The polling this year is insane. Same state, same time period, wildly different results from poll to poll.
Rasmussen has Romney up 2 in Wisconsin; Marquette has Obama up 8. The Boston Globe (UNH Polling) has Scott Brown up 2; Suffolk has Elizabeth Warren up 7. And so on, and so on.
2012 will be to polling industry credibility what 2008 was to mainstream media credibility.
It’s not an issue of whether a particular poll is credible, but that the industry has lost legitimacy, much as trust in the mainstream media is at an all time low. The Jake Tappers of the world were tainted by the Andrea Mitchells of the world — you will see the same effect on the polling industry.
Or to put it diffently, the public is “averaging” the pollsters just as the pundits are averaging the polls, and on average the credibility of the industry is tanking.
The polling industry — but not necessarily every pollster — may always have been corrupted by partisan messaging. But this is the year in which the public woke up to it.
Here’s today’s case study:
So let’s wrap our mind around this one, gang. Princeton Survey Research Associates did the Pew poll from 10/24 to 10/28. Found a tie.
— Jay Cost (@JayCostTWS) October 31, 2012
Princeton Survey Research Associates did the National Journal poll from 10/25 to 10/28, found Obama up 5.
— Jay Cost (@JayCostTWS) October 31, 2012
There’s no better e.g. of how polling *assumptions* affect outcomes than Pew v. NJ. Same & data collector. Different LV screens.
— Jay Cost (@JayCostTWS) October 31, 2012
Use this as a cautionary tale, gang. Read in tandem with my piece today, reinforces this point: DIFFERENT ASSUMPTIONS = DIFFERENT RESULTS.
— Jay Cost (@JayCostTWS) October 31, 2012
I’ll take this one step further: predictions built on polling averages (of any kind), e.g. 538, are seriously problematic for that reason.
— Jay Cost (@JayCostTWS) October 31, 2012
DONATE
Donations tax deductible
to the full extent allowed by law.
Comments
It is looking like yet another area the democrats have taken over is being exposed. Thanks to the new media, we are able to see the fudging of the numbers and most importantly, expose it.
With any poll now, the first thing I do is look for the internals. I know I am not the only one who is suspicious of all polls and disregards the media polls.
Cheer up everyone Happy Halloween!
As a Pole I resent the implications…oh, wait. Poll?
But, are you painless?
(sorry, could not resist.)
Why?
Is he a dentist?
In the Army?
With equipment?
Naughty girl, Valerie. Naughty girl!
It’s very clear Democrats and their sycophantic press (which includes certain polling organizations) don’t much approve the final narrative, less than a week out from election day, that Romney is surging – not only in conventional swing states, but into traditional Blue states.
So the Democrats and their allies have to counter it with a false narrative that the necessary states are falling into line. That’s what these polls are meant to suggest. Released no doubt after Hurricane Sandy, suggesting “Frankenstorm” reset the presidential race.
It is as much psyops to keep the Democrat’s base in the game as much as it is a last minute attempt to suppress Republican turnout.
This nails it. When I look at the RCP ‘averages’, I always take care to remove anything from PPP and CBS (for starters).
To borrow from Galileo, one person with the facts is right. I don’t care what the ‘average’ is, I care about who gets it right.
Agree with tazz. Whatits become is a “buyer beware” sort of thing. Stick with RCP and sift from there.
My standard is to ignore polls with a sample less than 1000-1500 or so. I ignore things like we see with the National Journal poll with small samples.
Rasmussen + Gallup + Pew Research pretty much paint an accurate picture of how things are. To improve things, a few suggestions: There should be some “standards” so we know were comparing “like” apples and apples.
Polls should have available the demographics/ makeup of Republicans and Democrats in the sample. Further, Id like to see the breakdown of the area codes for the respondants. The area code breakdown would surely help folks understand the sample more completely…so were not looking at results driven by location in an uneven way.
Obviously if a sample is composed of 40% New York area codes then we have to know that impact.
RCP are not screening polling outfits.
Today RR average dropped because of one unknown lotosying a 5 + Obama national
Lotto = ‘polling outfit’.
Crazy
Vote to preserve individual dignity and to recognize an intrinsic value of human life. Vote your conscience.
That said, no one can divine the future. Life is an exercise in risk management. The best we can hope for is to mitigate its consequences.
The lesson is a simple one, poll results are only valuable to the person who paid for them, to everyone else they are of NO real value.
What about newsmedia polls? They might as well publish their daily lunch selections, those might actually have some use.
Publicly released polling is about as useful as planning EMS services based upon a random set of EKGs.
I don’t need a stinking poll to tell me what is going to happen on Tuesday. Think of the elections since 2008, Christy, Brown, 2010, Chick Fil La, that is all you need to remember. Nothing has changed since then, nothing at all.
Polls are statistical, and as we all know, 87% of all statistics are made up! Just like the one I just cited 🙂
Exactly. The idea that you would take Nate Silver (or RCP) more seriously than you would Gallup is ludicrous. Gallup has been in the business since 1936 and they have gotten a total of two elections wrong: the infamous Dewey vs. Truman in 1948 and Carter – Ford which they had 48 – 49, but ended up 50.1 – 48.1.
There were a few others that they didn’t get exactly right:
Bush – Kerry in 2004 they had 49-49, it ended up 50.7 – 48.3.
Bush – Gore in 2000 they had Bush 48-46, it was 47.9 – 48.4. They got the Electoral result right and the popular vote wrong.
Its hard to ding them for the ones that they had as too close to call and the result ended up within margin of error.
Gallup has Romney up by 5 with a margin of error of 2.
The reason I use Gallup as an example is not because they have the largest lead for Romney but they have the longest track record, make their living on the trustworthiness of their polls, and don’t have a reputation for partisan bias.
