The US Election Polling / Data Thread

[GGG]

Originally Posted by Parallex

Well of course... any statement that proclaims there is a 0% chance of something is wrong on the face of it. But I don't think the likelihood is 50-50 either. Clinton easily just had the worst two weeks of her campaign and Trump didn't draw even. Sure she could have more bad weeks... but I don't think that's any more likely then Trump having more bad weeks.

Her campaign still has...

A: More money
B: A better ground game
C: Better (albeit of the less bad variety) favorable/unfavorable rating



And if there isn't any then she wins, if it's polling bias the other way she wins by an Obama 2012 margin. It's silly to fret about the maybe's of a polling bias because you don't know that any exist, that if it exists it's against Trump, and in all likelihood (in the aggregate) it doesn't. I'm as terrified about the prospect of Donald Trump having codes to a nuclear arsenal but now is not the time to freak out. If Trump is ahead in enough states to get 270 electoral college votes in a months time... that's the time to start building your bunker.

Id agree it's not 50/50 and think it's higher than the current model suggests. I'd buy Hillary stock at the current price of 63%. However I'd likely not go over the 75% point.

I do think polling inaccuracy has a greater probability here then in previous elections because of the 3rd party vote. It increases the variability of the results beyond the statistical +/- 3% that the sampling shows. Ground game being worth about 1-2% offsets some of this risk. But anyone saying no path is greatly under estimating trumps chances.

Up to what's likelihood would you bet Hillary?

I think there is an opportunity with the betting sites here as they appear to track the 538 forecast so public money is effectively betting on Nate's model so is there room for sharp money to do better based on things his model doesn't consider?

[Parallex]

Originally Posted by GGG

so is there room for sharp money to do better based on things his model doesn't consider?

Nate's models (remember he has three, PollsPlus/PollsOnly/NowCast) has the most inputs of any of the prediction models so I'd say no.

[Parallex]

Originally Posted by GGG

Id agree it's not 50/50 and think it's higher than the current model suggests. I'd buy Hillary stock at the current price of 63%. However I'd likely not go over the 75% point.

That seems about right. I mean if I were a gambling man and someone were to approach me with a straight up bet I wouldn't consider putting money on Trump unless I got around 4/1 odds or thereabouts (I still wouldn't because... bad kharma) but dispassionately speaking that's where I'd put money down if I were so inclined.

NY Times has a site that shows all the prediction models (alongside there own) if you're interested...

http://www.nytimes.com/interactive/2...forecast.html?

[rubecube]

Wasn't Nate's model pretty much bang on in 2012?

[GGG]

Originally Posted by Parallex

Nate's models (remember he has three, PollsPlus/PollsOnly/NowCast) has the most inputs of any of the prediction models so I'd say no.

His model doesn't consider ground game or I believe the trend of third parties voters to move back toward one of the two candidates

[Parallex]

Originally Posted by rubecube

Wasn't Nate's model pretty much bang on in 2012?

No, it was just bang on (not "pretty much" it called the electoral college perfectly). Missed on some of the senate races thou (which is my way to segway into mentioning that their senate forecast is up... http://projects.fivethirtyeight.com/...recast/senate/)

[Flash Walken]

As the race reaches voting day, mis-steps become more critical and magnified. Part of the massive mis-step that was Palin was being overwhelmed by her gaffs right as election day approached. Had she been introduced earlier, say before the convention as a potential VP candidate, a lot of her terribleness would've been exposed earlier and probably (in my opinion of course) not have been so magnified as it was.

Because we know that Trump is basically immune to gaffs and mis-steps, Clinton will have to play out a perfect string to clinch the election. Any serious mis-step will be magnified by the US media because of how rare it will be.

I am resigned to a Donald Trump presidency. Much like Biden Vs. Palin, Palin picked up support from that debate just by not being as bad as people expected. Trump has nowhere to go but up and that is terrifying.

[Itse]

Originally Posted by Flash Walken

Trump has nowhere to go but up and that is terrifying.

