Pollster.com

Articles and Analysis

 

Morning Status Update for Saturday 11/1

Topics: Status Update

Not surprisingly, yesterday was another heavy day of new poll releases: 37 new statewide surveys and 10 national releases, yet these surveys indicate no clear trends and leave our bottom line electoral vote count unchanged. McCain does a little better on some polls and on some of our statewide trend estimates, Obama does a little better on others. The net result -- for today at least -- looks mostly like random noise.

If we do "apples-to-apples" comparisons, looking only at how the most recent surveys compare to previous tracks by the same pollster, the pattern is mostly random. Of the 37 new statewide surveys, 16 represent track updates from previous polls conducted released since October 15; McCain gains at least nominally on 7, Obama on 4 with unchanged margins on 5. If we look back a bit farther, and examine the 25 new polls that track previous polls released since October 1, the pattern is almost perfectly random: 10 show nominal gains for McCain, 9 show nominal gains for Obama and 6 show zero change in the margin.

081101 new polls

The pattern is similar in terms of how these new surveys affect our trend estimates in battleground states. Since yesterday, 9 states show slight increases for Obama, 6 for McCain. Compared to last Saturday, 14 states show slight gains for Obama, 9 for McCain.

081101 trends

Yesterday's polls did help shift two states from the lean to strong categories. In New Mexico, a new PPP survey shows Obama leading by 17 percentage points (58% to 41%). That margin is significantly better than any other New Mexico poll to date, although the Obama margin on our trend estimate had been increasing as a result of two recent Rasmussen surveys that put Obama ahead by smaller double-digit margins. Obama now leads by nine points (52.6% to 43.6%) on our trend estimate, enough to qualify for dark blue "strong" Obama status.

In Mississippi, polls have been relatively rare (just four in October), a condition that makes our trend line more sensitive to recent results. The new Daily Kos/Research 2000 survey, showing McCain leading by 13 points (53% to 40%), helps boost McCain's lead on our trend estimate back to nearly ten (50.9% to 41.3%) enough to shift back to "strong" McCain. It had been "lean" McCain for a single day, due to a closer Rasmussen result released on Thursday.

For some reason, yesterday probably marked a new general election record for new releases in a single day in New Hampshire, where six new surveys show Barack Obama leading comfortably, by margins varying between 7 and 15 percentage points. The new results increased Obama's margin on our trend estimate only slightly. He now leads by 11.4 points (52.5% to 41.1%) in a state that John Kerry carried by a single percentage point in 2004 (of course, those with a healthy sense of irony and caution will recall another big day for polls in New Hampshire earlier this year).

The trend estimates indicate continuing progress for McCain in Pennsylvania, a state that has been the focus of increased Republican candidate time and television advertising. A new poll from Republican affiliated Strategic Vision and a new tracking release by Muhlenberg College both show Obama leading by single digit margins with McCain at 44%. Our trend estimate shows Obama leading by 9.6 points (51.8% to 42.5%) enough to remain in the "strong" Obama column.

Yesterday's new national poll releases, however, yesterday provided no good news to those looking for a late shift to McCain. Obama's margin on our national trend estimate clicked up slightly yesterday (to +6.0) for the first time in a week. The table below shows how yesterday's new polls compare to the previous non-overlapping samples from the same pollsters. The pattern of change in this table looks mostly random -- 5 nominal shifts to Obama, 3 to McCain and one unchanged.

081101 national

I did a separate update last night comparing the Friday daily tracking releases to those from Thursday. Again, if anything, the pattern suggested a slight improvement for Obama.

 

Comments
jaxspellinar:

First...? :)

I was just listening to Chuck Todd on Today. He seemed strained to say that there'd be a lot for the news to talk about Tuesday after about 9pm except for the Kentucky Senate race, unless they can't call Virginia for Obama by around then.

In Chuck Todd we trust, everyone else must bring data.

____________________

Chicago:

I'll believe it when I see it. Polls are not votes but young voters think they are the same and will stay home Tuesday.

