Green Room

CNN Polling: 2 + 2 = 11?

posted at 1:59 pm on November 5, 2012 by

CNN has decided to throw in with Marist and Quinnipiac for the ‘Which Pollster Can Be The Most Wrong” contest, funnily enough being decided tomorrow night as well. They’ve got the race tied at 49 all, but with a very interesting sampling number that stands out in their methodology.

Among those likely voters, 41% described themselves as Democrats, 29% described themselves as Independents, and 30% described themselves as Republicans.

You read that right. CNN, the night before the election, has a turnout model in their poll that is 57% higher for the President than it was four years ago. So the economy hasn’t improved, wages haven’t improved, energy production hasn’t improved, the Middle East hasn’t improved, Defense hasn’t improved, health care costs haven’t improved, the debt hasn’t improved, the deficit hasn’t improved, partisanship hasn’t improved, MSNBC’s ratings haven’t improved, Iran hasn’t been deterred from their march towards nuclear weaponry, and we’re still heading towards the entitlement fiscal cliff.

But CNN is happy to report that the country is no longer a 50/50 country, and that the spread in turnout Barack Obama got in all his hopeychangey glory is now half again wider? And the race is still tied?

By the way, Scott Rasmussen finished October with a report that shows party ID favoring the Republicans by 5.6%. That’s almost a 17 point swing from what CNN shows. Just a reminder for today and tomorrow. Do not let what a poll tells you cloud your decision on whether to turn out or not.

Recently in the Green Room:

Blowback

Note from Hot Air management: This section is for comments from Hot Air's community of registered readers. Please don't assume that Hot Air management agrees with or otherwise endorses any particular comment just because we let it stand. A reminder: Anyone who fails to comply with our terms of use may lose their posting privilege.

Trackbacks/Pings

Trackback URL

Comments

Duh….

Turtle317 on November 5, 2012 at 2:03 PM

By the way, Scott Rasmussen finished October with a report that shows party ID favoring the Republicans by 5.6%.

Most signficant polling data of the day, imho.
If even remotely true, a Romney landslide is in the works.

Worth repeating: Rasmussen was the most accurate pollster in 2008.

Norwegian on November 5, 2012 at 2:03 PM

No cloud here.The wife and I voted straight R last week here in the NW burbs of Chicago. Fat chance that R/R will pick up Ill-noise but, at least I can let my displeasure known. I am sure the D’s have already cancelled my vote out with a few dead people.

BobK on November 5, 2012 at 2:05 PM

Math is hard

cmsinaz on November 5, 2012 at 2:06 PM

CNN, the night before the election, has a turnout model in their poll that is 57% higher for the President than it was four years ago.

This is totally insane. We all know that’s not going to happen.

INC on November 5, 2012 at 2:06 PM

By the way, Scott Rasmussen finished October with a report that shows party ID favoring the Republicans by 5.6%.

Yet it’s tied?

WTF?

portlandon on November 5, 2012 at 2:06 PM

Looks to me that The Dems have it in the bag, Most of them should stay at home, their vote will not be needed. If you believe the ranges in the CNN polling.

Guest1.1 on November 5, 2012 at 2:07 PM

You read that right. CNN, the night before the election, has a turnout model in their poll that is 57% higher for the President than it was four years ago. So the economy hasn’t improved, wages haven’t improved, energy production hasn’t improved, the Middle East hasn’t improved, Defense hasn’t improved, health care costs haven’t improved, the debt hasn’t improved, the deficit hasn’t improved, partisanship hasn’t improved, MSNBC’s ratings haven’t improved, Iran hasn’t been deterred from their march towards nuclear weaponry, and we’re still heading towards the entitlement fiscal cliff.

But CNN is happy to report that the country is no longer a 50/50 country, and that the spread in turnout Barack Obama got in all his hopeychangey glory is now half again wider? And the race is still tied?

Perfect.

It could not be better said.

Bravo.

farsighted on November 5, 2012 at 2:08 PM

Serious question for Duane, or anyone else:

Since the media has an obvious objective by skewing the sampling of these polls, why wouldn’t they lie about the sample? Wouldn’t the results of the poll be more effective for their narrative if they misrepresented the sample? Is their some law they must abide by to keep them from doing this? Just something I’ve always wondered…

revolutionismyname on November 5, 2012 at 2:10 PM

By the way, Scott Rasmussen finished October with a report that shows party ID favoring the Republicans by 5.6%.

