One More about the UnSkewed v. Reality Discussion
Wherein I get a bit petty (but to make a point and, maybe just because it amuses me).
I was pleased (and a bit surprised) to see Doug Mataconis’ post in which Dean Chambers noted three key things:
1. He was wrong about the alleged “skew” in all the polling:
2. That Silver was right (although I think Chambers owes him an apology for the ad hominem in his Examiner piece); and
3. That Rasmussen has some ‘splainin’ to do.
Now, readers of OTB are quite aware that I found the entire unskewed bit to be problematic from the start because it was based on flawed premises and even more flawed methodologies. Further, over time, it became quite clear that a lot of people simply did not understand partisan identification as a variable.
These posts caused a number of commenters to get rather upset, because they were certain that mainstream media outlets were paying the polling companies to cook the books to produce pro-Obama numbers as a means of depressing the Romney vote (that this was counter to both the interests of the pollsters in question, as well as to decades-worth of polling, did not matter).
All of this was profoundly frustrating to me because the positions being taken were the stuff of alternative realities founded on fantasies and preferences rather than empirics and sound thinking. And I would underscore: I would have been similarly frustrated had Romney been up in the polls and Democrats been feverishly unskewing the samples.
Of course, those who thought the polls were wrong thought I was upset because of what they assumed was my preference in the election. However, as I noted in one of the comment threads: I am not a cheerleader (and I am certainly not a partisan pundit seeking to boost my side). So while yes, I have preferences, my main goal is understanding. As I have noted on multiple occasions: I would far rather be correct in my analysis than to have my preference come to pass.*
(So here’s the petty part that you came for).
One commenter in particular, one Smooth Jazz, harangued we incessantly in these posts. He assured me that Nate Silver was a “hack” (he had to be dontcha know, because he works for the New York Times!):
Nate Silver is a left wing hack, whose complex model puts more weight on a month old PPP or Marist OH model that showing Obama up 10 with a goofy and outrageous partisan split, compared to a more recent Rasmussen sample that shows the race in OH tied because he prefers the more partisan left leaning pollsters such as PPP and Marist compared to Rasmussen. ANY model is susceptible to “garbage in, garbage out” so spare us the propaganda.
[…]
The NY Times is an activist organ devoted to the takedown of anything and everything Repub. I realized you were far left but give us a break: The NY Times is every bit as, or more, partisan as Fox.
Indeed, he frequently noted that all the mainstream, established polling was wrong because, well, just because. For example: “They call it as they see it, unlike the NY Times & NBC that has to rig their polls to push their agena.” As well as:
Finally, I and many other Repubs do not trust Bill McInturff or the news Division of the Wall Street journal. McInturff wants NBC/MSNBC $ and sponsorship and I wouldn’t put it past him to fall in line with his Obama pay masters at NBC/MSNBC, an entity known for doctoring tapes to push the Liberal agenda. Besides, McInturff is the classic “Inside the Beltway Repb” that sucks up to Liberals so they can get invited to DC/NY cocktail parties and get their $.
Indeed, it was supposed to be a pretty big conspiracy (only a few can be trusted!):
I have absolutely no problem trusting Gallup. Unlike Quinnipiac, Marist & other polling entities sponsored by Obama sychophants NY Times, CNN, ABC, NBC, CBS, etc, Gallup has no vested interest in juicing their polls or screening more Libs to push a Liberal agenda their paying sponsors broadcast to the world. And YES, when Gallup had Obama JA at 43% I said that was trouble, and if he got near 50% he would be in great shape like all incumbents in our history running for reelection.
If Rasmussen starts showing the same trend as Gallup, as both entities did after the Dem convention, then YES I would concede Romney is solidly behind, especially if the turnout models show a reasonable Dem 2%+ versus Repubs give or take. These are the only regular national polls I trust as their turnout models tend to be more representative. All I’m saying is that Dems voting by 9% – 10% more than Repubs this year, when in 2008 it was Dems 7%, is not credible. As boring as Romney is, there are still a lot of Repubs psyched about throwing out Obama that Rep turnout should carry over some momentum from their heavy turnout barely 2 years ago. Reps are going to turn out in masse this year because they hate Obama, not because they like Romney. Quinnipiac polling for NY Times and finding 9% – 10% Dem turnout margins in generally even starts like OH & FLA is not credibe.
I would trust Quinnipiac and Marist more if they weren’t polling for NBC, CBS, NY Times,etc.
