-
Hey, guest user. Hope you're enjoying NeoGAF! Have you considered registering for an account? Come join us and add your take to the daily discourse.
PoliGAF 2016 |OT10| Jill Stein Inflatable Love Doll
- Thread starter Aaron Strife
- Start date
- Status
Part I caught
Terrible to poll over a holiday weekend because higher educated are more likely to be on vacation.
nature boy
Member
Assuming way too many white non-college voters, in a scale never seen before (50%, in 2012 around 1/3).
Shouldn't have done it during Labor Day, rich folks out of town.
If adjusted to 2012 demographics, Clinton up +4 (46-42).
Trump in big trouble with white college voters, both female and male. Romney won that group bigly and still lost.
The white electorate has also been shrinking little by little with each presidential election. Very interesting! Thanks for sharing!
whyamihere
Banned
I'm less okay with adjusting the poll based on demo to that polling over Labor Day weekend is a bad idea!
Plinko
Wildcard berths that can't beat teams without a winning record should have homefield advantage
Yes...I'd call it retirement.
This...is actually a really good point. Demographics were all over the place, though.
I just hate the media's general self righteousness. Jake Tapper started his Zika segment with a snarky bit about how while Congress was on vacation, x number of women were infected by Zika and emphasized how lazy Congress is. Meanwhile, Jake Tapper was on a three week vacation just like the rest of Congress, because that's what a lot of people do in August.
I'm no fan of Congress, but the media likes to pretend that their shit does stink when in reality, it's some of the most foul smelling of the bunch.
Black Mamba
Member
Chuck Todd/NBC now unskewing polls?
Don't like it.
The CNN poll is most certainly wrong, but it's not worth unskewing. If Hillary is up 5, then you're going to get a random Trump +2 in there.
The media's treatment of polling is so fucking bad.
Don't like it.
The CNN poll is most certainly wrong, but it's not worth unskewing. If Hillary is up 5, then you're going to get a random Trump +2 in there.
The media's treatment of polling is so fucking bad.
bangbang i'm a timpani
Banned
probably a mix of "its only one poll", "trump has no ground game", "trump/clinton floor/ceiling"
Hillary just can't win. Lady on Chuck Todd saying that she handled bringing up Pam Bondi horribly and she had to be careful about what she said because she has no moral authority on the subject and the stories of her own scandals are so bad. Then Chuck chimes in with even though there's no evidence of wrongdoing on her part she has to be careful because... optics. Ugh.
Dubbedinenglish
Banned
Like I'd buy 30 secs before the SOTU every year on Fox just to eat a sandwich on air kind of money. Hopefully the rest get paid too.
balladofwindfishes
Member
As someone who chronically coughs for no real medical reason, just as a personal tic, this whole coughing saga is pretty awful.
Actually, it should be awful for anyone, whether they cough a lot or not.
Actually, it should be awful for anyone, whether they cough a lot or not.
http://www.nytimes.com/2016/06/10/u...n-people-think-thats-good-news-for-trump.html
nyt said:
Note that the article goes on to say that Trump can't lose nonwhites and college-educated whites as badly as he seems to be and win the election. But exit polls simply aren't comparable to pre-election polls, and they are not that reliable. For obvious reasons! Their methodology is extremely messy!
whyamihere
Banned
Yeah. I think you're seeing a combination of: tightening, which would lead to some polls with Trump up, a Labor Day sample, or just random outlierness. Or a combination of all 3! But let's not, you know, reset a poll with 2012 exit poll demos.
Black Mamba
Member
But while I find Cohn's argument about there being more white voters being true, I don't believe it's in Trump's favor. I think there's more white voters that vote Democrat that are being missed in exit polling (Cohn even hints at this with Obama's numbers).
When Cohn wrote this article, I think his evidence was good but conclusion was bad. I don't think this makes Trump more likely to become President. To the contrary, I think less likely. His argument relies on Trump making reliable Democrat voters switch and I don't see it.
Remember, the Trump/GOP argument is that there are more white non-voters out there, not that they can switch current voters.
newshinycd
Member
No, unskewing is using feels to change the poll result into something that confirms one's preconceptions.
Also:
Ultimately, this is what the taco truck election is about. Notice how much better America is than all European countries.
http://www.theatlantic.com/politics/archive/2016/09/taco-trucks-and-the-soul-of-america/498845/
Ultimately, this is what the taco truck election is about. Notice how much better America is than all European countries.
http://www.theatlantic.com/politics/archive/2016/09/taco-trucks-and-the-soul-of-america/498845/
Black Mamba
Member
The whole adjusting for past electorates is dumb.
I don't need a poll to know who will win if the electorate is the same as 2012. Spoiler: It's Hillary.
Pollsters have to guess the electorate and we have to live with that because we don't know what the electorate will be (though, we can laugh at really bad ones or disagree with their model).
At the end of the day, the election comes down to this. Will enough of the Obama coalition show up to vote? if yes, Hillary wins. End of story. People who voted Obama twice are not voting Trump.
There are no missing white voters. If the Obama coalition turns out, she wins. If it doesn't, she can lose. If it turns out AND she gets a bigger bump from minorities in turnout alongside a lack of enthusiasm for Trump, it will be a slaughter.
