Well it looks like Dean Chambers has surfaced. Business Insider managed to get him on the phone and he admitted he was wrong — what choice did he have? — though he still manages to hypocritically point the finger at Republican-leaning polling companies and blame them for it.
“Nate Silver was right, and I was wrong,” Chambers said in a phone interview.
Chambers’ method of “unskewing” polls involved re-weighting the sample to match what he believed the electorate would look like, in terms of party identification. He thought the electorate would lean more Republican when mainstream pollsters routinely found samples that leaned Democratic.
But as it turned out, the pollsters were right — self-identified Democrats outnumbered Republicans by 6% in election exit polls.
“I think it was much more in the Democratic direction than most people predicted,” Chambers said. “But those assumptions — my assumptions — were wrong.”
And yet:
But he said he probably won’t go back to “unskewing” polls next time. He actually thinks conservative-leaning pollsters like Scott Rasmussen have a lot more explaining to do.
“He has lost a lot of credibility, as far as I’m concerned,” Chambers said. “He did a lot of surveys. A lot of those surveys were wrong.”
And the fact that you were even “unskewing” those polls and saying they were underpredicting the Republican landslide? You were even more wrong. And we’re all laughing at you.
16 comments
Skip to comment form ↓
holytape
November 8, 2012 at 11:10 am (UTC -4) Link to this comment
You don’t understand. It’s not the unskewing polls that was wrong. It’s this fake reality that is wrong. Obama cheated, and that is the only explanation. I don’t know how, or even care. but somehow that fascist-negro-socialist-muslim-atheist-communist-kenyan-anti-american-black-power-sold-out-white-dictator-spineless-wimp, along with ACORN and msnbc had tens of thousands of homeless illegal welfare mothers voting under fake names. These fake voters were shipped between polling stations by FEMA armored vehicles, which will now be used to ferry conservatives to death camps to learn French so that Satan can increase taxes on the wealthy hard working true Americans.
imrryr
November 8, 2012 at 11:15 am (UTC -4) Link to this comment
Dean Chambers is still as classy as ever. This is what currently greets you at the UnSkewed Polls website:
Actually, the comments page he points to is mostly just filled with spam, but still, that’s a hilarious bit of projection coming from the man who said this:
dingojack
November 8, 2012 at 11:17 am (UTC -4) Link to this comment
holytape – “… so that Satan can increase taxes on the wealthy hard working true Americans”.
Or ‘Stan’ as his friends call him.
Nah, everybody just calls him Satan, he doesn’t have any friends. (He’s been waiting for his brother, Jesus to come around for 2000+ yrs, not even a lousy phone call). :)
Dingo
blf
November 8, 2012 at 11:18 am (UTC -4) Link to this comment
Reptilians.
greg1466
November 8, 2012 at 11:25 am (UTC -4) Link to this comment
I suspect that part the problem with his method is people like me. Once upon a time, way back before I started actively trying to apply skeptical thinking, I found myself aligned with the Republican party and registered that way. I even voted for W the first time around. As I started to apply my natural skepticism to more/all of my beliefs, I realized that I am really a progressive Independent. That, in combination with the GOP going bat shit crazy, has turned the Republican party into toxic sludge as far as I’m concerned. So while I am registered as a Republican, and would answer so on a poll, I don’t vote Republican. I just have never bothered to change my registration.
Doug Little
November 8, 2012 at 11:54 am (UTC -4) Link to this comment
It’s gonna be interesting next election cycle to see if these guys commit the same mistake again. You would expect their predictions to start to align with Mr Silvers if they learned anything, I would bet good money that they won’t.
Dr X
November 8, 2012 at 12:07 pm (UTC -4) Link to this comment
Part of the problem is also that pollsters record but don’t weight values based on self-identified party affiliation. Particularly when people are surveyed, they prefer to identify themselves as independents. A lot of people who voted Republican did that this year. Self-identified Party affiliation is not predictor of outcome. So the pollsters weight values based on demographic characteristics, matching them to census figures. That generally cleans up the sampling biases up.
eric
November 8, 2012 at 12:21 pm (UTC -4) Link to this comment
greg1466:
Not as far as I know. The data used by everyone was from polls, not registrations. Chambers’ “unskewing” was based on the idea that either people lied to pollsters or that pollsters had a systemic bias against calling/talking to conservatives. So, doesn’t matter what you were registered. Assuming you would’ve responded honestly to a poll, you were accurately represented.
Doug Little:
I think Chambers will, because I think he’s a sincere believer; he really does think that the mainsteam polls have some sort of bias against conservatives. He’s going to keep mathematically correcting for a bias that isn’t there (or is statistically insignificant).
I think there are other conservative analysts out there now (and before the election) who support mainstream analytics. They are just ignored/invisible because FOX, Beck, etc. are basically ‘good news only’ purveyors. So the real interesting question is, IMO, not whether there are some conservatives who will actually pay attention to data. IMO, there are. The real interesting question is whether FOX etc. will change their operations enough to let realistic conservative voices have a stage.
Deen
November 8, 2012 at 12:41 pm (UTC -4) Link to this comment
Well, he’s right about that, of course. The thing is, Nate Silver and many others have pointed this out all along, explaining that those polls are consistent outliers, which is why these polls get low weighting values in their models. I hope Chambers is going to reconsider who to trust from now on…
garnetstar
November 8, 2012 at 12:47 pm (UTC -4) Link to this comment
holytape, you forgot “homosexual”.
cry4turtles
November 8, 2012 at 12:58 pm (UTC -4) Link to this comment
Hoo hoo hoo, hee hee hee, haa haa haa! That comes straight from the belly!!
Area Man
November 8, 2012 at 1:47 pm (UTC -4) Link to this comment
In another article, he brags that he “only” got 4 swing states wrong. That’s kind of like bragging that your pick for winner of the Superbowl was correct plus or minus 1 team.
tacitus
November 8, 2012 at 2:53 pm (UTC -4) Link to this comment
And that’s only because he realized that he was going to look even more like a complete fool if he continued to “unskew” the polls by the same amount all the way up to election day.
He claims that Rasmussen polls shift by several points toward Obama in the last week, but in reality that shift was far smaller than anything that would account for his reskewing of his unskewed polls. It was basically a preemptive CYA exercise.
(and here I have to refrain from commenting about how much “A” he has to cover…)
wscott
November 8, 2012 at 8:07 pm (UTC -4) Link to this comment
CBS News just had an interesting article on how & why the Romney team ignored the polls and overestimated their chances. One tidbit I hadn’t thought of: many polls showed Romney leading among “Independants.” And in past years, the candidate who pulled the most Independants won. But it turns out a large chunk of those Independants were former Republicans who just no longer identify as such, but still tended to lean right. The bottom line is there just aren’t as many registered Republicans as there were 8 years ago.
varith
November 8, 2012 at 10:44 pm (UTC -4) Link to this comment
I spent yesterday mocking the guy, but he did come out and admit he screwed up. If he wants to say he relied too much on Rasmussen, I’m not going to contradict him. He was being a crazy wingnut, but when faced with reality, he manned up. So he is one of the people I would at least engage.
martinc
November 9, 2012 at 1:48 am (UTC -4) Link to this comment
Doug Little @ 6:
Their predictions will align with Silver’s when Silver predicts a Republican landslide. The ‘alignment’ is purely of the ‘stopped clock’ variety.