly wrong

Update: 3:45 PM, added Gallup and Hotline

Updated 6pm: Added ABC and TIPP

Updated 6pm: Added ABC and TIPP

I thought you weren't including TIPP. Any reason for the shift?

at this point Obama will win unless he is caught with a dead hooker in a hotel room.

at this point Obama will win unless he is caught with a dead hooker in a hotel room.

Or if we experience a major terrorist attack.

I could see the nuking of Indianapolis. Major city with lots of people, but not extremely important to the economy like Chicago, New York, San Francisco, et al would be.

That, I think, is probably what would be required to let McCain turn things around.

Either that or massive, extremely obvious fraud, which ends up causing huge riots...

Either that or massive, extremely obvious fraud, which ends up causing huge riots...

I doubt it.

I doubt it.

Really? All signs point to a clear Barack Obama victory at this point. In order for McCain to win, either everyone has to suddenly change their mind, or there has to be massive fraud.

Said fraud would not only be patently obvious, it would piss people off after the THIRD election in a row with serious problems with the voting, and unlike the other two, this one would be seen by everyone as obviously stolen.

Couple that with the amount of excitement people have over Obama, and how angry they'd feel about it, plus everything else, I don't see how we could AVOID mass riots.

I doubt it.

You present your case so convincingly.

I thought you weren't including TIPP. Any reason for the shift?

Intellectual honesty. I figure, this is a thread for reporting poll results. I might not agree with it, I might think their methodology is flawed, I might think that they're biased, but it's still a poll, and this is a poll thread, so I include it to allow the readers to make their own determination.

Today is Saturday, October 25th, there are 10 days of campaigning left before the election

OK folks, here's the deal, keep in mind that TIPP hasn't come out yet, and Battlegrounds doesn't publish on the weekend, and those are the two that have Obama generally lower than the rest, so when you view these, keep that in mind.

With that being said...here's today's polls

Zogby: Obama + 9.5 (down .8% from yesterday. 51.1%/41.6% from 51.3%/41%)
Research 2000: Obama +12 (unchanged 52%/40%)
Battlegrounds: Does not publish on weekend
Rasmussen: Obama +8 (up 1 from yesterday 52%/44% was 52%/45%)
Hotline: Obama +7 (unchanged from yesterday 50%/43%)
Gallup: Obama +8 (up 1 from yesterday. 51%/43% was 51%/44%)

That's it by 2pm. Now that's...pretty good for Obama. Now again as I said, this isn't showing polls that tend to be lower for him but....Rasmussen AND Gallup +8? Zogby +9.5? OK, sure, Zogby went "down" but at less than a point, it's more likely statistical noise than anything else.

I usually discount high polls like Research 2000, especially with its Daily KOS affiliation, but it's getting harder and harder to do that when they're not far off the mainstream.

Rasmussen is especially telling of the state of the race. There were some tlak of predicted "tightening" of the race, but that doesn't seem to be the case at a national level. For a full month now, McCain has been within 44% and 46% and Obama has been between 50% and 52%. At no point in the last month has either candidate moved outside their narrow 2 point area. Now, true, that area represents a full spread anywhere from 4 point difference to an 8 point difference, but if anything, Obama has been IMPROVING, since only 8 days ago Rasmussen had him at +4. Gallup is similar, showing him at only +4 on the 16th, but moving to +8 now.

Evidence that the race was tightening SEEMS to be running in reverse now, with Obama now expanding his lead in the final run to the election. If McCain doesn't do something drastic SOON, he will VERY shortly find himself in the position of having more points in deficit then days until the election.

At which point, he'll be pretty much done.

I have been letting this post slack a bit, in fact it's been exactly a week since I posted. Which makes for a perfect opportunity to compare.

About a week ago, people were saying McCain NEEDED a strong week, well, let's see if he got it. I'll post today's results, compared to one week ago:

Zogby: Obama +7 (last week Obama +10
Research 2000: Obama +6 (last week Obama +12)
Battlegrounds: Obama +4 (last week Obama +3
Rasmussen: Obama +4 (last week Obama +7)
Hotline: Obama +7 (unchanged)
Gallup: Obama +9 (last week Obama +7)
TIPP: not published (last week Obama +3.5)
ABC: not published (last week Obama +9)

So in the last week, Obama has raised in 2, down in 3, unchanged in another. What's that mean? Well, let's look at the polls that went DOWN, of the 3 published so far, two of them has Obama in the rather bizarre double didgets, now a more sensible +7 and +6 (meaning even DROPPING they still show a lead greater than the lead Clinton had over Bush Sr.). Meanwhile, Tipp and Battlegrounds which HAD been showing pretty close numbers to justify the "OMG the polls are tightening, McCain SURGE!!!!!" of the right have backed away from their "too close to call" numbers and are posting a smaller than the rest, but still significant lead for Obama.

