Pollster.com

Blue states holding firm, a critical red state trending blue. The Obama electoral strategy is working.

Come on McCain, keep wasting time & money in these states!

McCain has no choice but to spend money in Colorado, it's practically a must win state with Iowa & New Mexico looking very good for Obama.

You know, without knowing the party affiliation weightings it really is impossible to vouch for the validity of these polls.

I always have to wonder why anyone would do a poll and not simply release the party identification of the respondents since it is probably the #1 most important thing affecting the outcome (and the information is easily accessible).

I mean, what use is a poll unless I know the party distribution was fair and accurate? We have seen with the national polls what a HUGE difference simple party identification discrepencies can play.

mccain has one chance in PA

Change from previous Q polls

CO- Obama +3, McCain -1
MI- Obama +2, McCain +2
MN- Obama +1, McCain +1
WI- Obama -1, McCain +3

I can't see this as anything but good news for Barack.
1) Great shift in the mood in Colorado.
2) Fewer undecideds in MI and MN basically keeping the status quo
3) He was/is so far up in WI that it isn't a concern.

Hopefully, we'll start seeing some light blues up there.

@Change

PA may look appealing but it is a death hole for McCain. He has never had a lead in that state. Just saw an NBC Mason/Dixon poll with Obama up 2 in FL.

Quinnipiac has MN close; however, it is the same margin as it was back in July, Obama 46% vs. McCain 44%. We all saw Rasmussen numbers yesterday. I'm waiting to see how the campaigns take this numbers. If McCain and Obama start visiting the state, that could be an indication that in fact something is going on there. Otherwise, I will suspect that it is not this close.

Quinnipiac has Obama leading WI; however, it is tighter than it was back in July, Obama 50% vs. McCain 39%. Other polls have shown the race tighting, but the thing that makes me more aware of McCain's chances on the ground is McCain/Palin's presence in the state. They have been campaigning in the state lately. Likewise Michelle Obama and most recently Barack. I would dare to say that WI is indeed becoming a little more competitive for McCain.

Unless a poll shows party idenitifcation to provide a frame of reference it is USELESS.

When they don't show party identification although that information is surely available to them, we must ask ourselves WHY this key data is beig intentionally withheld from the analysis.

Why not just show it?

no way not in FL, careful with PA alot of racist folks in the rustbelt

It's funny how we are making all of these prognostications and we haven't even had the first debate. Historically the debates have resulted in swings of 5 to 9 points.

@change:

Obama gets many more votes because is IS black than he loses because he IS black.

And DailyKos continues to believe only 26% of Americans are Republicans. Why do they even waste our time?

riverrun:

Research2000 daily tracking once again shows Obama/Biden at +6%. Internals are fascinating once again: Obama and Biden net favourables are high (21%, 20% respectively), McCain and Palin putrid: McCain just 1% net favourable, Palin negative (-4%).

So what if a NH poll shows a small lead for McCain? It's unlikely to hold up, but let's be generous and give it to him: Obama doesn't need it.

Pollster, 538 and RCP averages all show Obama comfortably ahead in all the remaining Kerry states as well as Iowa and New Mexico, and also (less comfortably) in Colorado - but strengthening there. Nate Silver's sophisticated projections on 538 also show him ahead in Virginia, Nevada and Ohio.

McCain is making an arse of himself every time he opens his mouth on the economy, while Obama looks increasingly presidential. Team Obama will hold all the states where they are currently leading, and move ahead in Indiana, Florida, and even North Carolina and Missouri.

Does he need such a big margin? You bet! Bush and his allies have done so much damage, the country needs not just a new leader, but a complete change of mindset. A sweeping victory across many previously red states, supported by gains down the ballot, will help to bring about this sea change.

There you go again guys!!!!

@billy change

send me the link for Mason-dixon poll

Some people have been complaining about how the map in pollster.com has been slow to change in their favor. I for one prefer that they stay conservative in their estimation and keep things neutral - unlike say FiveThirtyEight, which I frankly find laughable. Check out their "polling average" (neutral presumably) and their cooked up "projection" for the states that are not solidly Republican or Democrat. They might as well simplify and give Obama a fixed bonus points.