Here’s a fun one about Ohio:
he is not winning OH, not matter how many Dem 10%+ polls CNN trots out. Remember, it is barely 2 years since OH elected Gov Kasich & Sen Portman, and those Rep voters will be back out in force in 2012.
Here is a stat that should make you shutter: According to the various Secy of State County officials in OH, as of 10/26 220,000 fewer Democrats have voted early in Ohio compared with 2008. And 30,000 more Republicans have cast their ballots compared with four years ago. That is a 250,000-vote net Rep increase for a state Obama won by 260,000 votes in 2008. That is not based on a bogus CNN & MSNBC poll with a juiced Dem sample; That is a cold hard stat from OH election officials.
Good luck drinking the Nate Silver 99% probability Obama wins koolaid.
I would note, and I can think of no better way of saying this, but in regards to the quality of these outlets: I told you so. Indeed, Silver’s map was remarkably on track: he got VA correct and had Florida a light blue, which comports with the inability of anyone to call it yesterday.
(BTW, there are just some of SJ’s greatest hits. I don’t have time—indeed have taken too much already—to look for more).
May I note for the future: if one wants to make arguments, evidence and knowledge are quite keen. Bluster and ignorance? Not so much.
*While one caveat: I often have to pick against the Cowboys in NFL Pick’em, and in those instances I would rather be wrong and have my preference come to pass. I would gladly have traded the point I earned weekend before last for a Dallas win over the Giants. Of course, I don’t see politics as sports, which a lot of people do.
Speaking of reality…Where’s Jan???
The real question is: what will the anti-math and anti-science Republicans do now that their chief validator Rasmussen has suffered a fatal blow to his credibility?
Scott Rasmussen is less credible than the Unskewed Polls guy.
Not only in terms of results (he was wrong in 88% of swing states) but in terms of methodology (weighting by party ID). And his assumption of R+4 party ID is perhaps the single most aburd and discredited metric this election cycle.
Gallup at least has their RV results to fall back on (RV called the race very well, LV were a debacle though).
Hooray for Nate Silver …
…Welcome to something new for Republicans … Reality!
The thing that Smooth Jazz and other commenters elsewhere that I have dealt with on this don’t seem to get is that Pollsters are in the business to make money, and while election year polling is their biggest score, they do polls constantly for local governments, news outlets and businesses during the down cycles.
No local government looking for a tax increase is going look to pay a bad pollster. There’s no “reward” to a polling firm for being overly friendly to Obama, or liberals, or the SEIU or whatever boogeyman. The only reward is being seen as trustworthy enough that someone will want to hire you next month.
I guess they assume that because Rasmussen polls are designed primarily to make republicans feel good about themselves, that all of the other companies are doing the same for the opposition.
I’m sure Smooth Jazz will have the integrity and courage to acknowledge he was wrong.
Princeton Election Coalition was spot-on as well.
I had been very nervous about this election.
When NH came across in the range predicted by the aggregators…I knew the statistical bias theory was bunk and it was all going to be OK.
But no possible vindication will ever compare to the Rove “Whiney-Assed-Titty-Baby” tantrum. Pure unmitigated…and yes, petty…schadenfreude.
@C. Clavin: I caught the Rove thing live, an d it was something to see. In honesty, I was less struck by Rove, who at least was trying to make sense of the numbers, and more taken by Chris Wallace’s attempt to make it into Florida 2000 based on an e-mail from the Romney campaign, but the most amazing part was Megan Kelly grilling Fox’s decision desk guys and treating them skeptically.
@Steven L. Taylor:
But that did produce the most awesome quote in the history of Fox News (displacing even “terrorist fist jab”):
They soon will have a whole lot of company in not listening to or caring about what Karl Rove says.
While Mr. Taylor may be drinking Nate Silver’s Kool Aid.
Look’s like SJ et.al. are drinking the Fox News Fool Aid.
@ SLT and Geek…
Rove took hundreds of millions of dollars from donors…and returned nothing…less than nothing, really…because they lost a Senate seat.
I suppose that if I failed on such a massive level I would be throwing tantrums too!!!
@C. Clavin:
Maybe he can get a gig working for Stan O’Neal or Chuck Prince.
@Steven L. Taylor: The Big Tell about the Karl Rove meltown is when Meagan Kelly got half way through her frog-march down to the math den in some bunker below Fox news and said, “during rehearsal, this is where my ear piece stopped working.” So the whole thing was staged, likely devised to humor the easily-humored for what they knew was coming.