____________________

merveilleux:

Can you please fix the poll number typos?

____________________

BarackO'Clinton:

LOL young voters staying home. We just enough to turn out, not an overwhelming margin.

If they can stand outside for 4 days for an iPhone, or camp out for 6 weeks to see Star Wars, I can imagine them waiting in line for a couple hours to vote for history.

____________________

bythesea:

Has anyone been watching the changes in the state polls and correlating them to campaign activities to determine what campaign activities actually have an impact? Its clear that candidates run around frantically trying to have an impact and campaigns move their ad dollars and tactics trying to shift some voters, but what actually works? My casual observations suggests that visits by candidates move the state numbers a couple of points but I haven't monitored it closely enough to say for sure. Can anyone offer anything definitive on this question?

____________________

Fioreb:

I am happy to see polls continue to be rather stable when it comes to trends, at the same time I remain a bit apprehensive because so many things could still go wrong, including some issues like the

@bythesea

I would be surprised if any major impact was to be felt, at this point only less than 6% seem to still be uncertain and chances are that if a differential exists is likely to break at no more than 2:1. That said polls may continue to show some marginal gain by Sen McCain (historically polls margin narrow as election day approaches) but should the last undecided voters show up at voting booths 'en masse' and go to McCain with the above quoted 2:1 ratio that would still mean that Sen Obama would hold ~ 1-2% vote advantage nationally.
That should award him the presidency at last.

____________________

thoughtful:

Remarkably stable.

Is there going to be a last day surge for either ticket?

____________________

One more good aspect of this election. Kay Hagan leads in 5 polls over that White trash southern conservative Lizardbeth Dole. It looks like she will have to go back to the white cross, err red cross. One less Klan member in the Senate, Yay.

____________________

RossPhx:

Obama is strong in states with 264 electoral votes, according to this trend estimates chart. What if he carries only those, and Nevada with 5? No, let's not even think about that.

____________________

Whoever wins the election, neither Obama nor McCAIN support single payer universal health care. I would like people to click on my username and answer a single poll question on Single Payer Universal Health Care.

Thank you.

____________________

kth:

Ross, it would be somewhat unlikely for Obama to take Nevada (and NM) but not Colorado. And even in the highly unlikely event of a 269-269 tie, it goes to the incoming House of Representatives, which will be even more tilted to the Dems come January than it is now. 270 really is do-or-die for McCain.

____________________

blakec:

One more good AZ poll and it should flip to toss up.

____________________

spork:

During the primary, I seem to remember Obama always polling a few points higher than he ended up with. Is there a reason that this should not be the case now? Anyone seen an analysis of this data or see it applied to the current polls?

____________________

balancebudget:

Anecdotally, the main impact of campaigning in local markets is the free coverage from the local TV/news. Obama was doing much better with this because he had distributed his resources, while McCain was holding Palin close (figuratively? literally?). He lost valuable coverage by trying to control her and improve his "macho" credentials.

____________________

Breaking News !!

Obama has released the SECRET WEAPON against McCain.

This is the first time that a U. S. President speaks in Spanish to Latinos. WOW !!

Obama is speaking in Spanish to Latinos. He is in front of the Camara talking in perfect Spanish. He kisses latino children and talks of education and health care for children.

There are mariachis ( mexican band with big hats and so on )

Washington post says this is an "unauthorized leak" ... Republican Nel Martinez of Florida is furious.

I try to make a fuzz of this and follow this attempt to conquer the Southwest and Florida, where the ad is aired.

This is the first time that a U. S. President speaks in spanish to Latinos. WOW !!

See the YouTube video and comments on the impact of Obama in Spanish here

http://raciality.blogspot.com/

Vicente Duque

____________________

phillybooster:

If we look back a bit farther, and examine the 25 new polls that track previous polls released since October 1, the pattern is almost perfectly random:

5 heads and 5 tails is no more "perfectly random" than '6 and 4' or '3 and 7'.

