If you're a Clinton supporter, how would you describe your mood about the polls?

[Attorney General & PPT Dwarven Dragon]

They should be at least nervous. I think Trump is going to vastly outperform the polls; I think we'll see that there are a lot of Trump voters out there who weren't willing to admit to supporting Trump to pollsters but will do so in the secrecy of the voting booth.

thus far there has been no evidence of that. Only evidence of the opposite: people who are saying they'll vote for Trump, but as soon as they get in that voting booth, reconsider that choice. And admittedly, even evidence of that is scant.

People who were planning to/open to voting for Trump but made a different decision at the booth definitely do exist - even in primaries he won convincingly, I don't remember Trump ever winning among late deciders according to exit polls, and I almost always watched the network coverage.

That's a very good point.  There was a well-documented Shy Anti-Trump effect in the primaries until NY, so we actually have more evidence in favor of that than we do in favor of a Shy Trump effect.  In light of this, I'm surprised the Shy Trump idea's gotten so much traction in the first place.  When in doubt, I would assume the polls are accurate in aggregate.

I checked the exit polls for NY and later among late deciders. These are the results:

NY: Trump 44%, Kasich 39%, Cruz 17% (Anti-Trump 56%, Trump 44%)

PA: Trump 37%, Kasich 29%, Cruz 29%  (Anti-Trump 58%, Trump 37%) (5% refused to answer)

CT: Kasich 45%, Trump 38%, Cruz 12%, Uncommitted 4% (Anti-Trump 61%, Trump 38%) (1% refused to answer)

MD: Trump 37%, Kasich 35%, Cruz 22% (Anti-Trump 57%, Trump 37%) (6% refused to answer)

IN: Cruz 45%, Trump 42%, Kasich 11% (Anti-Trump 56%, Trump 42%) (2% refused to answer)

NE: Trump 59%, Cruz 19%, Kasich 10% (12% refused to answer)

WV: Trump 77%, Cruz 7%, Kasich 5% (11% refused to answer)

And comparison with the actual Trump %:

NY: 60% vs. 44% (-16)
CT: 58% vs. 38% (-20)
PA: 57% vs. 37% (-20)
MD: 54% vs. 37% (-17)
IN: 53% vs. 42% (-11)
NE: 61% vs. 59% (-2)
WV: 77% vs. 77% (+/- 0)

Outside of Nebraska and West Virginia, this is really pathetically weak. Particularly CT. Imagine if only late deciders had been allowed to vote in NY, CT, PA, MD, and IN. Cruz/Kasich would have stayed in and we would have probably gone to an open convention.

[Devout Centrist]

Suicidal/borderline homicidal.

[Willmiller22]

I'm honestly more concerned about the GOP holding the Senate

same here

[Beet]

The primary isn't the same as the GE. Just because Republican primary voters didn't feel shame supporting Trump instead of other Republicans like Cruz or Rubio, it doesn't mean swing voters won't feel shame in supporting Trump instead of Clinton or Johnson. These are two entirely different populations of people, in entirely two different social environments.

The live phone / online poll distinction is plenty of evidence for a Shy Trump effect. The same distinction was seen during the lead up to Brexit polling.

The substantive reasons have to do with the kind of campaign Clinton is running: She's arguing that Trump is temperamentally unfit to be president, suggesting that he shouldn't even seriously be considered. This is an appeal to people's higher, more mature natures. Another example is her ad showing Trump mocking the disabled reporter and cutting to the reactions of kids. Clearly the implication is - if we elect adults him president, how can we tell our kids to have good manners?

The Clinton conceit is that there is some higher adult consciousness that we will ultimately fall back on, but the problem is that adults are really just overgrown children. We have more of a veneer of civility, but emotionally we're not any different. This, incidentally, explains the willingness of many people to believe things that are patently false. In human nature, emotions trump facts. The acceptance of a fact is predicated on the emotional receptivity to an adverse one, and that's not always there. Liberals always profess shock and claim that being educated/reading books will "cure ignorance", but that's not the problem. The problem is human nature. This is where conservatives are right.

[Devout Centrist]

Ok, what the hell is it with the Brexit nonsense? Brexit lead in the polls til election day and was favored to win by several models. Hell, even an exit poll showed a Brexit win. What the f**k guys?

[Attorney General & PPT Dwarven Dragon]

Ok, what the hell is it with the Brexit nonsense? Brexit lead in the polls til election day and was favored to win by several models. Hell, even an exit poll showed a Brexit win. What the f**k guys?

A poll or two may have had leave ahead, but most dud not: 

http://www.cnbc.com/2016/07/04/why-the-majority-of-brexit-polls-were-wrong.html

[AndyHogan14]

Somewhat relaxed...while the national polls and some of the state polls have tightened, she still leads in more than enough states (on average) to win by a larger margin than President Obama did in 2012. With states like Virginia and Colorado in the high single digits (on the border between safe/likely Democrat), it makes Trump's path extremely difficult. I also expect Hillary to get a decent bump out of the debates, which may put states like Georgia and Arizona more in play than they are now.

