Will Clinton/Trump Outperform the polls?

[The Other Castro]

Between theories of Silent Majorities, Shy-Trump voters, under-representation of Hispanics/Latinos, and ground game strength, there is a possibility that the polling average does not line up with the end result. If the RCP polling average on Election Day showed, for example, Clinton +6, by how much do you think she or Trump would outperform the polling average?

Bonus Question: Mention any specific states where you think either candidate will beat polling expectations and by how much.

For reference, Obama outperformed the RCP average by 3.2 in 2012, fell behind by 0.3 in 2008, and Bush outperformed by 0.9 in 2004.

[Joe Biden is your president. Deal with it.]

Trump will do worse likely, he's only polling in the high 30s and low 40s.

[Spark]

Trump +3

[Erich Maria Remarque]

At least as big as Brexit. Trump+6

[Seriously?]

The polls are a disjointed mess right now, but by the time November rolls around, more likely than not, they'll all have their methodology down pat and a consensus will form.

Barring any last minute "surprise" (e.g. Bush DUI-story the weekend before the election, Hurricane Sandy, etc.),  the polls on aggregate will be accurate within 1-2 points with LV.

The state polls won't have as many data points, so the individual states could be off a tad bit more.

It's hard to predict which assumptions are correct at this point as most national live pollsters have not employed their LV shield yet.

[evergreenarbor]

I expect Clinton to pick up more undecideds, so I think she'll narrowly outperform the polls.

[Wells]

Clinton +2.

[HagridOfTheDeep]

Clinton. Her ground game advantage has to be worth something.

[Xing]

Trump+10 because muh enthusiasm and primary turnout!!!!11!!!!! And emails!!!!!11!!!!

/s

Clinton+1 or 2.

[Erich Maria Remarque]

Clinton. Her ground game advantage has to be worth something.

Shouldn't her ground game advantage be revealed by polls?

[Horus]

About +2 Trump. There will be a "shy Trump" factor, but thankfully it won't be quite enough for him to win.

[ProudModerate2]

There are small indications on both sides that may effect (outperform) the polls.
But the net-effect compared to the final RCP national average will be Clinton +1 (my vote).

[twenty42]

I have a very hard time seeing the polls at Clinton +6 by the time November rolls around. Right now we are in a malaise period between the conventions and the debates, and I think a lot of Johnson and Stein supporters will reluctantly end up backing a real candidate as the leaves turn gold. I also see the debates helping Trump by default. They are the televised sporting events of the election, and I don't see Hillary edging Donald out in the game of reality TV. The way things are scheduled to happen, in my eyes at least, portends closer polls on Halloween than on Labor Day.

To make a prediction on the question, I can see Clinton outperforming the polls, but not from a +6 vantage point. I can envision a Clinton +2.5 lead going into the election and her winning by 3 to 3.5 in real life. I think the results will be very 2012-esque after the dust settles.

[Fmr President & Senator Polnut]

Clinton +2 to allow for ground-game and Hispanic under-polling.

[Desroko]

Clinton +2 to allow for ground-game and Hispanic under-polling.

[Erich Maria Remarque]

Clinton. Her ground game advantage has to be worth something.

Shouldn't her ground game advantage be revealed by polls?

If you accept that explanation for 2012, no.  If you think 2012 was just about Sandy or some other factor, then yes.

Do you have the same explanation for 2000? Then it was R+2.1 (in 2012 R+3.4) according to 538.

How did Obama's "ground game" change in 2012 exactly, accrodign to you, that it made polls underestimate him?

[psychprofessor]

I have a very hard time seeing the polls at Clinton +6 by the time November rolls around. Right now we are in a malaise period between the conventions and the debates, and I think a lot of Johnson and Stein supporters will reluctantly end up backing a real candidate as the leaves turn gold. I also see the debates helping Trump by default. They are the televised sporting events of the election, and I don't see Hillary edging Donald out in the game of reality TV. The way things are scheduled to happen, in my eyes at least, portends closer polls on Halloween than on Labor Day.