I wish I had a better sense of how likely a change it to be "random" and how likely it is to be "meaningful."

____________________

Mavericky:

Hi first time commenting,
ZOGBY: MCCAIN MOVES INTO LEAD 48-47 IN ONE DAY POLLING claims our friends at Drudge.

I'm reminded of a very old saying "Figures don't lie but Liars figure".

How appropriate!

____________________

Trosen:

RossPhx:
"Obama is strong in states with 264 electoral votes, according to this trend estimates chart. What if he carries only those, and Nevada with 5? No, let's not even think about that."

Then he's at 269, as is McCain. Tiebreaking vote goes to the Democratic senate, Obama is president, even under the most unbelievable nightmare scenario.

That better?

____________________

duneboi29:

I attended an Obama rally last night in Highland, (Northwest) Indiana with an estimated crowd of 40,000 people. Indiana truly has become a swing state in this election. I waited in line 3 1/2 hours to early vote -- in fact people in line were being bused to other polling places to alleviate the long lines. According to a recent polling, Obama is leading McCain 50 to 46 percent amongst those who have early voted in Indiana. I know of a fair number of Republicans who are splitting their ticket and voting for Obama. For Obama to carry Indiana, he would have to garner large pluralities in Northwest Indiana and greater Indianapolis. I know the Obama campaign is well organized in this part of the state, which borders Chicagoland. Do not be surprised, come Election Day, that Indiana is one of the last states to be called, and it will migrate to the blue column. And check out the local coverage of Obama's Highland rally at

www.post-trib.com.

____________________

tar_heel:

trosen: nope, the Senate would pick the VP (Biden), and the House would elect the President, but under the unit rule in which each state has only one vote. The Democrats, though, are certain to have the majority in 27-29 state delegation plus D.C. (not clear if their votes count).

____________________

benjam421:

Do we have any clear idea what happened in the NH primary to Obama's numbers?

____________________

Chicago:

Polls are not votes people, wake up. Exit polls on early voting are just polls, could be lies or bad polling. Remember recent history? Enthusiastic crowds are not votes. Registering is not a vote. "Supporting" a candidate on MY SPACE and Facebook is not a vote. the youth vote will disappoint again and we get 4 more years of McBush. Obama supporters will stay home thinking this is a landslide. You are in for a big sad surprise. I hope I am wrong, but add Rep dirty tricks and I doubt it.

____________________

Kenny:

Another newbie question:
Do you do strong/lean/tossup on the basis of some specific percentage? Some percentage that varies with the size of the state? Some percentage that varies with the size of the poll samples (say, one standard deviation for lean, two for strong)?
Is there a common practice in the polling business for the use of these terms, or do pollsters vary in their usage?

____________________

edo4cubs:

I am in Fort LAuderdale and this antedotal..but here goes... my brother in law and 2 cousins who have never voted anything but republican have voted 2 for obama 1 for nader... now these people are red as red can be and I know none of my friends who have switched from dem to mccain....so this is small but if you have say a 10 % defection to dems ... OBMAMA should win... so my call is FLorida for Obama by 5%.....

____________________

bassoonista:

@Chicago

"Polls are not votes people, wake up. Exit polls on early voting are just polls, could be lies or bad polling. Remember recent history? Enthusiastic crowds are not votes. Registering is not a vote. "Supporting" a candidate on MY SPACE and Facebook is not a vote. the youth vote will disappoint again and we get 4 more years of McBush. Obama supporters will stay home thinking this is a landslide. You are in for a big sad surprise. I hope I am wrong, but add Rep dirty tricks and I doubt it."

I don't know ANYONE that is staying home because they think it's in the bag. The Democrats I know (including myself) are very nervous for the outcome and are taking nothing for granted. I think turnout will be strong for Obama.

Many McCain voters think he has already lost, so there may be some apathy that direction also.

