Slog
News & ArtsTuesday, October 2, 2012
2012 Mitt Romney: Skewed or Screwed?
Posted by Goldy on Tue, Oct 2, 2012 at 9:21 AM
There are two ways for Republicans to approach President Barack Obama's growing lead in the polls. A) Introspection: a reasoned self-examination of why voters seem to be rejecting their message and their candidates; or B) Denial: a fit of paranoid delusion that blames the widening gap on a conspiracy among pollsters to skew the results in favor of the Democrats.
According to a new PPP poll (most likely skewed), 71 percent of Republicans have resorted to denial.
Okay. So let's say Republicans are right, and the presidential polls are skewed in President Obama's favor, intentionally or not. So how unskewed would these polls have to be to give Mitt Romney a reasonable chance of winning the White House? That's the question Darryl over at my old haunts HA asked himself, rerunning his regular polling analysis* while introducing a bias in Romney's favor. Darryl's conclusion?
- horsesass.org
- Recent polls would have to be skewed 6 points in Romney's favor to give him a narrow electoral lead.
The graph is clear…to eek out a win, Romney has to move the electorate across the board by a remarkable 6%. That is, he is 6% behind in the polling now.
[...] To win with at least a 95% probability, Romney needs to shift things by 8%. To have the kind of lead that Obama now enjoys—with a solid 100% probability of winning—Romney needs a 10% shift.
That's an awfully big skew Republicans are imagining. If Romney goes on to win in November, this could prove to be the biggest polling fuck-up since "Dewey Wins!" Or... this could just prove how incredibly pathological the Republican Party truly has become.
* FYI, Darryl is a professor at the UW who does this sort of statistical analysis for a living. In 2008 his meta-polling analysis predicted the outcome within two one electoral votes, closer than the much ballyhooed Nate Silver.
1
I'm worried about voter turnout. If Obama supporters believe strongly enough that he has it in the bag then they might sleep in on election day.
3
Polls that are in the tank for Obama include known liberal-leaning pollsters "Fox News" and, most recently, "Rasmussen".
How will 'wingers unskew the poll that they used to skew their reported polling results?
How will 'wingers unskew the poll that they used to skew their reported polling results?
4
Nothing can keep Obama supporters from the polls. They believe in this president. But we need "independents" or as I like to call them, people with short memories.
6
The poll denialism on the right is simply laying the groundwork for blaming Romney's (likely, but not inevitable) loss on the media.
If you can convince a substantial proportion of your supporters that the press is simply not to be trusted and are in the bag for the other guys, then you really are never going to be held accountable for anything that you say or do.
And of course, if Romney actually does win, they get to point to the inaccuracy of the polls as evidence of liberal media bias.
It's a hedging strategy, plain and simple.
If you can convince a substantial proportion of your supporters that the press is simply not to be trusted and are in the bag for the other guys, then you really are never going to be held accountable for anything that you say or do.
And of course, if Romney actually does win, they get to point to the inaccuracy of the polls as evidence of liberal media bias.
It's a hedging strategy, plain and simple.
7
@2 haha you are delusional
As long as Obama doesn't show up in Muslim garb carrying a Koran while flanked by the Kenyan embassy, he will win the election.
As long as Obama doesn't show up in Muslim garb carrying a Koran while flanked by the Kenyan embassy, he will win the election.
8
@5 He postulates the chance of a tie to be 0.6 percent likely. "Fortunately, such an outcome remains quite unlikely. Of the 25,001 simulations that we ran on Monday, a 269-269 tie came up in 152 model runs, or about 0.6 percent of the time."
9
GDfR do you even read the articles you site? To quote.
"Fortunately, such an outcome remains quite unlikely. Of the 25,001 simulations that we ran on Monday, a 269-269 tie came up in 152 model runs, or about 0.6 percent of the time.
Still, this probability has roughly doubled from a few weeks ago, when the chances had been hovering at about 0.3 percent instead."
I'll speak slowly for you now GDfR a 0.6% likelihood of something happening is very, very, very, very..... small.
Nate Silver's article was a pot boiling article. "Oh I've got a deadline and an article due, well there is nothing meaningful to write about so I'll write about something completely unlikely and meaningless. Gotta keep those checks coming in."
"Fortunately, such an outcome remains quite unlikely. Of the 25,001 simulations that we ran on Monday, a 269-269 tie came up in 152 model runs, or about 0.6 percent of the time.
