The US Election Polling / Data Thread

[Street Pharmacist]

Originally Posted by MikeN

Right, and dead people can vote via a Medium ?

Lol. Do you believe all the things you read on fringe websites? There's is a term that should help you determine what to believe: credibility

[MikeN]

Originally Posted by GGG

I would argue that the media has been pro trump.

Here you have a candidate which has no political experience, is an open bigot, has admitted to sexually assaulting women, has an open child rape case going on, has two policies that are universally panned by Enconmists and yet they treat him like a credible candidate and blatantly lies every time he opens his mouth.

The fact that the media doesn't start every broadcast with Here are the reasons that you can't elect trump makes them biased against Hillary. Against a generic republican you might have a point but against Trump. Come on.

The rape case has been dropped, how is he a open bigot? and Clinton is not?

Clinton is a criminal who is evil beyond words, and everything she gets her hands on turns to ####.

I'm not for Trump in away but he is by far the lesser of two evils. Clinton is bought and paid for by wall street and the middle east. Wikileaks has proven this. Add that with her having multiple federal investigations against her, a vote for her is a vote for Tim in the oval office.

[opendoor]

Originally Posted by MikeN

The rape case has been dropped, how is he a open bigot? and Clinton is not?

Clinton is a criminal who is evil beyond words, and everything she gets her hands on turns to ####.

I'm not for Trump in away but he is by far the lesser of two evils. Clinton is bought and paid for by wall street and the middle east. Wikileaks has proven this. Add that with her having multiple federal investigations against her, a vote for her is a vote for Tim in the oval office.

What the hell does this have to do with polling data?

[afc wimbledon]

Originally Posted by opendoor

What the hell does this have to do with polling data?

Or rational thought or reality for that matter.

[photon]

Originally Posted by MikeN

The rape case has been dropped, how is he a open bigot? and Clinton is not?

Clinton is a criminal who is evil beyond words, and everything she gets her hands on turns to ####.

I'm not for Trump in away but he is by far the lesser of two evils. Clinton is bought and paid for by wall street and the middle east. Wikileaks has proven this. Add that with her having multiple federal investigations against her, a vote for her is a vote for Tim in the oval office.

If you want to discuss your almost completely incorrect characterization of the candidates please use the election thread, this is a thread about polling.

[GGG]

Sorry about responding to that one.

[afc wimbledon]

Originally Posted by MikeN

The rape case has been dropped, how is he a open bigot? and Clinton is not?

Clinton is a criminal who is evil beyond words, and everything she gets her hands on turns to ####.

I'm not for Trump in away but he is by far the lesser of two evils. Clinton is bought and paid for by wall street and the middle east. Wikileaks has proven this. Add that with her having multiple federal investigations against her, a vote for her is a vote for Tim in the oval office.

There are no federal investigations against Clinton, none.

[photon]

Trump basically seems to need a miracle in Nevada, things are very similar to 2012 and Obama won by almost 7 points. Republicans would need to win election day by double digits AND have an astronomical turnout. Or he needs a huge percentage of those Democratic votes to be for him and none of the Republican ones to be for her.

http://www.ktnv.com/news/ralston/the...ly-voting-blog

EDIT: They had to extend the hours of one polling station in a Mexican supermarket in NV because there were 1000 people in line. When it's all said and done I wonder if the story will be that women defeated Trump, or Latinos? Latino turnout seems to be way up all over.

EDIT2: Aaahaha the Internet sent pizza to all those 1000 people waiting in line.

[MissTeeks]

[Drak]

Latest polls out of Florida

[photon]

Dems pull ahead in early voting in Florida

Total Ballots cast: 5,731,761
Total Vote By Mail: 2,370,567 (45%)
Total Early Vote: 2,897.183 (55%)

Democrats: 2,268,663 (39.58%)
Republicans: 2,261,383 (39.45)
NPA: 1,201,715 (20.97%)

Total Margin: DEM +0.13%

http://steveschale.com/blog/2016/11/...-football.html

[Senator Clay Davis]

So HuffPost posted an article with the headline "Nate Silver Is Unskewing Polls — All Of Them — In Trump’s Direction". Nate decided to tweet storm it up in response. I won't post all the tweets but he seems pretty pissed.

https://twitter.com/NateSilver538

[photon]

Yeah I was following that when it was happening. HuffPo basically (intentionally or not) accused him of cooking the books (essentially lying about his model) though so I don't think his response was that out of line.

I read the HuffPo piece before I saw Nate's tweets start and I thought they fundamentally are missing or not undestanding a feature of his model. I've seen poll numbers adjusted up for Clinton in the past, the adjustments depend on the inputs, not on the whim of Nate. One can argue if the adjustments are valid or not, but their criticism wasn't that.