To make a prediction on the question, I can see Clinton outperforming the polls, but not from a +6 vantage point. I can envision a Clinton +2.5 lead going into the election and her winning by 3 to 3.5 in real life. I think the results will be very 2012-esque after the dust settles.

It's hard for me to see Trump outperforming the polls, especially with his Gallup favorables hovering around 30-32 percent. Those are locked in. He isn't going to miraculously turn those numbers around. The only hope he has is for Clinton's low 40's favorables to tank, which I also find implausible. I think the margin between the two candidates will be the margin on election day between their favorables - so right now I'd peg the race at Clinton +8-10.

[Erich Maria Remarque]

I have a very hard time seeing the polls at Clinton +6 by the time November rolls around. Right now we are in a malaise period between the conventions and the debates, and I think a lot of Johnson and Stein supporters will reluctantly end up backing a real candidate as the leaves turn gold. I also see the debates helping Trump by default. They are the televised sporting events of the election, and I don't see Hillary edging Donald out in the game of reality TV. The way things are scheduled to happen, in my eyes at least, portends closer polls on Halloween than on Labor Day.

To make a prediction on the question, I can see Clinton outperforming the polls, but not from a +6 vantage point. I can envision a Clinton +2.5 lead going into the election and her winning by 3 to 3.5 in real life. I think the results will be very 2012-esque after the dust settles.

It's hard for me to see Trump outperforming the polls, especially with his Gallup favorables hovering around 30-32 percent. Those are locked in. He isn't going to miraculously turn those numbers around. The only hope he has is for Clinton's low 40's favorables to tank, which I also find implausible. I think the margin between the two candidates will be the margin on election day between their favorables - so right now I'd peg the race at Clinton +8-10.

You do understand that Gallup is included in THE polls that possibly might underestimate him?

I see just two reason that could explain possible underestimation, Shy Trump and that polls miss some groups. In both cases Gallup would be effected as well, of course.

Or do you have some other reasons?

[psychprofessor]

I have a very hard time seeing the polls at Clinton +6 by the time November rolls around. Right now we are in a malaise period between the conventions and the debates, and I think a lot of Johnson and Stein supporters will reluctantly end up backing a real candidate as the leaves turn gold. I also see the debates helping Trump by default. They are the televised sporting events of the election, and I don't see Hillary edging Donald out in the game of reality TV. The way things are scheduled to happen, in my eyes at least, portends closer polls on Halloween than on Labor Day.

To make a prediction on the question, I can see Clinton outperforming the polls, but not from a +6 vantage point. I can envision a Clinton +2.5 lead going into the election and her winning by 3 to 3.5 in real life. I think the results will be very 2012-esque after the dust settles.

It's hard for me to see Trump outperforming the polls, especially with his Gallup favorables hovering around 30-32 percent. Those are locked in. He isn't going to miraculously turn those numbers around. The only hope he has is for Clinton's low 40's favorables to tank, which I also find implausible. I think the margin between the two candidates will be the margin on election day between their favorables - so right now I'd peg the race at Clinton +8-10.

You do understand that Gallup is included in THE polls that possibly might underestimate him?

I see just two reason that could explain possible underestimation, Shy Trump and that polls miss some groups. In both cases Gallup would be effected as well, of course.

Or do you have some other reasons?

Nope, why would his favorables be underestimated?

[Erich Maria Remarque]

I have a very hard time seeing the polls at Clinton +6 by the time November rolls around. Right now we are in a malaise period between the conventions and the debates, and I think a lot of Johnson and Stein supporters will reluctantly end up backing a real candidate as the leaves turn gold. I also see the debates helping Trump by default. They are the televised sporting events of the election, and I don't see Hillary edging Donald out in the game of reality TV. The way things are scheduled to happen, in my eyes at least, portends closer polls on Halloween than on Labor Day.

To make a prediction on the question, I can see Clinton outperforming the polls, but not from a +6 vantage point. I can envision a Clinton +2.5 lead going into the election and her winning by 3 to 3.5 in real life. I think the results will be very 2012-esque after the dust settles.

