Hillary's numbers are tanking

[bilby]

I keep seeing analysis's showing Clinton leading among women by about 15% while Trump leads among men by about 5% or so. Since women voters out number men voters, Clinton should be far ahead. But she is barely ahead in most polls. Something here is out of whack. I keep seeing various news articles about women who plan to vote Clinton, but are keeping that to themselves, which are true or not true. If so, Trump loses. What the hell is going on with all of this? We will have to wait til the week after election day to see how this resolves itself. But it may well be that Trump will have caused the pollsters to rewrite the conventional wisdom about polling. Only now is early voting getting analysis, but numbers involved are still too small to be predictive. There is a lot of wishful thinking and cherry-picking polls going on all around. I hope this pro HRC silent majority of women voters phenomena is true, but it may well be a fantasy.

Pollsters and those who report their polls all have a vested interest in a close race. It's all about excitement and entertainment - Nobody is going to attract viewers, listeners, readers or even voters by saying 'It's all over bar the shouting; Candidate A is almost impossible to beat at this stage".

Cherry-picking, both of polls to report upon, and of citizens to include in the guesstimates of 'likely voters' is inevitable. If that is happening, the result would be an apparent tightening of the race as the election draws nearer. And that's what we are seeing.

There is simply no way to separate the hypotheses that the polls are wrong in this way, or that the race really is tightening - other than by looking at the final result, when the votes are counted.

 

[Axulus]

I keep seeing analysis's showing Clinton leading among women by about 15% while Trump leads among men by about 5% or so. Since women voters out number men voters, Clinton should be far ahead. But she is barely ahead in most polls. Something here is out of whack. I keep seeing various news articles about women who plan to vote Clinton, but are keeping that to themselves, which are true or not true. If so, Trump loses. What the hell is going on with all of this? We will have to wait til the week after election day to see how this resolves itself. But it may well be that Trump will have caused the pollsters to rewrite the conventional wisdom about polling. Only now is early voting getting analysis, but numbers involved are still too small to be predictive. There is a lot of wishful thinking and cherry-picking polls going on all around. I hope this pro HRC silent majority of women voters phenomena is true, but it may well be a fantasy.

Pollsters and those who report their polls all have a vested interest in a close race. It's all about excitement and entertainment - Nobody is going to attract viewers, listeners, readers or even voters by saying 'It's all over bar the shouting; Candidate A is almost impossible to beat at this stage".

Cherry-picking, both of polls to report upon, and of citizens to include in the guesstimates of 'likely voters' is inevitable.

Except just two weeks ago, there were many articles that Hillary had it in the bag. I thought she pretty much had it in the bag. There was even speculation that she had a shot at Texas, a state that has not gone Democrat for president since 1972. Not anymore.

The new polls definitely seem to be pointing to diminished chances of Hillary winning.

 

[bilby]

Pollsters and those who report their polls all have a vested interest in a close race. It's all about excitement and entertainment - Nobody is going to attract viewers, listeners, readers or even voters by saying 'It's all over bar the shouting; Candidate A is almost impossible to beat at this stage".

Cherry-picking, both of polls to report upon, and of citizens to include in the guesstimates of 'likely voters' is inevitable.

Except just two weeks ago, there were many articles that Hillary had it in the bag. I thought she pretty much had it in the bag. There was even speculation that she had a shot at Texas, a state that has not gone Democrat for president since 1972. Not anymore.

The new polls definitely seem to be pointing to diminished chances of Hillary winning.

Sure. But my point is that this is consistent with Hillary's chances being diminished; but it is ALSO consistent with the media seeking a more exciting narrative than the one presented by reality. Two weeks ago, the audience was 'people who really care about politics'. Today, the audience is the much larger demographic 'people who only care about politics when there's a close fought election'. Competition for this larger audience may be sufficient to explain some or even all of the difference.

 

[Jimmy Higgins]

In 2012, 2 days before the election CNN, Politico and Monmouth all had it even, Rasmussen and Gallup had Romney +1, ABC and Pew had Obama up 3 and NBC had Obama +1. The election ended with Obama +4.

The way I see this year is that if everyone that voted for Obama in 2012 votes for Hillary she will win. If a lot of people stay home, Trump could pull it off.

yes, but there was one poll that was predicting a 4+ Obama victory in 2012 and that was the LA Times/USC poll. That methodology is still being used by them and has consistently shown a Trump lead. Maybe though this year that method really won't work and 2012 was a fluke.

