There is one HUGE difference in the polling for Biden in 2020 vs Clinton in 2016

[Roland99]

Biden has been at or above 50% for some time.

SO, even if all "undecideds" broke for trump (and ignored the third parties), trump would still not be able to catch up!

Biden is polling well above where Clinton was and has been for much of the year.


Nationwide:

Link to tweet

[Wounded Bear]

1. Yeah, 50% is kind of a magic number...

but anybody who hasn't voted yet had better be determined to do so.

It ain't over until it's over.

[regnaD kciN]

4. I'd say 52% is the magic number...

If Joe hits that number, the closest Trump can come is 4 points (52%-48%). Why is this important? Because, as Nate Silver pointed out, any popular margin less than 4% results in a situation where there's a very high chance of another popular/electoral vote split, handing Trump the presidency just like in 2016. Every percent above 4% we can win by makes such an event more unlikely.

[Roland99]

5. at 538, Biden has been averaging 50% or higher since Aug 10

[Roland99]

2. not as rosy for battleground states, though

Altho for some key battleground states, RCP showing trump slightly up (I think this is mostly due to several pollsters adjusting for errors made in 2016)

Link to tweet

Altho for some key battleground states, RCP showing trump slightly up (I think this is mostly due to several pollsters adjusting for errors made in 2016)

Even the pro-trump Trafalgar has Biden +2 in PA! That's HUGE!. He takes PA and it's most likely all over for trump.


ALSO, unclear as to what the date cut-off is for the 2020 numbers. Not sure if it includes the last day or two

ALSO ALSO, keep in mind that RCP tends to skew a bit further right than 538. That's why I like to kinda refer to it more. When I see good numbers there, I feel much more comfortable.

[Rstrstx]

14. Clinton was at her high water mark at this time in the fall of 2016

The Access Hollywood tape had just come out and Trump’s numbers tanked for a week or so.

Also, the Trafalgar Group has Biden slightly ahead not only in PA but in MI and WI as well (they also predicted in Trump would win NV in 2016 so their record was not perfect that year, I’m suspect about their polls in Western states). Their margins in MI and PA for Trump last year were about a point higher or more than what the actual numbers were. They also missed some Senate races in 2018. So their results are usually the best case scenario for Trump.

[Roland99]

16. that's why I'm a bit more optimistic about Biden in PA and the upper midwest

I'm writing off FL but I think PA is definitely looking good. That would be the kicker as I think MI/WI would also go Biden

Not counting on OH but I think AZ/NV will go Biden.


My current worst-cast and best-case Electoral College scenarios:

Link to tweet

[DonaldsRump]

3. That is an enormous point

In 2016, there were so many undecideds. Trump just had to convince them to vote for him.

With Joe over 50%, Trump has to convince Joe's voters (a) not to vote for Joe; and (b) to vote for Trump. That is a much harder task.

[sweetloukillbot]

6. I've seen the point made a lot that the margins are the same as this time in 2016

The margins may have been, but there's a big difference between 52-42 and 47-37. A helluva a lot less undecides this time around. And they all broke for Trump last time.

[Wounded Bear]

7. Yes, and that should save us...

but anybody who hasn't voted yet should be determined to do so.

[Cicada]

8. 53% is the new 50%, per Nate Silver

Because of the electoral college bias Biden needs 53% of the popular vote

Hillary had 52% and lost

[Roland99]

10. yeah...and I believe I'd read where Nate had mentioned forecasting changes done to account for that

[TDale313]

11. So friggin' irritating that we basically spot them 3%

The Electoral College sucks. Just sayin’

[sweetloukillbot]

15. When did Hillary have 52%?

She won 48-46.

[Roland99]

17. true. But, also, between Stein, Johnson and McMullin, 5% of the total vote went third-party

we'd be lucky to see 3-3.5% this time.

[Cicada]

18. Sorry. Hillary had 51% of the two party vote, Biden will need 51.5%, not 53%.

[BannonsLiver]

19. Hillary got 48 percent in 2016 not 52

Trump was at 46.

[cbdo2007]

9. Also look at the consistency of the Biden polling vs Clinton's...

in 2016 there were a few times where Clinton and Trump were within 3-4 points and then they would separate, I think even Trump took the lead for a short time.

Not with Biden - he has been 8-10 points ahead for months and has been above the 50% mark for 2 months. Also, no viable 3rd party candidates this year.

[Johnny2X2X]

12. Gary Johnson also!

19 days out in 2016 Gary Johnson was still polling at 6.2%, he would get 3% of the vote, that delta was the supposed "hidden Trump voter."

In fact Johnson was polling at 4.8% on election day, he got 3.0%, that 1.8% was coincidentally exactly the same margin the average polls were off. Average was Hillary +3.9, she won by 2.1.

[Roland99]

13. Great point. Combined, Jorgensen/Hawkins (Libertarian/Green) are polling 3%

about another 3% undecided.

With Biden at 51.4% at RCP and 52.4% at 538, even if EVERY SINGLE Libertarian, Green, or undecided voted for trump, he couldn't catch up. And odds are Biden will get at least 1/4 - 1/3 of those, pushing up probably closer to 53-54%.

And Hawkins is NOT on the ballot in all 50 states.

Link to tweet