It's hard for me to see Trump outperforming the polls, especially with his Gallup favorables hovering around 30-32 percent. Those are locked in. He isn't going to miraculously turn those numbers around. The only hope he has is for Clinton's low 40's favorables to tank, which I also find implausible. I think the margin between the two candidates will be the margin on election day between their favorables - so right now I'd peg the race at Clinton +8-10.

You do understand that Gallup is included in THE polls that possibly might underestimate him?

I see just two reason that could explain possible underestimation, Shy Trump and that polls miss some groups. In both cases Gallup would be effected as well, of course.

Or do you have some other reasons?

Nope, why would his favorables be underestimated?

Ehm...
You said that you don't see Trump outperforming the polls partly because of Gallup favorables.
But IF he will, than he will outperform Gallup is well. So it is a cyclic argument.

[Pragmatic Conservative]

Clinton will probably win a majority of undecided and slightly over perform the polls.

[dspNY]

I have a very hard time seeing the polls at Clinton +6 by the time November rolls around. Right now we are in a malaise period between the conventions and the debates, and I think a lot of Johnson and Stein supporters will reluctantly end up backing a real candidate as the leaves turn gold. I also see the debates helping Trump by default. They are the televised sporting events of the election, and I don't see Hillary edging Donald out in the game of reality TV. The way things are scheduled to happen, in my eyes at least, portends closer polls on Halloween than on Labor Day.

To make a prediction on the question, I can see Clinton outperforming the polls, but not from a +6 vantage point. I can envision a Clinton +2.5 lead going into the election and her winning by 3 to 3.5 in real life. I think the results will be very 2012-esque after the dust settles.

It's hard for me to see Trump outperforming the polls, especially with his Gallup favorables hovering around 30-32 percent. Those are locked in. He isn't going to miraculously turn those numbers around. The only hope he has is for Clinton's low 40's favorables to tank, which I also find implausible. I think the margin between the two candidates will be the margin on election day between their favorables - so right now I'd peg the race at Clinton +8-10.

You do understand that Gallup is included in THE polls that possibly might underestimate him?

I see just two reason that could explain possible underestimation, Shy Trump and that polls miss some groups. In both cases Gallup would be effected as well, of course.

Or do you have some other reasons?

Gallup polling almost 90,000 voters and finding Trump's disapproval #s at 65% pretty much removes a Shy Trump effect. The margin of error on that deep-dive poll (different than the Gallup favorable polls we see listed) is less than 1%

[Erich Maria Remarque]

I have a very hard time seeing the polls at Clinton +6 by the time November rolls around. Right now we are in a malaise period between the conventions and the debates, and I think a lot of Johnson and Stein supporters will reluctantly end up backing a real candidate as the leaves turn gold. I also see the debates helping Trump by default. They are the televised sporting events of the election, and I don't see Hillary edging Donald out in the game of reality TV. The way things are scheduled to happen, in my eyes at least, portends closer polls on Halloween than on Labor Day.

To make a prediction on the question, I can see Clinton outperforming the polls, but not from a +6 vantage point. I can envision a Clinton +2.5 lead going into the election and her winning by 3 to 3.5 in real life. I think the results will be very 2012-esque after the dust settles.

It's hard for me to see Trump outperforming the polls, especially with his Gallup favorables hovering around 30-32 percent. Those are locked in. He isn't going to miraculously turn those numbers around. The only hope he has is for Clinton's low 40's favorables to tank, which I also find implausible. I think the margin between the two candidates will be the margin on election day between their favorables - so right now I'd peg the race at Clinton +8-10.

You do understand that Gallup is included in THE polls that possibly might underestimate him?

I see just two reason that could explain possible underestimation, Shy Trump and that polls miss some groups. In both cases Gallup would be effected as well, of course.

Or do you have some other reasons?

Gallup polling almost 90,000 voters and finding Trump's disapproval #s at 65% pretty much removes a Shy Trump effect. The margin of error on that deep-dive poll (different than the Gallup favorable polls we see listed) is less than 1%

IF there is a significant Shy Trump effect (probably there isn't), than Gallup would not know it even if they do poll every single person. That is the definition.