It's just too close to call. It will depend on turnout in key battleground states - which some signs indicate are favoring the democrats. Fortunately I've already voted. Of course living in the frickin' reddest state in the country (i.e. the one that elected Republicans head of all three branches of government only to have everyone of them frickin' impeached or about to be) my vote is pretty much useless.

SLD

The LA Times has a single young black Trump supporter that is terribly skewing their poll.

 

[Cheerful Charlie]

The idea that this is all about the media wanting a close race seems to me to be misguided. The pollsters do the polls, the media reports. Sure you can get cherry picking such as Faux. But the pollsters have a vested interest in getting things right. Screw up enough and you can hurt your reputation. Zogby for instance has stopped this sort of polling.

And then we have ground game. Possibly polls may be misleading if people don't get out to vote. In 2012, Obama's superior ground game allowed him to beat the aggregate polls by 3% Possibly Trump could lose out here.

And we have the discouraged partisan effect. Some people do not respond to polls after their candidate is featured unfavorably in the latest news stories. Skewing the polls. So the new e-mail "scandal" may lower HRC's poll numbers but no necessarily her vote numbers.

So at this point there are too many theorticals to say what is really going on and what will happen November 8. All one can do is hope.

There is a lot of partisan hoping going on now.

 

[Will Wiley]

https://www.google.com.au/trends/explore?date=today 1-m&q=change early vote

Try putting "change early vote" into google trends

 

[Cheerful Charlie]

Apparently Nate Silver is throwing up his hands and admitting that the polls alone aren't truly predictive. This election is a whole new ball game. We will see on November 9th who were the pollster eagles and who were the pollster goats.

 

[Warpoet]

yes, but there was one poll that was predicting a 4+ Obama victory in 2012 and that was the LA Times/USC poll. That methodology is still being used by them and has consistently shown a Trump lead. Maybe though this year that method really won't work and 2012 was a fluke.

It's just too close to call. It will depend on turnout in key battleground states - which some signs indicate are favoring the democrats. Fortunately I've already voted. Of course living in the frickin' reddest state in the country (i.e. the one that elected Republicans head of all three branches of government only to have everyone of them frickin' impeached or about to be) my vote is pretty much useless.

SLD

The LA Times has a single young black Trump supporter that is terribly skewing their poll.

I thought you were taking the piss, but I did some quick googling just to be sure, and holy shit, you're right. I'm amazed this hasn't been more widely reported as it really discredits the poll; I don't see how any poll can keep re-contacting the same people and be taken seriously. But then, one need only remind themselves that Donald Trump is one of the two candidates for the presidency to realize that normal rules of logic don't apply at the moment.

 

[Tom Sawyer]

Apparently Nate Silver is throwing up his hands and admitting that the polls alone aren't truly predictive. This election is a whole new ball game. We will see on November 9th who were the pollster eagles and who were the pollster goats.

Well, there's no need for him to get into such a huff about it. Why doesn't he just call up Putin and ask what the final vote results are going to be?

 

[Horatio Parker]

Pollsters can influence polls by the assumptions or "weighting" made. Or the samples they choose. Eliminate cell phones, and bye bye to everyone under 30.

I follow Sam Wang, because it's his hobby and he has no commercial desire to increase traffic to his site. His algorithms are also transparent, and the statistics geeks often come out in comments. Over my head, but whatever.

One point he makes is that the more polls done, the more results.

Another that I've seen recently is that people respond less to polling when their candidate is not doing well, and more when they are. This can create the appearance of movement.

We'll see how it shakes out, but HRC has been ahead by 3-5% for a year. Wang says it's the most stable race in 50 years. Kinda hard to reconcile that with the media craziness.

 

[SLD]

The idea that this is all about the media wanting a close race seems to me to be misguided. The pollsters do the polls, the media reports. Sure you can get cherry picking such as Faux. But the pollsters have a vested interest in getting things right. Screw up enough and you can hurt your reputation. Zogby for instance has stopped this sort of polling.

And then we have ground game. Possibly polls may be misleading if people don't get out to vote. In 2012, Obama's superior ground game allowed him to beat the aggregate polls by 3% Possibly Trump could lose out here.

And we have the discouraged partisan effect. Some people do not respond to polls after their candidate is featured unfavorably in the latest news stories. Skewing the polls. So the new e-mail "scandal" may lower HRC's poll numbers but no necessarily her vote numbers.

So at this point there are too many theorticals to say what is really going on and what will happen November 8. All one can do is hope.