I'm not so sure. I would suspect most people have only read about Trump on Facebook or on TV (if they're old folk), and so they've just seen some very short clips. That means a lot of people will really only see the Trump show themselves for the first time during the debates, and I suspect a lot of them will simply be disappointed. Political debates are generally damned boring if you're not into politics (and often even if you are), so people tuning in expecting to be entertained will most likely just tune out in fifteen minutes or less. And the one they will be disappointed in will be Trump, because he was supposed to be the interesting one.

I also don't think Trumps charisma reallys last in big doses. He's kind of a one-trick pony as a speaker, and in my opinion tiresome to listen to for any length of time. I've tried to watch his RNC speech a couple of times, and I never make it to the end because i's mostly just really boring. He's kind of a slow talker and repeats himself too much. Even his voice gets kind of annoying.

A lot of people here have been saying that Trump essentially just needs to not make a fool of himself, but I think that opinion is mostly shared by people who already think he is a fool (or worse), and those people are not terribly likely to vote for him anyway.

On the other hand, nobody really excepts Clinton to be anything more than a competent politician, and that's much easier to deliver. She basically needs to be herself and not faint on stage. If she can throw in a good zinger that gets lots of Facebook coverage, she'll probably be just fine.

I think Trump is also potentially a good opponent for Clinton. Acting like a bully is a big reason why Trump has that aura of power, but doing it to Clinton could easily lose her too many of those female votes he desperately needs.

Clinton on the other hand has a problem of being (or at least seeming) condescending. Seeming condescending towards a smart and respectful guy like Obama looked bad. Being condesdencing towards a loud bully would look strong. What's worse for Trump, if you try to bully someone and they're just unimpressed, you'll seem kind of pathetic.

Clinton also only needs to split the rest of the undecideds to win the election, and that isn't super hard. That's why the gap narrowing isn't too worrisome for Clinton.

[Gozer]

Originally Posted by Itse

if you try to bully someone and they're just unimpressed, you'll seem kind of pathetic.

Like the Gore-Bush chest-bump.

[GGG]

Originally Posted by Parallex

No, it was just bang on (not "pretty much" it called the electoral college perfectly). Missed on some of the senate races thou (which is my way to segway into mentioning that their senate forecast is up... http://projects.fivethirtyeight.com/...recast/senate/)

He was within 2% on his state by state results slightly under representing Obamas votes in swing states.

A 2% swing is essentially the difference between Hillary winning easily and trump taking New Hampshire. I think in general (and I'm guilty of it too) we put to much faith in the oracle of elections.

When some of it like calling Florida correctly when your model says coin flip does come down to luck.

[afc wimbledon]

Originally Posted by Flash Walken

As the race reaches voting day, mis-steps become more critical and magnified. Part of the massive mis-step that was Palin was being overwhelmed by her gaffs right as election day approached. Had she been introduced earlier, say before the convention as a potential VP candidate, a lot of her terribleness would've been exposed earlier and probably (in my opinion of course) not have been so magnified as it was.

Because we know that Trump is basically immune to gaffs and mis-steps, Clinton will have to play out a perfect string to clinch the election. Any serious mis-step will be magnified by the US media because of how rare it will be.

I am resigned to a Donald Trump presidency. Much like Biden Vs. Palin, Palin picked up support from that debate just by not being as bad as people expected. Trump has nowhere to go but up and that is terrifying.

I'm not sure the Trump is immune part is true though, I tend to take the reverse view from Trumps campiagn, I suspect people are saying they will vote for Trump as he is the 'up yours' candidate but I don't think they will actually vote for him when push comes to shove, I suspect he polls way higher than the actual numbers of voters that will turn out for him.

[Senator Clay Davis]

New four way national poll shows the Johnson/Stein support is predictably thinning out as we get closer to election day. The bad news for Hillary is it doesn't appear to help her like other polls suggested.

Clinton 49, Trump 47 two way; Clinton 46, Trump 44, Johnson 5, Stein 1 four way (ABC/WaPo)

Also the closest Pennsylvania poll we've seen in a while

Clinton 40, Trump 38, Johnson 8, Stein 3 (Morning Call)

So Hillary really needs a good performance tomorrow night. Any Trump "win" and he probably takes the lead in polling.