So in the last week, McCain has managed to do...not much really. The race has tightened according to some polls, but those polls were probably too high to be real anyway, and the polls that were OH SO CLOSE when Obama as posting consistant > +8 leads, instead of showing a tie or slight McCain lead we'd think if the polls really had closed in on an already short race they wer predicting...have widened.

So has the race tightened? Maybe, it seems maybe the race has gone from a high of maybe +7 to a low of maybe +5. Which means McCain, in a week, has managed to shave off 2 points from Obama's lead. A week to take off 2 points. He's got 3 more days to take off the last 5 points.

Also keep in mind that todays polls are the first polls to take in a full day of polling following his 30 minute TV special, later polls may reflect a widening level of support, or a stall in McCain's slow shaving off points.

In short, McCain has 3 days to do twice what he MIGHT have managed to do, in a week.

I have been letting this post slack a bit, in fact it's been exactly a week since I posted. Which makes for a perfect opportunity to compare.

About a week ago, people were saying McCain NEEDED a strong week, well, let's see if he got it. I'll post today's results, compared to one week ago:

Zogby: Obama +7 (last week Obama +10
Research 2000: Obama +6 (last week Obama +12)
Battlegrounds: Obama +4 (last week Obama +3
Rasmussen: Obama +4 (last week Obama +7)
Hotline: Obama +7 (unchanged)
Gallup: Obama +9 (last week Obama +7)
TIPP: not published (last week Obama +3.5)
ABC: not published (last week Obama +9)

So in the last week, Obama has raised in 2, down in 3, unchanged in another. What's that mean? Well, let's look at the polls that went DOWN, of the 3 published so far, two of them has Obama in the rather bizarre double didgets, now a more sensible +7 and +6 (meaning even DROPPING they still show a lead greater than the lead Clinton had over Bush Sr.). Meanwhile, Tipp and Battlegrounds which HAD been showing pretty close numbers to justify the "OMG the polls are tightening, McCain SURGE!!!!!" of the right have backed away from their "too close to call" numbers and are posting a smaller than the rest, but still significant lead for Obama.

So in the last week, McCain has managed to do...not much really. The race has tightened according to some polls, but those polls were probably too high to be real anyway, and the polls that were OH SO CLOSE when Obama as posting consistant > +8 leads, instead of showing a tie or slight McCain lead we'd think if the polls really had closed in on an already short race they wer predicting...have widened.

So has the race tightened? Maybe, it seems maybe the race has gone from a high of maybe +7 to a low of maybe +5. Which means McCain, in a week, has managed to shave off 2 points from Obama's lead. A week to take off 2 points. He's got 3 more days to take off the last 5 points.

Also keep in mind that todays polls are the first polls to take in a full day of polling following his 30 minute TV special, later polls may reflect a widening level of support, or a stall in McCain's slow shaving off points.

In short, McCain has 3 days to do twice what he MIGHT have managed to do, in a week.


McCains got us right where he wants us Neo. He loves to fool the pollsters.

Something I've wondered about:

The last election, all the polls had Kerry winning. And then the exit polls had him winning.

I don't have the figures for the last election on poll rejection rates (i.e., how many people tell the pollster to fuck off), but it's around 70 to 75 percent today, and far more Republicans than Democrats say they won't fill out an exit poll.

With skews in the sample that are so large, I am assuming that some people are lying when they answer polls as well.

I don't believe that the samples permit the polling to be accurate - at least not accurate enough to predict something like the current election.

The polls don't even agree with each other as tightly as I would believe - probably because the sampling is varying all over the place.

http://www.tipponline.com/

was the most accurate poll in the last election. I am tending to go with this one.

Something I've wondered about:

The last election, all the polls had Kerry winning.

At this point in the election? No, they did not.

then the exit polls had him winning.

The final results of which falling within the margin of error.

The final results of which falling within the margin of error.

Then why did so many Democrats say the last go-round that the polls said they would win, and they lost? Plenty of tinfoil fodder came out that time...

http://www.tipponline.com/

was the most accurate poll in the last election. I am tending to go with this one.