There is a lot of partisan hoping going on now.

One of the problems with polling is that it may actually be the cause of shifts. And that makes it difficult to predict ahead. When polls show the race tightening, people suddenly might be interested in voting and realize their vote does count. Unmotivated registered voters in North Carolina right now must be shaking their heads and saying, uh-oh, maybe I ought to get out there and vote.

I certainly hope you are right about the Democratic ground game. They are more organized than Trump and hopefully that will make a difference. But I wonder to what extent the pollsters already have put that into their models? Does anyone know? Hopefully not!

SLD

 

[Cheerful Charlie]

Nate Silver has admitted nobody knows what is really going on, so its hard to work that into the models.
It's going to be a watershed election for pollsters. All the old methods are obsolete.

2018 will be weird as hell as the GOP tries to retake the Senate, no holds barred. How the pollsters handle that will be interesting to watch. Will the extreme GOP obstructionism cost them? How much? Will the womens' vote stay away from the GOP? Or drift back to the GOP? Will the GOP really try to stonewall the supreme court nominations for the next 2 years?

I suspect this off year election will be the most bitter, savage, expensive and crazed off year election ever. And the most heavily polled.

 

[blastula]

I suspect this off year election will be the most bitter, savage, expensive and crazed off year election ever. And the most heavily polled.

You forgot to give percentage odds.

 

[ElectEngr]

Nate Silver has admitted nobody knows what is really going on, so its hard to work that into the models.
It's going to be a watershed election for pollsters. All the old methods are obsolete.

2018 will be weird as hell as the GOP tries to retake the Senate, no holds barred. How the pollsters handle that will be interesting to watch. Will the extreme GOP obstructionism cost them? How much? Will the womens' vote stay away from the GOP? Or drift back to the GOP? Will the GOP really try to stonewall the supreme court nominations for the next 2 years?

I suspect this off year election will be the most bitter, savage, expensive and crazed off year election ever. And the most heavily polled.

Is Cruz and McCain (et al) threatening to block both HRC's court and cabinet nominations or just the court ones if they keep control of the Senate?

Later,
ElectEngr

 

[Cheerful Charlie]

So far, McCain and McConnell have mentioned Supreme Court Justices. But there are about 53 unfilled federal judgeships blocked by the GOP. I do not expect changes there. As for cabinet positions. who knows?

This all smacks of the sort of semi-coups we see down in South America. Banana republic stuff for sure. The GOP has grown comfortable with this approach over the last 8 years and it just keeps getting worse. They keep getting by with this. I would hope the American public would grow weary of this and act, but the popularity of Trump seems to indicate to me that a lot of voters want this type of obstructionism.

 

[credoconsolans]

Depends on what poll you're looking at.

Others show Hillary still with a lead.

And why do Trump supporters care anyway? Didn't Trump say he didn't believe in polls? Or does he only believe in them when they show he's ahead?

Ya, it's amazing how quickly the polls stopped being rigged. It's probably because he called them out on it, so they changed their models to have them stop faking the numbers.

I'll be so glad next week when either nobody needs to give a shit about Donald Trump anymore or we get to watch him try and be President. The former would be a massive relief and the latter would be so fucking funny and entertaining that everybody who's not an American be able to have a good laugh.

How sweetly naive of you. Trump loses, we're not rid of him. He'll attack President Clinton every chance he gets, compare what HE would have done and then run again in 2020 since he 'came so close' and obviously only lost because the election was rigged.

 

[Cheerful Charlie]

I doubt Trump will run in 2020. He'd be in his 70's and he's going to be busy with various legal problems soon. But he'll try to make mischief. He'll try to set up political PACs and organizations to harass Clinton. He'll be like Nutty Newt, all over the right wing media giving his wretched opinion on everything.
He'll "write" (ghostwritten) books ala Ann Coulter. He's not going to just go away and never be heard of again.

 

[Jimmy Higgins]

Ya, it's amazing how quickly the polls stopped being rigged. It's probably because he called them out on it, so they changed their models to have them stop faking the numbers.

I'll be so glad next week when either nobody needs to give a shit about Donald Trump anymore or we get to watch him try and be President. The former would be a massive relief and the latter would be so fucking funny and entertaining that everybody who's not an American be able to have a good laugh.

How sweetly naive of you. Trump loses, we're not rid of him. He'll attack President Clinton every chance he gets, compare what HE would have done and then run again in 2020 since he 'came so close' and obviously only lost because the election was rigged.

Trump will be 76 I think.