Following was as of this morning:

Arkansas
Polling average: McCain +11.9
Projection: McCain +11.1

Colorado
Polling average: Obama +1.3
Projection: Obama +4.3

Florida
Polling average: McCain +3.2
Projection: McCain +1.2

Indiana
Polling average: McCain +2.5
Projection: McCain +1.3

Iowa
Polling average: Obama +9
Projection: Obama +10.5

Louisiana
Polling average: McCain +12.4
Projection: McCain +7.4

Michigan
Polling average: Obama +3.5
Projection: Obama +5.5

Minnesota
Polling average: Obama +4.7
Projection: Obama +7.3

Missouri
Polling average: McCain +4.9
Projection: McCain +2.2

Montana
Polling average: McCain +6.1
Projection: McCain +2.5

North Carolina
Polling average: McCain +5.5
Projection: McCain +2

North Dakota
Polling average: McCain +10.9
Projection: McCain +8.3

New Hampshire
Polling average: Obama +0.4
Projection: Obama +2.1

New Jersey
Polling average: Obama +7.3
Projection: Obama +8.8

New Mexico
Polling average: Obama +6.2
Projection: Obama +7.7

Nevada
Polling average: McCain +2.3
Projection: Obama: +0

Ohio
Polling average: McCain +2.1
Projection: Obama +0.4

Oregon
Polling average: Obama +6.2
Projection: Obama +10.6

Pennsylvania
Polling average: Obama +3.1
Projection: Obama +4.8

Virginia
Polling average: McCain +0
Projection: Obama +3.4

Washington
Polling average: Obama +6.5
Projection: Obama +10.3

Wisconsin
Polling average: Obama +4.3
Projection: Obama +7.4

Their recent change (see below) doesn't fully explain the inexplicable bias in swing states.

http://www.fivethirtyeight.com/2008/09/allocating-undecideds.html

Tuesday, September 23, 2008

The Rasmussen Reports daily Presidential Tracking Poll for Tuesday shows the race for the White House is tied once again—Barack Obama attracts 48% of the vote and so does John McCain (see trends). Both men are now viewed favorably by 54% and Rasmussen Markets data currently gives Obama a 51.5% chance of victory.

Republicans typically win a majority of the White vote and this year is no exception. Among White Democrats, 78% say they’ll vote for Obama, 17% for McCain. Obama leads among non-White voters by a 71% to 24% margin.

@riverrun:

Do you actually believe the Research2000 assertion that only 26% of the electorate is republican?

You can't be serious?

@ Change,

Yeah, all those racists in Philly and Pitt. They're pretty extreme. They even go to the extreme of permanently darkening their skin! I've never seen such dedication to a political candidate.

McCain/Palin '08/'12!!

boomshak,

1. Rasmussen does not release the internals of their state levels polls. By your logic, they too should be useless.

2. Debates do NOT historically result in 5-9 point swings. See Nate's writeup of a study by Tom Holbrook.

See, the thing is, Mr. Holbrook did an actual STUDY, you just pulled **** out of you a$$.

http://www.fivethirtyeight.com/2008/09/debates-may-not-be-decisive-after-all.html

@ Boom

"boomshak:
It's funny how we are making all of these prognostications and we haven't even had the first debate. Historically the debates have resulted in swings of 5 to 9 points."

FAIL.

http://election08data.blogspot.com/2008/09/debate-effects.html

Obama only getting 58% of Clinton supporters.

http://apnews.myway.com/article/20080923/D93CC8A80.html

haha, PA-John, psych!

@PA-John:
@ Boom

"boomshak:
It's funny how we are making all of these prognostications and we haven't even had the first debate. Historically the debates have resulted in swings of 5 to 9 points."

FAIL.

http://election08data.blogspot.com/2008/09/debate-effects.html

I disagree with that analysis.

Rasmussen is buddies with Hannity and has a love fest on his radio and / or TV show twice a week. I consider anything he reports to be Obama's low end in the polls and manipulated from time to time to serve a purpose.