[Street Pharmacist]

Originally Posted by Senator Clay Davis

New four way national poll shows the Johnson/Stein support is predictably thinning out as we get closer to election day. The bad news for Hillary is it doesn't appear to help her like other polls suggested.

Clinton 49, Trump 47 two way; Clinton 46, Trump 44, Johnson 5, Stein 1 four way (ABC/WaPo)

Also the closest Pennsylvania poll we've seen in a while

Clinton 40, Trump 38, Johnson 8, Stein 3 (Morning Call)

So Hillary really needs a good performance tomorrow night. Any Trump "win" and he probably takes the lead in polling.

I disagree about a debate "win" as this isn't really a competition. Trump just has to not look foolish or bad tempered and he'll "win" because that's all that's really holding him back

[Itse]

Originally Posted by Street Pharmacist

I disagree about a debate "win" as this isn't really a competition. Trump just has to not look foolish or bad tempered and he'll "win" because that's all that's really holding him back

I don't agree. What's holding him back is that people hate and despise him, and if enough confirmation bias sets in about that, it really won't matter what he does any more. (In fact, he might already be in too deep in that regard.)

At the very least he needs to be less unlikeable than Clinton, and by a good margin too. All Clinton needs is ~50% of the undecided voters to win. Being generic and bland is usually enough for that.

Candidates don't come from behind to win very often.

[GGG]

Originally Posted by Street Pharmacist

I disagree about a debate "win" as this isn't really a competition. Trump just has to not look foolish or bad tempered and he'll "win" because that's all that's really holding him back

A trump " not pooping himself"win"" will give him a bump in the polls. Though debate bumps often are fleeting. And I would say especially in this case as both candidates are known.

[Senator Clay Davis]

Well after some decent polling news last week for Hillary, this week starts rather poorly. Colorado is definitely in play and Pennsylvania is showing cracks now too. If Trump scores a big debate win tonight Hillary is in major trouble.

CO: Trump 41, Clinton 37, Johnson 6, Stein 6 (Gravis); Clinton 40, Trump 39, Johnson 7, Stein 2 (CBS/YouGov); Trump 42, Clinton 41, Johnson 13, Stein 3 (CNN/ORC)

PA: Clinton 45, Trump 44, Johnson 6, Stein 3 (CNN/ORC)

National: Trump 43, Clinton 41, Johnson 8, Stein 4 (Bloomberg); Clinton 44, Trump 43, Johnson 8, Stein 2 (Q)

[Senator Clay Davis]

Keep pollin, pollin, pollin'...but yeah lots of mixed messages coming from these polls. I guess that's the sign this is indeed a very tight race. And that we're seeing some unusual/lesser known pollsters too

National: Clinton 46, Trump 42, Johnson 8, Stein 2 (Monmouth)
Florida: Clinton 43, Trump 41 (Florida Chamber of Commerce....ok then)
Minnesota: Clinton 46, Trump 39, Johnson 6, Stein 2 (KSTP/SurveyUSA)
Iowa: Trump 38, Clinton 38, Johnson 9, Stein 1 (Loras)
Virginia Clinton 39, Trump 33, Johnson 15, Stein 3 (Christopher Newport University)
North Carolina: Clinton 43, Trump 42, Johnson 10 (High Point University)

[mrkajz44]

The race seems very close indeed. I always thought the debates will be the deciding factor, and the first one is finally here. Seems the expectations are much lower for Trump than Clinton, but it'll be interesting to see how the one-on-one goes. The performances in the debate could go a long way to shifting the undecideds at this point.

[nik-]

I think a lot of Trump's polled support is just poll support. I have a feeling a significant portion of independents that poll Trump will have a "I just can't do it" moment at the actual ballot box. Pragmatism will win lots of these people in the end.

[Flash Walken]

Originally Posted by nik-

I think a lot of Trump's polled support is just poll support. I have a feeling a significant portion of independents that poll Trump will have a "I just can't do it" moment at the actual ballot box. Pragmatism will win lots of these people in the end.

I think a lot of republicans will end up staying home altogether.