Pew National Poll: Clinton +4

[Erich Maria Remarque]

They didn't bother with a two-way race like last time? *sigh*

Because it is not.

[Erich Maria Remarque]

It terrifies me that despite leading Hillary in the low 40s. Looks like she won't get 50% in the GE.

Remember, 3rd parties always fade a bit into the fall, but yes it's highly unlikely anyone gets a majority this year.  However, I think Clinton will easily be over 45%.

Clinton and  especially Trump are the most disliked presidential candidates. Ever.
And if rember correctly, 3rd parties usually fade directly after the major party conventions. Right now it kind of opposite. 3rd parties grows...

[Erich Maria Remarque]

You'd think the third party fade would absolutely come if Johnson doesn't get into the debates, as has happened historically. I'm feeling less sure than ever, though, as I see the Johnson/Stein numbers holding up weirdly consistently in every poll.

Has it? I thought, that the numbers were pretty stable after conventions.
Nader, 2000
Perot, 1996
None of them participated in the debates (though Perot did it in 1992 = high name recognition?).

I'd also say that 3rd party candidates get more media coverage in this year; it might help them.

[Erich Maria Remarque]

But the point is data shows us LV screens in most national polls under-egged Obama by 2-3% in the end. The numbers might look better but that doesn't mean the LV polls are closer to the truth. My bet is that Trump does better in LV screens in diverse (both race and education) states (VA, NV) than he does in more educated and/or whiter states.

Usually it does, since LV polls are usually much/slightly better than RV.


http://fivethirtyeight.com/features/registered-voter-polls-will-usually-overrate-democrats/

And you cannot extrapolate Obama's numbers like you appear to do.

And yes, this election is quite unique.

[Erich Maria Remarque]

RV voters overstate, LV understate, I was too flippant, but building your house of the sand of LV polls is an error, as is investing in RV polls.

My other point about variations between states is starting to appear.

LV polls do better on average.

[Erich Maria Remarque]

LV models don't really matter anymore now that Democrats have mastered the art of targeted early and absentee voting. The people will vote.

Have they really?  Why did they not use that in 2014? Does it work only presidential election?

When did they mastered it?  What evidence can you provide?

[Erich Maria Remarque]

LV models don't really matter anymore now that Democrats have mastered the art of targeted early and absentee voting. The people will vote.

Have they really?  Why did they not use that in 2014? Does it work only presidential election?

When did they mastered it?  What evidence can you provide?

2008 and 2012? There's a lot of articles on the Obama campaign data and outreach operation that Clinton has now inherited. It's how the Democrats have locked up D+6/7 ID advantages in two straight elections.

For whatever reason, Democrats ran and hid during midterms refusing to attach themselves to the President or using the same kind of electioneering.

Articles?

In 2008 RV overestimated Obama with 2.6%. LV with 0.2%.
In 2010 and 2014 both RV overestimated Democrats heavily, while LV were very good.

In 2012 - yeah. But you can't extrapolate just one data point. Even if you might be right, you're probably not.

[Erich Maria Remarque]

WOW!!!

Must be a hot late Summer.....

We have a ton of people arguing over RV/LV models based on one poll, while everyone is really waiting for more A Grade polls, preferably post Labor day.

Regardless of various squabbling from various Red and Blue avatars, over a potential 1-2% swing in national polling numbers, at this point we haven't seen any conclusive evidence that the race is tightening in either national nor statewide polls. So why bother arguing so much over one particular poll, when we haven't really had much to see over the past week to support evidence one way or another?

No, no , no! It's not about one poll, but how it works in general. Some red hacks are saying LV/RV choice doesn't matter, despite the fact that LV polls historically were better. That's it.

Whether LV will be more favourable towards Clinton or Trump is a different question.