The one that shows Obama at +4.5 today, a point up from this time last week, and almost exactly as I said about the race being about a +5. It's amusing how conservatives "discovered" Tipp just over a week ago when their October 23rd poll showed the race at about Obama +1.1. Suddenly, when ONE poll showed a near dead heat, conservatives in the right wing were all over themselves screaming about how Tipp was "the most acurate poll" in 2004 and that the race was tightening and McCain was surging to the lead. McCain within a point in a half with two weeks to go!.

Unfortunatly for the right wing nuts, like Drudge, while they were so busy touting the credentials of "the most accurate poll" they couldn't explain why this oh so accurate pollster with their oh so accurate October 23rd, Obama +1.1 internals had the 18-24 year old demographic breaking 74% in favor of McCain. They were so busy calling Tipp so bloody accurate, they couldn't at all explain why this so very accurate poll had McCain capturing three fourths of the youth vote.

Of course, now that they hung their hats on their so very accurate poll, they're having to eat it as it starts to back off from that result, and is now showing a fairly significant Obama +4.5 and far more sensible youth numbers.

Then why did so many Democrats say the last go-round that the polls said they would win, and they lost? Plenty of tinfoil fodder came out that time...

some because they were overly hopeful about an outlier, the smarter ones because taking ohio looked like a legitimate chance.

http://www.gallup.com/Home.aspx

Obama flying ahead in all categories.

Something I've wondered about:

The last election, all the polls had Kerry winning. And then the exit polls had him winning.

False.

According to electoral-vote.com

Nov 4 - Kerry 252 Bush 279
Nov 3 - Kerry 252 Bush 259
Nov 2 - Kerry 262 Bush 261
Nov 1 - Kerry 298 Bush 231
Oct 31 - Kerry 283 Bush 246
Oct 30 - Kerry 243 Bush 280
Oct 29 - Kerry 236 Bush 281
Oct 28 - Kerry 260 Bush 254
Oct 27 - Kerry 257 Bush 274
Oct 26 - Kerry 247 Bush 285
Oct 25 - Kerry 247 Bush 285
Oct 24 - Kerry 253 Bush 254
Oct 23 - Kerry 257 Bush 254
Oct 22 - Kerry 257 Bush 271

Two weeks worth of polling leading up to the election. Kerry is up by a significant amount on two days out of those two weeks. Kerry is up overall 5 days out of the two weeks. So, out of 14 days, Kerry is up in 5 of them. ie. He was up 37% of the time, and never by very much.

So to say ALL the polls had Kerry winning is just completely wrong. And where he is winning, it is well within the margin of error, and never by a significant amount. The polls never had Kerry winning over 300 EV's for over a month. The polls always showed a close race between Kerry and Bush that would be too hard to really call.

Whereas McCain hasn't been ahead in any of the polls (in any significant way) since the GOP convention. I think he may have at one point been ahead by one or two points in one or two polls once or twice, but that's amounts to nothing more than statistical noise and/or bad polling. The average lead has continually been in Obama's favour since the GOP bounce faded, often by a significant amount.

Put simply - the polls said the election would be close in 2004, and it was. The polls said the election would be close in 2000, and it was. The polls say the election will be a landslide to Obama in 2008. Hmm...

Something I've wondered about:

The last election, all the polls had Kerry winning. And then the exit polls had him winning.

I don't have the figures for the last election on poll rejection rates (i.e., how many people tell the pollster to fuck off), but it's around 70 to 75 percent today, and far more Republicans than Democrats say they won't fill out an exit poll.

With skews in the sample that are so large, I am assuming that some people are lying when they answer polls as well.

I don't believe that the samples permit the polling to be accurate - at least not accurate enough to predict something like the current election.

The polls don't even agree with each other as tightly as I would believe - probably because the sampling is varying all over the place.

http://www.tipponline.com/

was the most accurate poll in the last election. I am tending to go with this one.

Straws. You're grasping at them.

Something I've wondered about:

The last election, all the polls had Kerry winning. And then the exit polls had him winning.

I don't have the figures for the last election on poll rejection rates (i.e., how many people tell the pollster to fuck off), but it's around 70 to 75 percent today, and far more Republicans than Democrats say they won't fill out an exit poll.

With skews in the sample that are so large, I am assuming that some people are lying when they answer polls as well.

I don't believe that the samples permit the polling to be accurate - at least not accurate enough to predict something like the current election.

The polls don't even agree with each other as tightly as I would believe - probably because the sampling is varying all over the place.

http://www.tipponline.com/

was the most accurate poll in the last election. I am tending to go with this one.