That said, Democrats cannot be complacent...as the Republican nominee, Trump can still win this thing and HRC and the Democratic Party should be treating it as if they are running from behind.

Ok, what the hell is it with the Brexit nonsense? Brexit lead in the polls til election day and was favored to win by several models. Hell, even an exit poll showed a Brexit win. What the f**k guys?

A poll or two may have had leave ahead, but most dud not: 

http://www.cnbc.com/2016/07/04/why-the-majority-of-brexit-polls-were-wrong.html

If I remember correctly, if you took all of the polls from the beginning of the Brexit campaign to the end there were definitely more pro-Remain results...HOWEVER, I believe the trend line was definitely towards the leave side and while many were saying that the tides had turned against Johnson, Farage, et al at the end of the campaign, I remember being confused because the polls did not reflect what the pundits were saying.

[Old Man Willow]

Nervous. Hillary needs to find a way to skip the debates. I've no doubt she'll crush Trump among the 0.1% of the populace that cares about substance, but the majority of people only care about charisma and emotion, so I'm worried he'll overtake her afterwards.

[Fmr President & Senator Polnut]

Nervous. Hillary needs to find a way to skip the debates. I've no doubt she'll crush Trump among the 0.1% of the populace that cares about substance, but the majority of people only care about charisma and emotion, so I'm worried he'll overtake her afterwards.

What? lol

[ursulahx]

Somewhere between relaxed and nervous. There isn't really a category in the poll which describes my position. As Dems tend to over-worry, I've gone for 'relaxed' as a balance.

[ursulahx]

Incidentally, the CNBC story is wrong about the "vast majority" of polls favouring Remain. While it's true that the last few polls before the vote favoured Remain (not always by much, usually within MoE), a look at all the polls shows a more even distribution. Sorry, I can't post links yet, but there's a table of all the polls on Wikipedia.

While the polling average might have given Remain a lead, it was always a narrow lead. You only have to be off by a percent or so for the result to tip the other way.

Also there wasn't an exit poll.

(Another thing to bear in mind: we weren't choosing a president, the public received very little information about what they were voting on, and the British electorate has a very different composition from the American.)

[DavidB.]

Extremely relaxed. Nothing points to a Trump victory.

[Erich Maria Remarque]

Incidentally, the CNBC story is wrong about the "vast majority" of polls favouring Remain. While it's true that the last few polls before the vote favoured Remain (not always by much, usually within MoE), a look at all the polls shows a more even distribution. Sorry, I can't post links yet, but there's a table of all the polls on Wikipedia.

Every last week poll underestimated the margin. Every single one.
On average they showed Bremain +3, the final results was Brexit +4.

It had probably something to do with killing of Jo Cox, but still Brexit have likely been underestimated all the way.

And Brexit/Johnson and Co was much, much, much less controversial than Trump. Nigel Farage is freaking politicall correctness itself comparing to Trump. Believe me, folks!

And Cameron has been much more successful and popular than Hillary. He is a much better and charismatic orator than Hillary (she really sucks at this LOL) (but Farage is at least that good as Cameron. And sarcastic I will miss his debates in European Parliament. My favourite one is Nigel Farage insults Herman van Rompuy, calls EU President a "DAMP RAG")

[Mr. Morden]

They should be at least nervous. I think Trump is going to vastly outperform the polls; I think we'll see that there are a lot of Trump voters out there who weren't willing to admit to supporting Trump to pollsters but will do so in the secrecy of the voting booth.

thus far there has been no evidence of that. Only evidence of the opposite: people who are saying they'll vote for Trump, but as soon as they get in that voting booth, reconsider that choice. And admittedly, even evidence of that is scant.

Well there can't be any definitive evidence for it (or against it for that matter) until the results of the election are in. But we do have some circumstantial evidence that supports my concern, and that is that Trump does better I online polls than in live polls.

That's what people always say when there's a difference between live interview polls and IVR or internet polls.  The implication then is that the IVR and internet polls are the ones we should believe, because people will lie to other humans but won't lie to a robot.  But where is the evidence that IVR and internet polls are more accurate than live interview polls?  If anything, it's the opposite.  Live interview polls tend to be a bit more accurate, on average.  The differences between them are not necessarily due to poll respondents lying, but may be due to differences in sample construction.  I mean, online pollsters definitely face a bigger challenge in trying to select a representative sample of the electorate.

[Amenhotep Bakari-Sellers]

It's Trump. But, Clinton was being held down in the Quinnipiac polls until recently. The presidential is following the Senate elections.

[Amenhotep Bakari-Sellers]

Clinton needs to stop campaigning in NC and focus on NM, CO & NV

[Erich Maria Remarque]

That's what people always say when there's a difference between live interview polls and IVR or internet polls.  The implication then is that the IVR and internet polls are the ones we should believe, because people will lie to other humans but won't lie to a robot.  But where is the evidence that IVR and internet polls are more accurate than live interview polls?  If anything, it's the opposite.  Live interview polls tend to be a bit more accurate, on average.  The differences between them are not necessarily due to poll respondents lying, but may be due to differences in sample construction.  I mean, online pollsters definitely face a bigger challenge in trying to select a representative sample of the electorate.