Here's an interesting sccenario.

Obama wins the Kerry states, plus Colorado and Iowa minus New Hampshire and one electoral vote from Maine.

You know what that gets us 268 Obama 270 McCain. That's actually what the electoral map looks like right now with uncertainty about the Maine 1 EV, not enough data, but its close enough there for it to be possible

NH also hasn't had many polls the only LV recent poll has MCain ahead, but its distinctly possible it goes McCain.

An interesting map don't you think?

@Mike_in_CA:
boomshak,

1. Rasmussen does not release the internals of their state levels polls. By your logic, they too should be useless.

Mike, if you actually paid attention occasionally, perhaps you wouldn't appear so foolish.

I have said on numerous occasions that state polling (including Rasmussen) should be taken with a grain of salt due to the high MOE.

I'm still a bit nervous about the Minnesota numbers. True, McCain hasn't been ahead in any poll here- but it seems like we have been flooded with Republican time and none from the Democrats.

boomshak,

I'm foolish? You just proclaimed your disagreement with an academic study about debate bounces based on ACTUAL polling data, just "because." Have you conducted your own study? Please elaborate on your disagreements and why your opinion is gold. Just saying you don't agree with something means nothing in the world of actual rational human discourse.

@change:

The Mason-Dixon florida poll is listed on the RCP state polls table - interviewing 09/16 - 18. The full link (headline: "Sunshine for Obama")is

http://firstread.msnbc.msn.com/archive/2008/09/23/1435316.aspx

I would not be surprised to see Gallup fall to +2 Obama today (who the hell knows what party affiliations he has).

We have seen both Battleground and Rasmussen move to McCain today with BG +1 and Ras +1 from yesterday.

@boomshak

"I disagree with that analysis."

Analysis? It's statistics based on history. Do you have different stats? Otherwise it's pretty straight forward.

Boomshak:

I am not sure if quinnipaic do weigh for party-id but my back of the envelope guess for their four samples based on their cross-tabs is CO +1R, MI +3D, MN +4D, WI +9D. In the 06 elections the exit polls had the party id's at MI +7D, MN +4D, and WI +5D. There was no 06 exit poll for CO but most polls have the party-id within about one.

@boomshak

Rasmussen is completely in the tank for mccain. i mean their virginia poll flies in the face of survey usa, their florida poll flies in the face of time/cnn and now mason dixon .. nationally they show no bounce for obama while other polls do..then they decide to team up with fox .. something is up

@Bonzi77:
@boomshak

"I disagree with that analysis."

"Analysis? It's statistics based on history. Do you have different stats? Otherwise it's pretty straight forward."

He uses some "poll of polls" in his analysis. As we have seen this year, many polls use biased sampling and are unreliable. I have no idea which polls he included in his analysis and how reliable they are.

For instance, we have this from 2004 on CNN:

"WASHINGTON (CNN) -- "Kerry had been lagging behind Bush in Gallup polls until he met the president September 30 at the University of Miami to debate foreign affairs.

A poll taken September 24-26 had the Massachusetts senator trailing Bush by a 44-to-52-percent margin. After the first debate, those numbers moved to a 49 percent tie."

As you can see, Kerry got a 5 point bounce from this poll.

@John:

I guess my point is they KNOW the party affiliation, why do they withhold it since it is key data to understanding the poll? I ask the same question of Gallup.

Since Rasmusson is a three day tracking poll it means that McCain must have had a very good day yesterday in the polling to go from -1 to even. Does anyone believe that McCain had a good day? I say this down tick is an aberration. Let's see what the Gallup survey shows later in the day.

@ change:
Rasmussen is not in the tank for McCain. Everybody seems to have a short memory in regards to this. The week after the Repub convention when ALL (supposedly democratic leaning pollsters) had McCain leading by a margin of 3 - 5 points, Rasmussen only had McCain up by 1-2 points. Dems weren't complaining then of any biases that he may harbour...

Whilst I may not like his affiliations, I think he is probably the most transparent, and probably accurate pollster out there.