Rove’s melt-down? Well, that’s another matter. Upon seeing his cred as a pundit collapse in about five seconds Rove received a mighty blow, especially after spending so much money. The zero-sum of politics is just brutal, but there I go with the math again…
@MM:
Funny isn’t it the party who claims to believe in the “invisible hand” – that competition between buyers and sellers that channels the profit motive of individuals on both sides of the transaction such that improved products are produced and at lower costs – didn’t believe the data the pollsters were producing.
@Ben Wolf:
I’m not holding my breathe. Just like I’m not holding my breathe that Jan will confess to the error or her ways, even though she said she would.
Wimps!
I’m a mathophobe who can barely multiply by 10 in my head. My one statistics class was a nightmare. Don’t even ask me about my Econ 1 classes. But even I understand a bit more about probability than a number of pundits (some of whom continue to misunderstand Silver).
If you tell me I have a 2/3 chance of winning $100 and I don’t win the $100 that doesn’t mean you lied. (You might have, of course, but we’re assuming a fair game here.) I also had a 1/3 chance of not winning.
People like Joe Scarborough completely baffle me. He’s got to have bet on something in his life (football, basketball, the stock market, pool, something). But he and a whole lot of other people seem not to understand that what Silver did, basically, as I understsand it, was to take the data from all the polls, assign weights and probabilities based on various factors (the “art” part of the process) and run simulations to see how often Obama came out on top and how often Romney did.
It is possible that all the polls were biased or had faulty methodology, esp. with a 10% response rate and the issue of cellphones. It is possible that during one of these elections there will be a major meltdown. It is also possible that one pollster might, consciously or unconsciously, produce biased samples. But as along as there are so many polls from so many different sources, one can pretty much depend on the raw data.
I think our pundit class needs to spend a couple hours flipping a fair coin and writing down the outcomes.
@LC:
Actually, a lot of that was based on statistical analysis of polls over time. There was one post (that I cannot now find) that showed what the probabilities are of a given poll being correct for each percentage point above a tie that a given candidate has. In other words, while there is some art to constructing the model, most of it is based on statistical tools.
Indeed. Along with rolling various D&D dice.
Oh please…have you seen the whining today from conservatives and Republicans? Reality is still a place they live very far away from…
This, of course, makes perfect sense, considering how he was such an integral part of the Bush presidency…
I wonder whether some commenters are paid shills. There’s an article to that effect that’s been going around the internets. It sounds like some conspiracy stuff, but I’d actually be surprised if various entities WEREN’T hiring people to influence opinion in blog comments.
Chris Hayes tweeted the rhetorical question , ” Can you die of schadenfreude? ”
I think that Steven, and the rest of the Team Data are overdosing on schadenfreude today. :-).
@C. Clavin: She’s off tending her unicorn ranch, I guess…
To OTB:
Your polls were correct. The polls, analysis, and gut feeling I relied on were entirely wrong.
You’re happy. I’m not. In fact I am quite discouraged.
However, there are three aspects of this election to be grateful for:
1) By Romney not winning there are no riots (that were previously threatened), nor any charges of suppressing the vote.
2) The civil candidate exchanges following elections, including a peaceful transition to another term of power.
3) The continuation of some kind of checks and balances of government overreach, because of the ‘loyal opposition’ in the house House, in whose hands it still remains.
@jan:
Jan dear, yesterday should have been a learning experience for you. It’s time to start getting your news from non crazy sources. Assuming you prefer accurate information that is.
Yes Jan riots were predicted…no, wait…Civil War if Romney did not win.
Obama, Head said, will “try to give the sovereignty of the United States away to the United Nations. What do you think the public’s going to do when that happens? We are talking civil unrest, civil disobedience, possibly, possibly civil war … I’m not talking just talking riots here and there. I’m talking Lexington, Concord, take up arms, get rid of the dictator. OK, what do you think he is going to do when that happens? He is going to call in the U.N. troops, personnel carriers, tanks and whatever.”
https://www.outsidethebeltway.com/texas-judge-says-obamas-re-election-will-lead-to-civil-war/
Except for this riot.
@ Jan
So, lesson not learned, you are going to keep right on going with the crazy uncle BS. Your own version of “Four more years”, I suppose…
@Nick:
I don’t think there is much doubt about that. Remember “Moderate Mom” here at OTB. I’m guessing 23 years old, male, anger management issues and an inflated view of just how clever he is. Could’ve been working for free though…
Nate Silver: 1. Assertion based punditry: 0
What a laugh. I’m sure you got this information from the same set of hacks you used to predict a big Romney win. Which means you’ve learned exactly nothing.