Still, this probability has roughly doubled from a few weeks ago, when the chances had been hovering at about 0.3 percent instead."
I'll speak slowly for you now GDfR a 0.6% likelihood of something happening is very, very, very, very..... small.
Nate Silver's article was a pot boiling article. "Oh I've got a deadline and an article due, well there is nothing meaningful to write about so I'll write about something completely unlikely and meaningless. Gotta keep those checks coming in."
10
In other Nate Silver news, his nowcast on the Presidential race shows 97.8% chance of President Obama winning re-election, 2.2% for Mitt.
If Romney tries hard tomorrow, he can probably get that down to 1%, which would be fitting.
Still gotta get out and vote, though.
If Romney tries hard tomorrow, he can probably get that down to 1%, which would be fitting.
Still gotta get out and vote, though.
11
If there's any systematic skew, it would be more likely in under-sampling those whose contact is the more technologically current. That is, as far as i know there are is no polling being done (or allowed?) via texting and the like. So which way would that skew the results? yep... toward those with landlines.
12
Though the prospect of a Republican win is certainly frightening, you may want to correct the "eek" to "eke".
13
@9: Yes, that why I chose the word 'postulate'. For example, scientists postulate the existence of water on Mars. That doesn't mean that they believe there is water on Mars.
14
GDfR, once again you only have misinformation. I have to believe you are actually understanding what you read, and purposely misrepresenting it when you make your comments here. Or do you just skip over the actual details? (Like Romney does)
16
@13 that's not how that word works. Nor is that how science works. Lordly, dude. Just admit when you're wrong. It'll make you look a lot less stupid.
17
@9 I'm beginning to dislike Nate Silver's work for the nytimes, it seems to consist 1 thoughtful post a week followed by a bunch of half-trolls
18
Just when you thought Gay Dude could not lie any harder...
I don't even know what to make of this one. #13 breaks new ground in bald-faced bullshittery
I don't even know what to make of this one. #13 breaks new ground in bald-faced bullshittery
19
@14: See @13, or let's put it in a medial context for you. Someone is experiencing shortness of breath and palpitations. The postulate that they might heart disease.
20
@18: I linked to an article without personal commentary -- do you really have a problem with that?
21
19, I work in health. We don't gloss over the facts and ignore details to figure out a patient's diagnosis. Only fraudulent hacks would do such a thing.
22
@ 11, yup, and also if polls would be intentionally skewed by "the mainstream media" it stands to reason that the race is actually even more a lot less close than being presented, close races sell newspapers (or to be more correct "generates publuc interest and thus ad income" as this is also about radio & tv.
Of course there can be also instances where polled voters can give incorrect data, but those are generally in cases like support for straight up extreme-right/fascist parties and referenda on civil rights, romney is a huge dick but not really yet something one would be embarrassed about to admit support for.
Long story short, the polls are more likely to be skewed in favor of romney than against...
Still, don't get lazy, go and vote..
Of course there can be also instances where polled voters can give incorrect data, but those are generally in cases like support for straight up extreme-right/fascist parties and referenda on civil rights, romney is a huge dick but not really yet something one would be embarrassed about to admit support for.
Long story short, the polls are more likely to be skewed in favor of romney than against...
Still, don't get lazy, go and vote..
23
@20: Except you lied about what the article said, and then essentially lied about what "postulate" means and how it applies here. You can not just say a writer postulates something that they did not even say.
Silver claiming a .6% likelihood of something happening is not him postulating that it will happen. But you know that already. You just lie.
It amazes me that you think you can point to the sky, tell us it is purple, and expect us to believe you.
Silver claiming a .6% likelihood of something happening is not him postulating that it will happen. But you know that already. You just lie.
It amazes me that you think you can point to the sky, tell us it is purple, and expect us to believe you.
24
@9 That's showbiz for you.
Seriously, Goldy, don't be dissing on Nate. He's one of the wonders of our political analysis landscape, a self-made statistician who grew up in baseball, doing awesome sports predictions. As his modeling techniques got more and more sophisticated, he realized he could probably apply them to other fields that had lots of statistical data, like political polling and demographics. Considering he came out of, you should pardon the expression, left field, and has bested some of the best seers in the business, give him some love.
As for him having to crank out something every day to amuse his fans and followers, it certainly keeps me amused anyway, and he does come up with the occasional gem. Last time around, he spent a lot more time showing us raw data and explaining in great detail how his model works. I actually enjoyed that more, but I suspect his current approach is more accessible to most people, and his model has gotten so complex that it would probably explode my head, anyway.