There are two funny things.

1 There were no big difference back in Juny, but it is now
2 When you compare non-live polls with live polls, Trump is doing better in blue states, while Hillary in red states, i.e. it might indicate on Shy effect for both candidates.

[MasterJedi]

Slightly nervous but I don't think I should be. Nothing really points to a Trump victory in the end.

[ursulahx]

Every last week poll underestimated the margin. Every single one.

No. Of the twelve polls conducted the week before the vote, three showed a lead for Leave, and one was a tie. I agree that's hardly impressive, but it wasn't every single one. And in the two weeks prior to that, a considerable majority of polls showed a lead for Leave. Who knows why this was the case? Herding? Sampling error? Weighting error?

You must be logged in to read this quote.

You mean the polls swung to Remain because of the murder? I don't know about that; I'd say that's unprovable at best.

You must be logged in to read this quote.

I'd challenge that, but again it's a subjective argument and in any case we're not comparing like with like here.

Are you saying Trump is going to win because there are lots of Shy Trump voters who didn't vote in the primaries? There's plenty of evidence to suggest otherwise.

I'm convinced that Brexit is a very poor example to use as a "why Trump can win" argument.

[Erich Maria Remarque]

Every last week poll underestimated the margin. Every single one.

No. Of the twelve polls conducted the week before the vote, three showed a lead for Leave, and one was a tie. I agree that's hardly impressive, but it wasn't every single one.

Yes.
Brexit won by 3.8%. Every single poll underestimated Brexit. Some polls by 2% (the best one), some by 14% (worst one), on average by 6%. So yeah, every single poll underestimated Brexit. During the last week.

You must be logged in to read this quote.

Polls got him almost perfectly right, when he was getting about 30%-40% (~his hard-core base).
When he started to get 40-60% (after NY), i.e. even those who hadn't Trump as first choice started to vote for him, the polls started to underestimate him.

It is not an evidence, but it might be an indication.

[Sir Mohamed]

Actually relaxed because I am 99.8% confidant that she will win. But I'm surprised that it is still relatively close in most polls.

[‼realJohnEwards‼]

Every last week poll underestimated the margin. Every single one.

No. Of the twelve polls conducted the week before the vote, three showed a lead for Leave, and one was a tie. I agree that's hardly impressive, but it wasn't every single one.

Yes.
Brexit won by 3.8%. Every single poll underestimated Brexit. Some polls by 2% (the best one), some by 14% (worst one), on average by 6%. So yeah, every single poll underestimated Brexit. During the last week.

You must be logged in to read this quote.

Polls got him almost perfectly right, when he was getting about 30%-40% (~his hard-core base).
When he started to get 40-60% (after NY), i.e. even those who hadn't Trump as first choice started to vote for him, the polls started to underestimate him.

It is not an evidence, but it might be an indication.

OTOH, Trump was rising steadily in the polls during that period, so it is understandable that he would end up higher on Election Day.

[Mr. Morden]

Suffice it to say, the fact that polls include people counted as "undecided" or "don't know", all of whom either have to vote for *someone* or stay home, means that most candidates will be "underestimated by the polls" (as they're presumably going to get some votes out of the "undecided" pie).  The fact that Trump only did about as well as the polls predicted (and not better) during most of the primary season was itself unusual.  Most candidates "beat the polls" by getting some undecideds.

[ursulahx]

Brexit won by 3.8%. Every single poll underestimated Brexit. Some polls by 2% (the best one), some by 14% (worst one), on average by 6%. So yeah, every single poll underestimated Brexit. During the last week.

My bad, I misunderstood what you were saying.

You must be logged in to read this quote.

I'm not sure about this, but I don't have the data to hand. Did Trump actually outperform his polls in any state, while the race was still competitive?

Even if the answer is yes, the idea that there are enough 'Shy Trumpeters' to put him over the top in enough states where he's behind is.. optimistic.

[King]

Always surprises me how much undeserved confidence the Republicans always have on this forum when it's obvious that the party is going to lose. We had people in 2012 that were convinced Romney was going to win in a landslide right up until the polls closed. We had people in 2008 saying there was a noticeable shift in Pennsylvania and it was up for grabs and McCain was going to win and the upset the election still. We had people in 2006 saying how the polls are overestimating Democrats and the Republicans carried a good chance of retaining Congress still.

And here we are in 2016 with a new Trump brand of Republican supporters. Down 270- 179 in the State polling average and down at least 4 in the National polling average, and they are convinced that it is Democrats who should be nervous and on the verge of tears because they're not up by 20 points. If Trump were up 270 - 179 and holding on to a slim popular vote lead, we would never hear the end of it. The election will be called the greatest lock in history. A landslide! Trump doesn't even need to participate in the debates! The election is over!