Looking at MN, maybe that Rasmussen Poll is a little inaccurate. The last 3 polls, (Star Tribune, Battleground, and Quinn.) have just shown a 2 point lead for Obama....Good numbers for WI, more on WI later. I don't buy the FL numbers though....

I still say the numbers are too close considering what happen last week. If there is an international incident before the election, then it's going to be very hard for Obama.

Rasmussen has the race tied again, Battleground +2 for McCain. I

@ vmval1:
I have noticed the same. Ras had the highest poll for Obama here in Minnesota.

But Ras state polling samples are small and his MOE of +-5 is high. Take it for what it's worth.

At this point and with the debates coming, it's anybody's game.

And this doesn't even include any potential October Surprises.

I think you're right Boomshak. It's WAY too early for either side to being feeling good about things yet.

Even if there are no October surprises, I'm sure the respective camps will come up with a fantastic highlights package - i.e. god damn america, called a woman a pig, Ayers, Rezko vs Fundamentally Sound Economy, 7 houses, 13 cars etc etc

It'll make for a very intelligent and measured discussion...

Actually... change... Rasmussen did show a little itty bitty bounce for Obama. You just were not paying attention. Now it is gone. McCain was up, then they were even, Obama was up, and now they are even again. Take a look at their trend numbers--

http://www.rasmussenreports.com/public_content/politics/election_20082/2008_presidential_election/general_election_match_up_history

Again I remind you that Gallup is "registered" voters which favors Democrats more because of the greater number of registered Democrats...
AND Rasmussen is "likely" voters which tends to even out but Democrats are still ahead of Republicans. The two relevant factors issue are--
1. Number of Democrats voting GOP and
2. Who fares better with Unaffiliated vote.

So quit complaining about the pollsters. Query why Obama cannot get those Democrats and unaffiliated voters who now favor McCain.

vmval1

Lets not forget the Keating 5

Rasmussen carries a 3 point "house effect" republican, CBS the same democrat.

The least impacted by house effect are..

Diaego, Cook & CNN, Quinn, ABC, Dem Corps, NBC/WSJ (strangely Dem Corps has a Republican bias)

The problem with states like MI and MN (and also with IN and NC for Obama) is that they're teases. They show varying degrees of defeat for the candidate - sometimes decisive, sometimes close, thus encouraging spending lots of resources chasing a win that's not going to happen absent a victory that would have been decisive anyway.

we should seriously have an election system like the UK or Australia. Where each party is allowed an equal amount of money to run their campaign for a period of 4 weeks only - i.e. mandatory public financing.

It is absolutely ridiculous that in our elections, the guy with the most money is going to be the one most likely to win. And then when he does win, he really only has about 6 -8 months to govern before he has to think about the next set of elections.

Well, as the Obamaniacs would note, "these polls are old." (Note they begin on September 14.)

That being said, assuming they are accurate, it would make sense in that McCain didn't have a good week last week -- this, of course, not not indicate a "trend."

It is interesting to note that even considering McCain's bad week that MN remains a (statistical) tie.

Colorado is trending darker blue every day as is Virginia... Obama is playing close to the edge iN PA, WI, MN and MI, but then again, McCain is playing russian roulette in Indiana and North Carolina.

In the end, Florida is surging for Obama and McCain is reeling all over the place.

Since some people keeps arguing about if the debates did X or Y for Bush. Here I am posting all the polls since September, at the end of the Republican convention, the last convention that year. Draw your own conclusion:

ARG (1258 LV) 10/28- 10/30 Bush 48% Kerry 49% Kerry +1
FOX News(1200LV)10/28- 10/29 Bush 47% Kerry45% Bush +2
Newsweek(882 LV) 10/27-10/29 Bush 51% Kerry45% Bush +6
Rasmussen (3,000 LV w/leaners) 10/27 - 10/29 Bush 48.8% Kerry 48.3% Bush +0.5
TIPP (792 LV) 10/26 - 10/29 Bush 45% Kerry 44% Bush +1
Battleground (1000 LV w/leaners) 10/25 - 10/28 Bush 51% Kerry 46%, Bush +5
CNN/USAT/Gallup (1195 LV) 10/22 - 10/24
Bush 52% Kery 46%, Bush +6
Los Angeles Times (881 LV) 10/21 - 10/24
Bush 49% Kerry 48%, Bush +1
Newsweek (880 LV) 10/21 - 10/22
Bush 48% Kerry 47%, Bush +1
Time (803 LV) 10/19 - 10/21 Bush 52% Kerry 47% Bush +5
GW/Battleground (1000 LV) 10/18 - 10/21
Bush 49% Kerry 45%, Bush +4
Rasmussen (3,000 LV) 10/17 - 10/19
Bush 48% Kerry 47%, Bush +1
FOX News (1000 LV) 10/17 - 10/18
Bush 48% Kery 43%, Bush +5
CBS News (678 LV) 10/14 - 10/17
Bush 47% Kerry 46%, Bush +1
CNN/USAT/Gallup (788 LV) 10/14 - 10/16
Bush 52% Kerry 44%, Bush +8
Time (865 LV w/leaners) 10/14 - 10/15
Bush 48% kerry 48% TIE
Newsweek (LV) 10/14 - 10/15 Bush 50% Kery 45% Bush +5
GW/Battleground (1000 LV) 10/11 - 10/14
Bush 49% Kerry 46%, Bush +3
CBS News (760 LV) 10/9 - 10/11
Bush 47% Kerry 46%, Bush +1
ICR (763 LV) 10/9 - 10/11 Bush 49% Kerry 46% Bush +3
CNN/USAT/Gallup (793 LV) 10/9 - 10/10
Bush 48% Kerry 50%, Kerry +2
Rasmussen (3,000 LV) 10/7 - 10/9
Bush 50% Kerry 46%, Bush +4
Time (886 LV w/leaners) 10/6 - 10/7
Bush 47% Kerry 46% Bush +1
GW/Battleground (1250 LV) 10/3 - 10/7
Bush 49% Kerry 46%, Bush +3
Fox News (1000 LV) 10/3 - 10/4
Bush 48% Kerry 45%, Bush +3
ICR (762 LV)** 10/1 - 10/5 Bush 51% Kerry 46% Bush +5
ARG (800 LV) 10/2 - 10/4 Bush 46% Kerry 47% Kerry +1
CBS/NYT (561 LV) 10/1 - 10/3
Bush 48% Kerry 47%, Bush +1
Zogby (1036 LV) 10/1 - 10/3 Bush 46% Kerry 45% Bush +1
CNN/USAT/Gallup (772 LV) 10/1 - 10/3
Bush 49% Kerry 49% TIE
Newsweek (1013 RV) 9/30 - 10/2
Bush 46% Kerry 49% Kerry +3
Battleground (1000 LV) 9/27 - 9/30
Bush 51% Kerry 44%, Bush +7
LA Times (1100 LV) 9/25 - 9/28
Bush 51% Kerry 46%, Bush +5
CNN/USAT/Gallup (758 LV) 9/24 - 9/26
Bush 52% Kerry 44% Bush +8
IBD/TIPP (649 LV) 9/22 - 9/27
Bush 45% Kerry 46%, Kerry +1
Time (877 LV) 9/21 - 9/23 Bush 49% Kerry 43% Bush +6
FOX News (1000 LV) 9/21 - 9/22
Bush 45% Kerry 43% Bush +2
Battleground (1000 LV) 9/20 - 9/23
Bush 50% Kerry 45%, Bush +5
CBS News (931 LV) 9/20 - 9/22
Bush 50% Kerry 41%, Bush +9
Zogby (1066 LV) 9/17 - 9/19 Bush 47% Kerry 44% Bush +3
IBD/TIPP (650 LV) 9/14 - 9/18
Bush 46% Kerry 43% Bush +3
CNN/USAT/Gallup (767 LV) 9/13 - 9/15
Bush 55% Kerry 42% Bush +13
CBS News (1088 RV) 9/12 - 9/16
Bush 50% Kerry 42% Bush +8
Battleground (1000 LV) 9/12 - 9/15
Bush 49% Kerry 45% Bush +4
IBD/TIPP (674 LV) 9/7 - 9/12
Bush 47% Kerry 47% TIE
Newsweek (1003 RV) 9/9 - 9/10
Bush 50% Kerry 45%, Bush +5
Zogby (1018 LV) 9/8 - 9/9 Bush 47% Kerry 45% Bush +2
Time (857 LV) 9/7 - 9/9 Bush 54% Kerry 42% Bush +12
FOX/Opin. Dyn. (1000 LV) 9/7 - 9/8
Bush 47% Kerry 45% Bush +2
CBS News (909 RV) 9/6 - 9/8 Bush 50% Kery 42% Bush +8
CNN/USAT/Gallup (778 LV) 9/3 - 9/5
Bush 52% Kerry 45% Bush +7
Newsweek (1,008 RV) 9/2 - 9/3
Bush 54% Kerry 43%, Bush +11