Seriously, Goldy, don't be dissing on Nate. He's one of the wonders of our political analysis landscape, a self-made statistician who grew up in baseball, doing awesome sports predictions. As his modeling techniques got more and more sophisticated, he realized he could probably apply them to other fields that had lots of statistical data, like political polling and demographics. Considering he came out of, you should pardon the expression, left field, and has bested some of the best seers in the business, give him some love.
As for him having to crank out something every day to amuse his fans and followers, it certainly keeps me amused anyway, and he does come up with the occasional gem. Last time around, he spent a lot more time showing us raw data and explaining in great detail how his model works. I actually enjoyed that more, but I suspect his current approach is more accessible to most people, and his model has gotten so complex that it would probably explode my head, anyway.
27
@GDfR, the issue at hand is your usage of the word postulate. It doesn't make any given the nature of that article. When the odds are so ridiculously small. That's not postulating, that's simply stating that COULD happen. A dog COULD be born with a birth defect that makes it meow, but no scientist would postulate that. No one in the science community would work on that experiment.
Basically, you're getting flak for a misleading use of a word and a nonsensical defense of it.
Basically, you're getting flak for a misleading use of a word and a nonsensical defense of it.
28
And, Goldy, Nate correctly predicted 49 states out of 50 in 2008 in the Presidential election, and all 35 Senate elections.
If your UW guy got lucky enough to miss the Electoral College totals by only 2 votes, that means he had to have missed at least 2 states. (No state has less than 3.)
If your UW guy got lucky enough to miss the Electoral College totals by only 2 votes, that means he had to have missed at least 2 states. (No state has less than 3.)
30
@9 Silver's work clearly produces a lot of tid-bits that are interesting primarily to stats geeks and political junkies... which is what this is. There's not a whole lot for him to report between the end of the conventions and the first debate, so he might as well post some interesting diversions for folks (like me) who are into that sort of thing.
Nate is one of the most consistently clear and useful political commenters. Plus he's not half-shabby at predicting the outcome of the election.
And GDfR, you are so full of shit you squish when you walk.
Nate is one of the most consistently clear and useful political commenters. Plus he's not half-shabby at predicting the outcome of the election.
And GDfR, you are so full of shit you squish when you walk.
31
@24 Yep I know "that's show business". I also enjoy reading Nate's statistical musings from time to time.
In an abstract could it happen and what would it look like if it did, kinda of way, Nate's could the electoral college tie article was interesting.
However only an idiot or someone deliberately trying to mislead would claim that someone else's musing was a postulation.
In an abstract could it happen and what would it look like if it did, kinda of way, Nate's could the electoral college tie article was interesting.
However only an idiot or someone deliberately trying to mislead would claim that someone else's musing was a postulation.
32
@28 Oops. I forgot about Maine and Nebraska not being winner-takes-all in the Electoral College. Still, anyone know how many states Holman got correct in '08?
33
It's not about the polls, it's about delegitimizing the election and therefore Obama. Obviously, if the liberal press have lied in the polling and skewed things to favor Obama, the ultimate election win that reflects those bogus poll results must itself also be bogus, right? I mean, a real impartial election wouldn't produce skewed partisan results, would it?
34
@6 The poll denialism on the right is simply laying the groundwork for theft of the election by black-box voting machine software.
FTFY. (At the risk of sounding like a less verbose Sgt. Doom.)
FTFY. (At the risk of sounding like a less verbose Sgt. Doom.)
35
It's like de ja vu all over again. The workmen at the base of the Tower of Babel arguing over whether to take the clockwise or the anti-clockwise staircase to put the capstone on the top. It ain't over till it's over. Polls do not mean much except to lull some into not voting and others into getting out the vote. Maybe by November the electorate will wake up and vote for none of the above.
36
De ja vu all over again. Workers at the base of the Tower of Babel arguing over whether to carry stones up the clockwise staircase or the counterclockwise staircase. Debating which hair to split from off the barber shop floor. Polls function to lull some people into not voting and others into getting out the vote. I hope the electorate wakes up by November and when it comes time to wank the national chain, it will be for 'none of the above'
37
@27: Postulate is quite malleable, as transitive verbs are, in its definition and use. I meant in the sense of entertaining an idea or supposition, rather than actually believing it. In that spectrum lies our disagreement. Inherent in postulate is a reservation of doubt.