Keep the bitter stupid coming, Jan. it’s always good for amusement, kind of like Michele Bachman.
@jan:
I am going to have to join the chorus and call you out on this as well. All of that riot talk was generated by the same alternate reality media machine that had you convinced that Romney was going to win.
I am not asking you to change your political philosophy or preferences, but I am asking you to take what has clearly unfolded here and use it to reassess some of the sources from which you have been gathering information. There is now a pretty substantial pile of empirical evidence that you (and many others) have been mislead. And if you both want to be taken seriously outside of that particular media bubble (and if you want your candidates to win in the future), such a reassessment is very much in order.
Reality has a way of winning. And this is not a gloat about the results of the election–it is a heartfelt plea to recognize that there are a large number of people out their shoveling BS because it makes them money or because it makes them happy (or both). But that BS is leading you astray in your perception of the world around you.
Jan’s sad and idiotic “riot” comment demonstrates the power and the problem of extreme partisan media and it’s audience. Rather than appealing to Lincoln’s “better angels of our nature,” the type of sites that Jan follows appeal to the “deepest fears of it’s reader’s souls.”
Of course she believes there would have been (race) riots if Obama lost (though she and most of the sources she reads would never dare address the race part). That’s because she, like other extreme conervatives, has a fear of urban minority dwellers and sees them only as takers (with the exception of the few well spoken ones who are members of the Republican party).
It’s also why she truely believes that there’s a Voter ID problem in the county — it’s fear of her vote being stolen. Likewise it’s why she also believes that all of the Muslim world is out to destroy America. Or that there’s not possibility that man-made climate change is happening (that would make her part of the problem).
Thees are all fundamentally fear based positions (just as a lot of extreme left wingers were predicting that because of Tag Romney’s part ownership in a Electronic Voting company, the Republicans were going to steal this election). And what she is looking for is validation that those fears are real and grounded.
And because of that — because she wants writers to tell her what she wants to hear — she frequents their site and quotes them as Gospel despite their past records (see Bob Krumm who posted reams of complex goobldy gook on why Nate Silver was wrong… Jan failed to note that he has been spectacularly wrong on most of his predictions… see her posting stats from a facitious polling company as Gospel).
This is especially a problem for Republican and Conservatives right now because of how well established Conservative Inc is. It will become a problem for Liberals if Progressive Media reaches that same success (god I hope it doesn’t).
But the biggest problem is until Jan and those like her on both sides, are willing to actually be critical (read and seriously consider material that challenging their prejudices) and demand more of their sources, things aren’t going to get any better.
Summary version of my last comment…
@Steven L. Taylor:
The challenge is that the utter BS that Jan is reading makes HER happy as well.
Maybe not happy in the sense that she’s feeling good right now. But she’s reading what she’s reading and uncritically accepting it because it’s confirming everything that she believes about herself (that’s she’s a good hard working American, fighting on the right side of history against overwhelming odds and a nation and world that’s against her) and everything she fears (that the country is going to collapse and the commies, Islamists, third columnists, urban savages, and Chinese are going to come for her one day if we don’t get them first).
And that makes her feel good on some level. Or at least right. And it’s that comfort with always knowing that you are right that is the real enemy here.
The sad thing is that the gullibility we see with Jan is *exactly* what I deal with w/my own mother.
She sends me a wingnut email.
I debunk it.
Silence. No response.
Time passes.
Next wingnut email forward.
Or she gets scared by some “investment video” (total scaremongering the whole world economy is going to collapse in 6 months stuff). I tell her this is extremely overblown, transparent kookery (you’d have to see the video I’m thinking of, but man it’s just so obvious). I close with “remember this in six months.” And of course in six months the market was up and things, while not great, were muddling along. I think that sunk in. I think. And I think it’s because my mother knows full well that there are scammers out there who rip people off. I don’t think it occurs to her that the equivalent exists in politics.
In the end, these people are preying on gullible folks who have certain beliefs. People who really appear to want to be lied to. And like matt, I worry about the possibility of a liberal version in the future.
But I gotta say… the stuff my mom sends me sets off my bullshit detectors immediately. It would set them off were it “liberal” instead of “conservative.” Really. Truly. Honestly. There is a certain style to all of these things. It leaps off the page.
And it really works on some people. It mystifies me. And scares the shit out of me, too. One of my fears is that mom, as she ages, eventually starts really falling for the crazy “investment” stuff. So far, no. But man, there’s a chance.
@Rob in CT: I think I watched part of that video a couple of days ago. Is this the guy who is forecasting 30-100% inflation by 2015 or something?