From a total of 55 polls released in September and October, Bush led in 47 polls, giving him a 85% chance of winning. See? Kerry just led in 5, giving him a "whooping" 9% chance of winning. The reality is that Bush came out of the Republican convention with momentum and never lost it. The majority of polls in September, before the debates and October, after the debate, showed no shift in people's preference.

The money issue is not in Obama's favor anymore. As of the end of August both candidates and their party had the same balance.

What now may be in McCain's favor is that the Obama campaign is spending its money faster and Obama still has to do major fundraising.

Do the parties have any sort of "rules" to follow when spending money on behalf of a candidate?

whitetower

regarding MN, they had a RNC convention and Mac'Palin were there just few days ago.
But Rasmussen poll yesterday came out Obama lead Maccain 52:44.
So, I think MN is coming back home to Obama.

Anyone have the "weighting" #'s on the BattleGround national poll? Seems off that every poll, including Rasmussen shows a tie or Obama ahead up to 6 points yet this ONE poll shows McCain up 2. Wow great #'s out of Colorado, Florida, and NC today. McCain's in trouble!

It's downrightly ludicrous how Republicans on this site are trying to spin all polls showing Obama leading
How in the world they can argue such an evidence, it's UNBELIEVABLE!
After the Palin pick I admitted the momentum had shifted to McCain, now with this whole economy turmoil they still find a reason to believe that McCain is somewhat leading, lol!
By the way there's a new poll showing a tight race in NC, let's see how you're gonna spin this one

I think Obama should make some visits to PA, MN, WI, and MI just to shore it up. It is and will trend further away from the Thugs.

Then he should CONCENTRATE on Colorado, Colorado, Colorado, maybe a visit to NH, and then Virginia, and Florida.

Forget Ohio.

Obama needs to do something in Minnesota. I'm tired of seeing nothing but McCain ads that say "change is coming". Maybe he is refering to Obama.

Let me ask: As I understand the poll questions are asked by calling people in their home. As there are so many, especialy young people, that only use cell phones, can this be enough to tilt polls in McCains favour considering Obamas lead among young voters?

@ Licinvs

I would think so, but in 2004 the cell phone aspect was overplayed. That said though, there is data that says "cell phone only" use has gone up 15 per cent since that election. I'm an Obama supporter, 27 years old, with only a cell phone. I have never been polled in any election.

Also in the poll are 2 senate races:

CO Senate:
Udall 48 to Schaffer 40

MN Senate:
Coleman 49 to Franken 42

@mike

LOL!

Ya, get your facts straight. Boom has always proclaimed we take Rasmussen polls with a grain of salt.

For example, when Rasmussen came out with state polling YESTERDAY with McCain up in 3 states, here is what he said:

"boomshak:

@kerrchdavis:

Why should I argue with these polls? My guy is winning 3 and statistically tied in PA where he should be getting crushed.

Lol."

and

"boomshak:

@Robi:

Why the hell would I "admit bad polls?" I LIKE these polls!"

Notice there was no mention of these polls being suspect for a lack of internals as well as a TINY sample size.