38
@32:
Darryl got 49 out of 51. In addition to NE-02 (but not NE as a whole), he missed IN and MO. They had the same number of electoral votes (11), so in terms of the EV count the pair were a push. Overall, he predicted 364-174. The actual count was 365-173.
FWIW, in Darryl's most recent analysis on September 30, Obama won every single one of the 100,000 Monte Carlo simulations. The mean EV count was 344-194, the median was 347-191, and the mode was 348-190.
Darryl got 49 out of 51. In addition to NE-02 (but not NE as a whole), he missed IN and MO. They had the same number of electoral votes (11), so in terms of the EV count the pair were a push. Overall, he predicted 364-174. The actual count was 365-173.
FWIW, in Darryl's most recent analysis on September 30, Obama won every single one of the 100,000 Monte Carlo simulations. The mean EV count was 344-194, the median was 347-191, and the mode was 348-190.
39
@28 As N points out @38, I was wrong. Darryl missed by one electoral vote in 2008.
And I'm not dissing Nate Silver. But there's nothing magical about what he does.
And I'm not dissing Nate Silver. But there's nothing magical about what he does.
40
@37 words have distinct meaning and create context. I'm pretty sure you would agree that these three sentences convey very different contexts, yet more or less describe the same situation:
Silver postulates an electoral tie.
Silver supposes an electoral tie.
Silver entertains an electoral tie.
I hoep you see what I'm talking about. They all offer a degree of certainty. I know you seem to have a love for disregarding language. But language and its meaning is very important.
Silver postulates an electoral tie.
Silver supposes an electoral tie.
Silver entertains an electoral tie.
I hoep you see what I'm talking about. They all offer a degree of certainty. I know you seem to have a love for disregarding language. But language and its meaning is very important.
41
Hmmm.... Republicans in denial.... No one could have predicted the attack on 9/11. Saddam's WMD. Hurricane Katrina. Trickle-down economics. The best way to solve the drug problem is "Just Say No." America has the best health-care system in the world. Homosexuality is a "lifestyle choice." Evolution is a false belief system. Abstinence-only education reduces teen pregnancy & STDs. Obama is a Muslim. Obama was born in Kenya. Obama is a Socialist. Obama hates America. Obama is the worst president we've ever had. The Bush/Cheney administration failed because they weren't "conservative enough."
Seems to me Republicans have always been in denial. This is news?
Seems to me Republicans have always been in denial. This is news?
42
@5, 13, 37: You're using "postulate" incorrectly, as a few others have pointed out. To postulate something is to assert its truth as a given for the purpose of an argument, proof, or line of reasoning. Euclid postulated that given a line and a point not on that line, there is one and only one line parallel to the first line and passing through the point. He didn't give a proof of it, but rather submitted it as the basis for more elaborate arguments. That's how postulates work.
There aren't many scientists who would postulate the existence of water on Mars. That's something we'd decide through empiricism, not intuition.
There aren't many scientists who would postulate the existence of water on Mars. That's something we'd decide through empiricism, not intuition.
43
37, Maybe you don't understand what you read after all. You certainly don't understand the words you write.
44
@38 Thanks for the link! I can't get enough of this stuff. Love seeing numbers put to trends with intelligent algorithms.
@39 Well, I know it's not actually magic, but can't I enjoy it anyway?
@39 Well, I know it's not actually magic, but can't I enjoy it anyway?
Advertisement
Most Commented on Slog
-
GOP Congressman: We Have to Ban Abortion Because Fetuses Are Masturbating In There by Dan Savage
-
SL Letter of the Day: The Tell-Tale Handjob by Dan Savage
-
Foie Gras Tastes Like Cat Food by Callan Berry
-
Cap Hill Is Becoming a Little Like What? by Charles Mudede
-
SL Letter of the Day: Let Me Think... by Dan Savage
-
Pizza Update: Breaking the 5-Minute Barrier by Goldy
-
President Obama Says NSA Spying Is "Transparent" by Paul Constant
-
The Wednesday Morning News by Goldy
-
Jobs-Killing Socialism Drives Seattle's Unemployment Rate Down to 4.7 Percent by Goldy
-
Less Than a Quarter of Americans Have Confidence in Newspapers by Eli Sanders
Stranger Personals
The Stranger
RSSSeattle Art & Performance Quarterly
A complete guide to Seattle's current art, film, dance, and theater season.
Facebook
Twitter
Comments (44) RSS