Tipp had Bush winning over Kerry for over a month. So not all polls had Kerry winning.
http://www.tipponline.com/samplereport.pdf
And yesterday TIPP is +4.4 Obama.

Time for a new round of polls!

Zogby: Obama +5 (yesterday Obama +7)
Research 2000: Obama +7 (yesterday Obama +6)
Battlegrounds: Does not publish on weekends (yesterday Obama +4)
Rasmussen: Obama +5 (yesterday Obama +4)
Hotline: Obama +7 (unchanged)
Gallup: Obama +10 (yesterday Obama +9)
TIPP: Obama + 4.5 (unchanged)
ABC: not published (last night Obama +9)

We have had 6 polls published today, 2 of them show no change at all, 1 shows a drop and 3 show a net gain for Obama. What's even more interesting, is that of the 6 published polls, 4 of them (all but Zogby and Tipp) shows Obama above the magic 50%. What that means is, if a politician is going to get > 50% of the vote, it doesn't matter how the independants or undecides break, once you have more than 50% of support, you win.

Basically, with 2 and a half days left until election day, we're just not seeing any real momentum for McCain.

added Tipp, Obama +4.5. No change from yesterda

OK folks, this is it, the final poll results before the election! (there may be results out tomorrow which include polling in the field done today)

So without further adeu, there are NO MORE DAYS of campaigning until election day, as the candidates enter the home stretch, so I will be adding a bit of analysis to each pollster.

And we're off:

NATIONAL TRACKERS:

Zogby: Obama +7. Showing McCain at 44 and Obama at the magical 51%. Obama gained a point and a half since yesterday as McCain held steady. Zogby has had some issues, including one night of Republican based celebration when his friday sample showed McCain +1. Of course this didn't really help McCain much, as the aggregate averaged out, and Zogby has been fairly stable, ranging from Obama just under 5, to today, his highest range in a week at +7.1

Research 2000: Obama +6. 51 to 45. Up 1 since yesterday. A rather volatile poll that has shown leads for Obama as much as +12. In the last week it's evened out somewhat, although not without it's peaks and valleys, but hasn't fallen below a +5 for Obama. It did show a rather large gain for McCain towards the end of the month, but that leveled off the last few days.

Battlegrounds: Obama +6. A fairly volatile poll, on the 21th it showed a nail biting 48/47 +1 for Obama, it's backed off that. While Obama has remained fairly steady, not stepping out of the 48-50 range in two weeks, McCain has fallen from his high of 47 down to a low of 44, showing simultaneously Obama's highest numbers in 2 weeks at 50%, and McCain's lowest at 44%, opening up a 6 point lead in what was less than two weeks ago, a 1 point lead.

Rasmussen: Obama +6. Up one from yesterday. Rasmussen has been fairly stable, and with the exception of one single day from the last 40 days, has shown Obama to be between +4 and +8, never stepping out of the 50-52 range, and McCain never, except for one day, breaking out of 44-46. Of course, that range includes everything from the soaring +8 to the close +4 (and one day of +3 a week ago when McCain broke his range and hit 47%). Most interesting is that it has never, not even once, in the last 40 days, shown Obama to slip below 50%

Gallup: Obama +11. 53/42, an overal shift of +1 Obama and -1 McCain since yesterday. Another fairly pro Obama poll, which, with the exception of an Obama +4 day on the 18th, has not shown McCain within 6 points of Obama in the last month, and, in that entire time, never, once, showed Obama below 50%

Hotline: Obama +5 (50/45). Unchanged from yesterday, and a +1/-1 shift in favor of McCain from the 1st. Numbers have been fairly stable the last week, ranging from Obama +5 to Obama +7.

TIPP: Obama +4.5. Probably the most volatile and hard to read poll in the bunch. TIPP has had a lot of problems in recent days. For example, since the first, TIPP has gone from Obama +4.5 to +2.1 back to +4.5. Another example. On October 21st, TIPP had its biggest lead for Obama at +6. Not two days later, it showed an almost dead tie, with Obama +1. It then gradually raised it back to a 4.5, then saw a 2.5 point plunge the next day, which immediately and fully recovered the next. TIPPs polls are notoriously screwy this election cycle.

And there we have it so far today. With the exception of a rather high Gallup poll, everyone suggests that the race is between 4.5 and 7 points, going into election day tomorrow. Additionally, of the 7 polls so far today, 6 (all but the screwy TIPP) show Obama at > 50%.

Beaten by Neo Art dammit. However the BBC are saying that according to their poll it'll be 54 - 44 to Obama.