____________________

BACK22000:

It looks increasingly to me that if Obama wins either Ohio or Pennsylvania he is President. If McCain wins both then he will President.

back22000

____________________

carysr:

Back22000:
Check out the New York Times map at http://elections.nytimes.com/2008/president/whos-ahead/key-states/map.html It shows Ohio and Florida as toss-ups, but gives PA to Obama. Obama wins with 291. Take away PA and he wins with 270. (You can do the same with Pollster's map. Currently 311 for Obama. take away Ohio and PA and it's also 270). McCain would still need to take away another state where Obama is leading in order to win. Not going to happen.

____________________

SlimTim:

The smears will only last so long and the Obama infomercial drove the nail through the coffin. Even with the kindest of slants towards McCain you still have Obama at +4% in the popular vote and McCain having to win ALL the toss-ups (FL, GA, NC, IN, MO, AZ, MT, ND) AND Ohio AND Pennsylvania AND either of the Obama leaning Colorado/Virginia/New Mexico/Arizona...

And this inspite of the democratic turn-out having been much higher than the polls suggested in 2006 senate race...


Nothing short of a widespread voting fraud will turn this election as it stands.

____________________

Vicente Duque:


What you have to pay to win $100 at INTRADE.
Prices for bets on States, Sunday, November 2, Morning at 7 AM.


Toss Up States for Obama with Last Prices :

NewMexico (5EVs) ...Obama 92.5, McCain 15.5
Colorado (9EVs) ..... Obama 85.4, McCain 16.0
Nevada (5EVs) ....... Obama 80.0, McCain 24.8
Florida (27EVs) ...... Obama 70.0 McCain 31.7
Ohio (20EVs) ........ Obama 75.0, McCain 24.5

Iowa (7EVs) .............. Obama 92.0, McCain 8.0
New Hampshire (4EVs) . Obama 89.8, McCain 12.5
Pennsylvania (21EVs) ... Obama 85.0, McCain 18.0
Virginia (13EVs) ......... Obama 83.0, McCain 20.0

North Carolina (15EVs) .. Obama 63.9 McCain 40.5
Missouri (11EVs) ......... Obama 52.9, McCain 51.9


So Florida and Ohio look more secure than North Carolina and Missouri.

http://tossUpStates.blogspot.com/

http://milenials.blogspot.com/


Vicente Duque

____________________

cube:

Phillybooster,

In a random event generator that, on each coin toss is 50:50, out of 10 tosses, 5 heads and 5 tails is more likely than 6 and 4 or 7 and 3. In fact, 5 and 5 is the outcome that is most consistent with a random distribution. Trouble is, there are no simple words I can think of to say this.

Another way of saying the same thing is that, if the underlying distribution is NOT random and prefers heads by, say, 70-30, you are more likely to get 6:4 than 5:5.

"perfectly random" is a poor choice of words. One weird thing about statistics is that there is no way (I know of) to give the odds that a distribution is random. You can, however, reject with a certain degree of confidence the idea that the distribution is random.

____________________

RWCOLE:

ties are decided by the house- not the senate- and every state gets one vote...

____________________

RWCOLE:

On election night, I'll be watching Ohio, Pennsylvania, and Virginia...if McCain wins all three- we have a contest- otherwise the night should go to Obama

____________________

BACK22000:

CARYSR

Thanks for the link. I should have said that I am assuming a McCain victory in Virginia. So to clarify, McCain needs to win Virginia, Ohio and Pennsylvania to win. Looks like a longshot but still possible.

____________________



Post a comment




Please be patient while your comment posts - sometimes it takes a minute or two. To check your comment, please wait 60 seconds and click your browser's refresh button. Note that comments with three or more hyperlinks will be held for approval.

MAP - US, AL, AK, AZ, AR, CA, CO, CT, DE, FL, GA, HI, ID, IL, IN, IA, KS, KY, LA, ME, MD, MA, MI, MN, MS, MO, MT, NE, NV, NH, NJ, NM, NY, NC, ND, OH, OK, OR, PA, RI, SC, SD, TN, TX, UT, VT, VA, WA, WV, WI, WY, PR