Also:
Ultimately, this is what the taco truck election is about. Notice how much better America is than all European countries.
http://www.theatlantic.com/politics/archive/2016/09/taco-trucks-and-the-soul-of-america/498845/
Where is Canada
The original Unskewed Polls website was literally what Chuck Todd did -- they rebalanced a poll by changing the demographic weights according to an opinion about what the electorate would really look like.
As Mamba notes, it's the same thing!
ItWasMeantToBe19
Banned
Also:
Ultimately, this is what the taco truck election is about. Notice how much better America is than all European countries.
http://www.theatlantic.com/politics/archive/2016/09/taco-trucks-and-the-soul-of-america/498845/
smh, no Austria on there. Now we don't get a country where "The 1%" refers to the percentage of people that believe minorities aren't rapists and murderers.
Also:
Ultimately, this is what the taco truck election is about. Notice how much better America is than all European countries.
http://www.theatlantic.com/politics/archive/2016/09/taco-trucks-and-the-soul-of-america/498845/
I'm super proud of the 56% in that poll. Good job team USA!
whyamihere
Banned
Exactly.
Box of Kittens
Banned
There are a few examples of rural areas that vote liberal. Vermont is mostly rural yet one of the most liberal states in the nation. Likewise, Western Massachusetts is largely rural yet arguably the most progressive part of the state. Some parts of the Upper Midwest also fit the rural/progressive pattern.
I said this earlier. Clinton has every right to bring up the Bondi thing because the media pretty much refuses to. But I figured the media will try to turn this back on her and the emails just because.
NYCmetsfan
Banned
http://www.nationallawjournal.com/home/id=1202766785916?slreturn=20160806175231
Obama nominates the first Muslim judge to district court.
Don't know if the right is gonna go with "Sharia LAW!!!!" or going with a curiel attack
Obama nominates the first Muslim judge to district court.
Don't know if the right is gonna go with "Sharia LAW!!!!" or going with a curiel attack
Suikoguy
I whinny my fervor lowly, for his length is not as great as those of the Hylian war stallions
Also:
Ultimately, this is what the taco truck election is about. Notice how much better America is than all European countries.
http://www.theatlantic.com/politics/archive/2016/09/taco-trucks-and-the-soul-of-america/498845/
That's way better than I thought, I can only fault the 7%, and that's damn low.
Black Mamba
Member
IIRC, the unskewer did it by party ID. Which is even worse because Party ID has no merit in adjusting a poll.
All pollsters weight their responses on what they assume the electorate will look like in terms or age, sex, race/ethnicity, and geography. But not on party ID.
What NBC did is basically say "If I was a pollster and I got this data set, here are my weights," which is unskewing but it's at least within reason. The party ID thing was stupid.
I still think it is bad to unskew the numbers like how NBC did. You can do that with every poll. Just take their data sets and put in your own weights. I guess you could be a pollster without doing work that way. But to pick out 1 or 2 polls and do this is bad.
The reason pollsters use weights is because
A. their samples are not very strong or something even statistically significant among sub-groups
B. some groups are really hard to poll
But if a pollster would do say a 50k person poll, they wouldn't have to weight things at all. Just post the results.
Which is why I just prefer to allow the pollsters to weight it themselves, then you aggregate it all. Don't change the numbers, ever.
whyamihere
Banned
I truly don't know what to do about this. And his unskewing seems awful, too. Ugh. Whatever, Chuck.
newshinycd
Member
I don't have CNN (or cable), but the posts here made it sound like he was discussing the demographics of the poll, not actually unskewing.
NYCmetsfan
Banned
LOL
This I will agree with.
Here in Philly over the weekend I saw the "Hillary: this is my jobs plan" ad exactly once. But the "*Night shot of White House* Trump is dangerous!" ad runs pretty much nonstop.
Hillary needs to be more forceful in selling herself, because the media has made it clear they aren't going to do it for her.
Cable news is dead, please everybody stop watching it forever
ViewtifulJC
Banned
So Trump's gonna win right?
What the... Come on, nobody is a fan of that CNN poll, but to begin unskewing a rival's poll because your poll isn't grabbing the same amount of headlines, is kind of... unprofessional.
And pathetic.
Although, come on, we know he's right 
He's already President and we're already dead.
Spoiled Milk
Banned
nature boy
Member
So PoliGAF can unskew but NBC with some good pollsters can't?
GTFO.
GTFO.
VendettaRed07
Member
She is still leading by like 8+ points in states that add up to 270
Suikoguy
I whinny my fervor lowly, for his length is not as great as those of the Hylian war stallions
I can't tell if this is serious or not.
It's unskewing! He literally reweighted the poll according to a different demographic breakdown because he didn't agree with their voter model! And he did so based on exit polls, when as discussed, exit polls simply aren't comparable with pre-election polls!
nature boy
Member
I'm serious, while you can debate putting up that graphic, if the CNN poll has demographic issues (you had a REPUBLICAN pollster on MTP Daily questioning the demographics) why shouldn't it pointed even if it's a rival network?
- Status
Similar threads
- Locked
- Locked
- Locked
- Locked