Yet it’s tied?

WTF?

portlandon on November 5, 2012 at 2:06 PM

The tracking poll assumes a D+4 weight. If they were to re-weigh to R+6, Romney would lead 58-41.

Interestingly, the poll determining party ID has MUCH smaller MOE than the tracking poll. I believe something like 15,000 respondants were polled.

Norwegian on November 5, 2012 at 2:12 PM

Yet it’s tied?

WTF?

portlandon on November 5, 2012 at 2:06 PM

They are probably not using that party affiliation number in their polling – but it’s pretty amazing how that number has tracked the final margin of every election since 2004.

TarheelBen on November 5, 2012 at 2:14 PM

Look around and tell me how many Obama bumper stickers you see. If you manage to spot one, the driver is hiding behind sunglasses and a hoodie or a hat.

faraway on November 5, 2012 at 2:15 PM

Pleasepleasepleasepleasepleasepleaseplease…

Abby Adams on November 5, 2012 at 2:17 PM

Worth repeating: Rasmussen was the most accurate pollster in 2008.

Norwegian on November 5, 2012 at 2:03 PM

Rasmussen came close in 2008, but a number of other pollsters did better. Last I saw, Ras was ninth most accurate in that election.

AngusMc on November 5, 2012 at 2:23 PM

Well 2 + 2 = 100 (in binary). But to be truly accurate, it would be written as 10 + 10 = 100.

MichiCanuck on November 5, 2012 at 2:24 PM

It’s true – IN BASE 3.

Betcha you wingnutz didn’t think of that, didya, huh?

RINO in Name Only on November 5, 2012 at 2:28 PM

MSNBC’s ratings haven’t improved

Ha – good one, Duane!

Do not let what a poll tells you cloud your decision on whether to turn out or not.

Anyone who relies on polls to guide their voting decisions is a moron. Unfortunately, there are a lot of morons out there!

KS Rex on November 5, 2012 at 2:31 PM

The death knell of CNN is heartening. Go to the graveyard of history CNN and take NYT with you. Farewell

WV. Paul on November 5, 2012 at 2:31 PM

Rasmussen came close in 2008, but a number of other pollsters did better. Last I saw, Ras was ninth most accurate in that election.

AngusMc on November 5, 2012 at 2:23 PM

Ras was #1

El_Terrible on November 5, 2012 at 2:35 PM

1 + 1 = 11…

Ace ODale on November 5, 2012 at 2:37 PM

Look around and tell me how many Obama bumper stickers you see. If you manage to spot one, the driver is hiding behind sunglasses and a hoodie or a hat.

faraway on November 5, 2012 at 2:15 PM

Or involved in a car accident… my unscientific observation is that liberals can’t drive their Prius worth a da**ed.

SteveInRTP on November 5, 2012 at 2:38 PM

Anyone who relies on polls to guide their voting decisions is a moron.

I just flat do not get these folks, nor the ones who figure things are a foregone conclusion so they may as well stay out of the booth. All the prophecies in the world don’t matter if we don’t go and fulfill them.

mrsknightley on November 5, 2012 at 2:38 PM

Rasmussen came close in 2008, but a number of other pollsters did better. Last I saw, Ras was ninth most accurate in that election.

AngusMc on November 5, 2012 at 2:23 PM

Nope, Rasmussen was the most accurate.

http://www.fordham.edu/images/academics/graduate_schools/gsas/elections_and_campaign_/poll%20accuracy%20in%20the%202008%20presidential%20election.pdf

Rasmussen’s final Tracking poll in 2008 had Obama 52 – McCain 46.
Acutal result: Obama 52.9 – McCain 45.7

Whatever Rasmussen reports in the their tracking poll tomorrow is probably very close to the final margin. Given the movement towards Romney today, I expect a 50-48 or 51-48 Romney margin.

Norwegian on November 5, 2012 at 2:39 PM

Math is hard.

It is doubly hard when you are stupid.

It is downright impossible when you are a Progressive and in the tank for Barack Obama.