[Desroko]

I have a very hard time seeing the polls at Clinton +6 by the time November rolls around. Right now we are in a malaise period between the conventions and the debates, and I think a lot of Johnson and Stein supporters will reluctantly end up backing a real candidate as the leaves turn gold. I also see the debates helping Trump by default. They are the televised sporting events of the election, and I don't see Hillary edging Donald out in the game of reality TV. The way things are scheduled to happen, in my eyes at least, portends closer polls on Halloween than on Labor Day.

To make a prediction on the question, I can see Clinton outperforming the polls, but not from a +6 vantage point. I can envision a Clinton +2.5 lead going into the election and her winning by 3 to 3.5 in real life. I think the results will be very 2012-esque after the dust settles.

It's hard for me to see Trump outperforming the polls, especially with his Gallup favorables hovering around 30-32 percent. Those are locked in. He isn't going to miraculously turn those numbers around. The only hope he has is for Clinton's low 40's favorables to tank, which I also find implausible. I think the margin between the two candidates will be the margin on election day between their favorables - so right now I'd peg the race at Clinton +8-10.

You do understand that Gallup is included in THE polls that possibly might underestimate him?

I see just two reason that could explain possible underestimation, Shy Trump and that polls miss some groups. In both cases Gallup would be effected as well, of course.

Or do you have some other reasons?

Gallup polling almost 90,000 voters and finding Trump's disapproval #s at 65% pretty much removes a Shy Trump effect. The margin of error on that deep-dive poll (different than the Gallup favorable polls we see listed) is less than 1%

IF there is a significant Shy Trump effect (probably there isn't), than Gallup would not know it even if they do poll every single person. That is the definition.

There's no significant divergence between live interviewers and IVR this cycle, which proved to be a very good predictor of Trump's vote in the primaries. Internet surveys have been more favorable, but they also overestimated his appeal.

[Erich Maria Remarque]

I have a very hard time seeing the polls at Clinton +6 by the time November rolls around. Right now we are in a malaise period between the conventions and the debates, and I think a lot of Johnson and Stein supporters will reluctantly end up backing a real candidate as the leaves turn gold. I also see the debates helping Trump by default. They are the televised sporting events of the election, and I don't see Hillary edging Donald out in the game of reality TV. The way things are scheduled to happen, in my eyes at least, portends closer polls on Halloween than on Labor Day.

To make a prediction on the question, I can see Clinton outperforming the polls, but not from a +6 vantage point. I can envision a Clinton +2.5 lead going into the election and her winning by 3 to 3.5 in real life. I think the results will be very 2012-esque after the dust settles.

It's hard for me to see Trump outperforming the polls, especially with his Gallup favorables hovering around 30-32 percent. Those are locked in. He isn't going to miraculously turn those numbers around. The only hope he has is for Clinton's low 40's favorables to tank, which I also find implausible. I think the margin between the two candidates will be the margin on election day between their favorables - so right now I'd peg the race at Clinton +8-10.

You do understand that Gallup is included in THE polls that possibly might underestimate him?

I see just two reason that could explain possible underestimation, Shy Trump and that polls miss some groups. In both cases Gallup would be effected as well, of course.

Or do you have some other reasons?

Gallup polling almost 90,000 voters and finding Trump's disapproval #s at 65% pretty much removes a Shy Trump effect. The margin of error on that deep-dive poll (different than the Gallup favorable polls we see listed) is less than 1%

IF there is a significant Shy Trump effect (probably there isn't), than Gallup would not know it even if they do poll every single person. That is the definition.

There's no significant divergence between live interviewers and IVR this cycle, which proved to be a very good predictor of Trump's vote in the primaries. Internet surveys have been more favorable, but they also overestimated his appeal.

They were good, when only those who had really favorable view of Trump (when he were getting 30-40%), but when he start to get majority of voters (i.e. even those voters who'd rather have someone else as nominee) underestimation started to show up. It might indicate Shy effect. Or it might be explained by some other reasons.

My point here is that Shy Trump effect is not likely to be among those voters who really love him and go to his rallies, but those who for example likes his platform, but not the way he is talking about it.

I, for example, think that nor Trump, nor his platform are racist, but that he appeals to racists. It is something I can live with. And some would not like to associate themselves with those racists by admitting they vote/like Trump.