So, Boomshank, the fair and balanced poll analyzer is a complete hypocrite.

fail.

Q's polls have large samples (almost 3 times as large as many of right-leaning Rasmussen's polls), and Q's accuracy ranks 14th in 2008 (6 spots ahead of Rasmussen). See:
http://www.surveyusa.com/index.php/surveyusa-report-cards/

Q's CO poll has Obama up 4--almost outside the poll's margin of error. According to Q, Obama holds this lead despite the fact that he and McCain are tied among voters ages 18-34--one of Obama's strongest demographics. There appears to be potential for Obama to expand his lead in this state. This poll is the 3rd CO poll in a row showing Obama ahead there; PPP releases a CO poll later today that is expected to be the 4th in a row to do so.

In MI, Q (like Rasmussen yesterday) finds that both Obama and McCain have consolidated their bases--meaning there is little room for movement in the race there. Q shows Obama up 4 points in MI. This poll is the 7th in a row showing Obama ahead in MI.

MN continues to be close, according to Q. The last 4 polls of this state have had Obama ahead by 2.8, 8, 1, and 2. No MN poll has shown McCain ahead there since March.

Q shows Obama with a lead outside the margin of error in WI. Other polls of WI have shown a closer race. However, no WI poll since May has found McCain ahead there.

There is not likely to be a great deal of movement in these states, with the possible exception of MN. In CO, MI, and WI the two candidates are spending approximately the same amount of money. McCain has out-spent Obama in MN thus far, thereby keeping it close. With Obama moving ND resources into MN and WI, this McCain advantage will likely end. See:
http://politicalwire.com/archives/2008/09/17/battleground_not_much_bigger_than_2004.html

@bookshak:
"I mean, what use is a poll unless I know the party distribution was fair and accurate?"

What do you consider a fair & accurate distribution? Also, many states don't register by party, so how can you know what's accurate for those states?

Sorry if you've said before. If you did, I missed it. TIA

@ NW Patrick:

Forget Ohio? Are you insane? A state with 20 electoral votes where the race is in a statistical dead heat?

Obama's strategy differs from that of his predecessors in that he is concentrating on many more states than Democrats traditionally do. Not only does this force John McCain to spread his scantier resources more thinly across the map, it gives Obama more outs (to use a poker term) should something go wrong elsewhere.

Granted, McCain is probably up a couple points in Ohio right now, but it's not an insurmountable gap by any means. It's yellow on the map for a reason.

I used the cross-tab data and then adjusted the results using the 2004 exit poll demographics and found the following:

MN based on gender: Obama up 2
MN based on party ID: Obama up 1
CO based on gender: Obama up 5
CO based on party ID: McCain up 3
MI based on gender: Obama up 4
MI based on party ID: Obama up 6
WI based on gender: Obama up 7
WI based on party ID --- need to validate exit poll

Party ID continues to be a problem in these various polls. Sampling mismatches with the 2004 exit polls are common.

Some patterns are consistently emerging in national and state data.
1. McCain leads with males and Obama with females. Where Obama is leading a state, it is usually because he is far ahead with women. He has a built in advantage in this regard because more women will vote than men and in some states the difference is on the order of 54-46 or more.
2. Obama has a 90-10 or greater advantage with African Americans. In states with significant AA population, this advantage is very significant. He is also leading nationally and in many states with Hispanics.
3. In many states, almost all the Republicans are voting McCain and almost all the Dems are voting Obama, on the order of 90-10 or greater. Independents will make the difference in addition obviously to who actually votes.

I believe that we may not know the winner on Nov 4 or perhaps even Nov 5. I expect recounts and disputes. The other variable is the early voting which could mean 1/3 of all voters cast their ballot before the election day. Depending on which state and what events occur between now and then, early voting could affect the outcome.

I still believe Obama needs the Kerry states plus Iowa, New Mexico and Virginia. If Obama somehow wins Nevada, then he can lose New Hampshire. If he wins Virginia, then he can lose Colorado. Looks very close, on the order of a single state either way.