Resist We Much on November 5, 2012 at 2:44 PM

By the way, Scott Rasmussen finished October with a report that shows party ID favoring the Republicans by 5.6%.

Yet it’s tied?

WTF?

portlandon on November 5, 2012 at 2:06 PM

The tracking poll assumes a D+4 weight. If they were to re-weigh to R+6, Romney would lead 58-41.

Interestingly, the poll determining party ID has MUCH smaller MOE than the tracking poll. I believe something like 15,000 respondants were polled.

Norwegian on November 5, 2012 at 2:12 PM

Rasmussen is playing games here. What he’s doing is releasing “topline” numbers that look good for Obama showing ties, and then releasing data like this that show a Romney win (if not a landslide) in the works. Either way, when he’s asked about the election, he can claim he was “right.” If Obama wins in a squeaker: “My polls predicted a close race!” If Romney wins big, “Well, look at the party ID polling!”

My guess is Rasmussen may genuinely have no idea of what’s going to happen tomorrow night and is hedging his bets by playing games with the numbers.

Doomberg on November 5, 2012 at 2:52 PM

Math is hard.

It is doubly hard when you are stupid.

It is downright impossible when you are a Progressive and in the tank for Barack Obama…

Resist We Much on November 5, 2012 at 2:44 PM

… and a journalism school graduate.

farsighted on November 5, 2012 at 2:55 PM

Ok, I’m seriously, genuinely confused, and here’s why:

We have a slew of polls saying basically dead heat. Tied or someone has a 1 point lead. That’s the consensus today, and all the polls are hitting basically that conclusion.

But they’re all getting there via ridiculously different ways.
CNN? Tied with a D+11 sample.
Rasmussen: 50-49 with a D+4 sample
Gallup: 49-48 with a (if I remember right) R+1 sample
Then you’ve got NBC and WaPo saying Obama by 1 with (I think) D+5ish samples.

In other words, something funky is going on. They are clearly not interviewing the same types of people. If you apply CNN’s sample to Gallup’s methodology, Obama is probably winning bigger than 08. Gallup’s sample to CNN’s poll? 1980 all over again. I’m at the point of throwing up my arms in confusion, sitting down in front of my tv tomorrow night and praying for the best.

LukeinNE on November 5, 2012 at 3:28 PM

Reminder: The leftist-controlled press will do everything it can to convince Republican and Independent voters that Romney will lose so that you will stay at home and not bother to vote. If you believe these chicken-bone polls and not bother to vote then you are giving them the low turnout that is Obama’s only hope of winning.

CrustyB on November 5, 2012 at 3:28 PM

But they’re all getting there via ridiculously different ways.
CNN? Tied with a D+11 sample.
Rasmussen: 50-49 with a D+4 sample
Gallup: 49-48 with a (if I remember right) R+1 sample
Then you’ve got NBC and WaPo saying Obama by 1 with (I think) D+5ish samples.

In other words, something funky is going on. They are clearly not interviewing the same types of people. If you apply CNN’s sample to Gallup’s methodology, Obama is probably winning bigger than 08. Gallup’s sample to CNN’s poll? 1980 all over again. I’m at the point of throwing up my arms in confusion, sitting down in front of my tv tomorrow night and praying for the best.

LukeinNE on November 5, 2012 at 3:28 PM

You’re hitting on the reasoning why some people (myself included) are becoming convinced a Romney landslide could be coming. The polling this cycle has been bizarre and conservatives have been suggesting for some time that the pollsters have gone over the edge and joined the MSM as a Democrat advocacy/GOTV operation. They’ve basically put their reputations on the line with this polling.

I think that if it turns out to be correct that the pollsters have been co-opted by the left, we’re going to need to see about starting an explicitly conservative polling outfit in the future.

Doomberg on November 5, 2012 at 3:38 PM

This a case of knowing the answer they want, and then asking those who will give them that answer. The thrill is gone with far too many for this stuff to hold water.

RalphyBoy on November 5, 2012 at 4:13 PM

Obama loses the endorsement of the nation’s first elected African-American Governor, Doug Wilder, who endorsed him in 2008.

Wilder will make no endorsement this year, which really means he’s endorsing Romney, but would like to keep his car intact.

Resist We Much on November 5, 2012 at 5:45 PM