One reporter I like said on her twitter: According to my model both nerds are losing by virtue of the fact that they're arguing about math on a Saturday.

[GGG]

That Huff post article is pretty terrible.

Essentially they are saying that if Nate as a state at 50.1% and the 49.9% wins then a model suggesting the 49.9% chance had an 80% chance of winning is right. When the model is predicting something will happen 50.1% of the time it should be wrong 49.9% of the time.

It's also funny their argument that Trump can't win because he needs a polling error of 3 points in his favour which happened in 2 of the last 3 elections but because it needs to happen in trumps direction it's very unlikely. And also they are complaining about his trend line adjustment and unskewing based on historical sampling because the Romeny guys used it poorly. When it was that unskewing and trend line adjustment used in 3012 that lead him to predict Obama would out perform conventional wisdom.

If they wanted to write that article they should have had their pollster do it instead someone who doesn't understand what is happening. That article is clearly written on a hedge that if their model out performs Silvers they can write an I told you so article. And Silver is just biting on the troll hate.

It's also funny that they say the upshot model at 85 is reasonable and their 98 is reasonable but Silvers 65 is not. There is much more sound reasoning for 85 and 65 being more similar then 85 and 98.

[Acey]

With NV gone, it seems like the most likely path for Trump would be winning NH, NC, then one of PA and MI. Then it'd come down to FL which he'd also have to win.

[CorsiHockeyLeague]

Originally Posted by GGG

Essentially they are saying that if Nate as a state at 50.1% and the 49.9% wins then a model suggesting the 49.9% chance had an 80% chance of winning is right. When the model is predicting something will happen 50.1% of the time it should be wrong 49.9% of the time.

If only people could apply this astounding level of logical reasoning to hockey stats... alright, wrong thread, I won't gripe.

But of course you're right. And so is Nate Silver. I don't know if his model is flawed; I'm not competent to determine that. But that HuffPost article is not only dumb and wrong, it doesn't care that it's dumb and wrong. It's ironically very Trump-like. And it's essentially written for people either too stupid to understand basic statistics concepts, or for people who just want to read their own version of reality. Unfortunately, there's a fairly big audience available in both categories, so I've little doubt they'll keep publishing stuff like it.

Tuesday needs to come and go already, I'm really starting to hate people.

[octothorp]

Originally Posted by GGG

That Huff post article is pretty terrible.

Essentially they are saying that if Nate as a state at 50.1% and the 49.9% wins then a model suggesting the 49.9% chance had an 80% chance of winning is right. When the model is predicting something will happen 50.1% of the time it should be wrong 49.9% of the time.

I don't think 538 is above criticism, but this article is an absolutely awful attempt to do so. I see two main differences between 538 and the other forecasters. One is, as this article says, is trend-lines. But the bigger difference, which this article only vaguely touches on, is uncertainty. As I believe 538 explained it early in the campaign, their model uses a longer historical sample of elections for its understanding of how polls can move. While Huffpost and Upshot's models are based on 1980s onward, while 538's go back into the 60s and 70s, when there were more polling misses and more volatility. So this is definitely a judgement call that forecasters need to make.

But there is reason to think 538 made the wrong call on that; with the probability calls they've making, you'd expect them to be wrong far more than they've actually been wrong. It's entirely possible that while their model is great in predicting results, by including a more volatile era of electoral history their model is essentially sandbagging it a bit. I'm not saying for sure it does, but that's what I think a legitimate criticism of 538's model might focus on.

edit: I should add, I think 538 has a solid defense to this, which is that most elections, the polls will be pretty accurate; but that when they miss, they'll miss in a lot of states. So you wouldn't necessarily expect to see them miss a couple every election; but you would expect that every now and then, there's an election where the polls missed or the race swung late in a lot of states at once.

[Acey]

The 538 model a) does not do as well with all these more volatile swings in polls and b) does not and will not account for the massive amount of votes already cast when Clinton's numbers were higher, including the fact that NV is almost surely blue now (a scenario in which Trump's overall chances drop to 9%)

[GGG]

Originally Posted by Acey

The 538 model a) does not do as well with all these more volatile swings in polls and b) does not and will not account for the massive amount of votes already cast when Clinton's numbers were higher, including the fact that NV is almost surely blue now (a scenario in which Trump's overall chances drop to 9%)

There is a great article on the Nevada vote on 538 right now.

if the polling error that is reflected in the early voting data is local to Nevada then it only affects 2.6% of cases. If the polling error in Nevada is a sign of Clinton out performing her polls generally then the affect is significant as you have stated above.

This also assumes that people are voting along party lines which some polling shows may not hold as the white working class is going more to Trump than previously though this is much more of an issue in the Midwest then in Nevada where the immigrant vote governs.

[GGG]

Silvers reply to Huff Post

http://fivethirtyeight.com/features/...-a-close-race/