New Site: Teaparty News
http://www.tpnn.com/
^Maybe Prof J will have them linked on right
I like Rush Limbaugh’s explanation best, that there’s probably a method to these ridiculously skewed polls:
a) Drive the narrative, buoying Democrats and depressing Republicans;
b) If Romney wins, provide a basis for the Democrat/media/U.N. cabal to declare the election invalid/stolen, Romney a thief/usurper, and any mandate illegitimate.
2012 should also do marvelous things for Americans’ respect for, and belief in, the hollyweird degenerates.
I think everyone going to WI probably means that both camps think it’s possible O is winning Ohio, which would make WI very very important to the Romney camp. It’s just anecdotal. However, it’s becoming unrealistic to keep assuming that every single poll out of OH has to be wrong. In 2008, we played this wishcasting, the polls are screwed game and lost, but we’re a lot closer than we were in 2008.
I hope we win this but I am again getting eeyorish. I wonder if Ohioans are still among America’s least educated for 2012.
Ummm…yeah.
Romney leads independents by about 15 points (give or take) in Ohio polls. GOP turnout is expected to be high. And Romney is the one in trouble?
Real analysis: assuming I+15 for Romney (a nationwide pattern, not just Ohio), D+7 (37D 30R 33I), and slightly more Dem support for Obama than is likely (94%), it’s still a tie. Move that to a more realistic D+2 (36D 34R) and Romney gets 51%. Move that to a more likely D+0/R+0, and it’s Romney 52%. Assuming a very conservative I+10 for Romney, and you have to get D+6 turnout for Dems to win.
Now, there’s certainly fraud: Somalis being bused around by Democrats, voting machines registering Romney votes for Obama.
While nothing is certain, but I’m inclined to be optimistic.
You sound like a lemming. Come on let’s not act like we did in 2008.
You don’t sound eeyorish to me. You sound like you’re aware of the uncertainty and risk.
No I am equal parts pissed off. Equal parts sad. Watching Christie giving Obama a fellatio on national tv will do that.
And the polling out of the states especially OH is horrible for us. NOT all of it is wrong. But all of it is horrible.
Let’s NOT ACT like we did in 2008. This is just silly. Some of you, LITERALLY, want the news and reality embargoed.
You want the full cocoon, not the three-quarters cocoon a partisan advocacy blog already is.
And let’s not forget this quote, never underestimate the stupidity of the American electorate as a whole, this quote sums it up:
The danger to America is not Barack Obama but a citizenry capable of entrusting a man like him with the Presidency. It will be far easier to limit and undo the follies of an Obama presidency than to restore the necessary common sense and good judgment to a depraved electorate willing to have such a man for their president. The problem is much deeper and far more serious than Mr. Obama, who is a mere symptom of what ails America. Blaming the prince of the fools should not blind anyone to the vast confederacy of fools that made him their prince. The Republic can survive a Barack Obama, who is, after all, merely a fool. It is less likely to survive a multitude of fools such as those who made him their president.
The problem was all those dishonest people that vouched for him, and the compliant media that spiked stories about him.
We are in the process of replacing our news media, and the individual talking heads will never live down their ignorant pronouncements. People like me will see to it.
There is also this quote from Benjamin Franklin at the Constitutional Convention:
I trust it hasn’t come to that yet, but such things happen faster than is imaginable beforehand.
The train left the station a long time ago.
Plenty of amateurs have been severely injured by poles .
They need to be secured at the top & floor.
special grip fluid is needed to hold on.
Then there is the universal problem of not being able to do the splits.
Poles are the boulevards of broken dreams & bung showers.
Shoulders. But showers could be a pain with use of one arm .
I think that this whole “Obama is leading” Operation Demoralize is having the reverse effect. My wife has never been more involved and never given more time to a political campaign than 2012. Every day she watches the polls and every week in the last five she’s donated and gone out knocking on doors and every little dip in Romney’s polling makes her wonder how she can do more.
I can’t recall any fact based research that would lead the Dems and their media familiars to think that bogus polling and their transparent attempts to depress Republican turnout work. I’m glad in a way. They can keep humping this particular leg as long as they want. It stops them from finding a truly effective Republican voter suppression strategy.
An interesting alternative to the usual suspects of polling is the website “Unskewed Polls”.
http://unskewedpolls.com/
Unskewed polls was a parody account on twitter before they started unskewing polls. You can’t unskew polls.
So you are the resident expert on poll unskewing? I think you give yourself too much credit. As far as the news goes, we just want the truth and if you’re trying to get that from the MSM it ain’t happening.
If you are following https://twitter.com/NumbersMuncher he is a good source for poll analysis. He’s like our Nate Silver.
You can’t unskew a badly skewed poll, all you succeed in doing is doubling down on a bad sample.
Let’s NOT ACT like we did in 2008. This is just silly. Some of you, LITERALLY, want the news and reality embargoed.
You want the full cocoon, not the three-quarters cocoon a partisan advocacy blog already is.
Righties here: “The Coming Landslide” by Dick Morris in today’s ‘The Hill’. I’m not president of Dick’s fan club, but old(reformed)smarmy boy has been saying for longer than any of the other experts that Mitt is going to enjoy a real decisive victory. I’ve been smoking his hemp since mid-summer and and sleeping with a smile and inner comfort. Want to share the same pipe..? Go to today’s realclearpolitics.com and click the link for Dick’s column. Ahhhhhhhh…That’s some gooooood Republican stuuuuffffff…
I’ve been waiting for this post…
Iowahawk has your daily reminder.
SMILE: Karl Rove’s piece in the WSJ today is a good ‘bottom line’ back up for reality based optimism on Tuesday.
I’ll take it from Uncle Karl: R-52% O-48%