538 Update Biden 86 Trump 14 in Forecast, Biden 52.2 to 42.0 in the poll average Biden +10.2

[Johnny2X2X]

Last edited Fri Oct 9, 2020, 01:30 PM - Edit history (1)

https://projects.fivethirtyeight.com/2020-election-forecast/

https://projects.fivethirtyeight.com/polls/president-general/national/

[634-5789]

1. After the 2016 debacle...I will NEVER trust 538 again. n/t

[helpisontheway]

2. Right..It turned against Clinton fast during that final week. Many

thought Nate was just hedging his bets. However, he said the swing states were tightening. Then disaster...

[Johnny2X2X]

4. Why? They were the closest.

They gave Trump about a 1 in 3 chance to win on election day and he won, that means their prediction was closer than just about any other model. The other models showing Hillary at 90% or better were way off.

1 in 3 means that Trump needed just a few things to break his way and they did. 538 says right now if the election were today Hillary would be a 19-1 favorite.

If I say something has a 70% chance of happening and then it doesn't happen, that doesn't mean I was wrong.

[mucifer]

10. Yup they were the most accurate bearer of bad news

[yellowcanine]

6. Had a LOT to do with the Comey announcement though. And this race is not nearly as close.

I am not worried. Things tightened up a bit in early September but all of the movement nationally and in battlegrounds has been in Biden's direction since. Sometimes we can learn the lessons of the last election too well. That was actually Hillary's problem in 2016. She was still following Obama's path to victory and she wasn't Obama.

[Johnny2X2X]

7. Comey definitely swung the election

10 days out was a bombshell and the race became knotted up. And no one ever Mentions Gary Johnson, but Gary Johnson polled at 4.8% going into the lection and would get 3% of the vote, that 1.8% delta went to Trump.

Of note also, 538 says they improved their model. And the inputs to the model are the polls, and 50% of the polls have been adjusted for 2016 turnout models. So the inputs are better (or at least more favorable to Trump) now.

The chances of a surprise happening this time around are way less.

[yellowcanine]

9. "2016" turnout models is key. Because if we don't go well beyond the 2016 turnout we do not deserve

to win.

[unblock]

11. what we *deserve* is a government that doesn't go out of its way to make voting hard or impossible

what we *deserve* includes:

no voter id crap
no selectively reducing the number of voting places
no questionable voting machines
no de-registering people based on questionable name matches with dead people
no disenfranchisement based on criminal record
no effective poll taxes
no gerrymandering
no electoral college nonsense
no biased reporting
and so on.

if we've gotten to the point where every citizen over 18 can easily vote without much (if any) traveling or paying anything or otherwise overcoming various hurdles designed to skew the vote, and *then* we can't muster enough votes, then maybe at that point, we deserve to lose.

until then, we deserve a better democracy than what we have. donnie has polled underwater consistently throughout his 4 years, and if he is deemed the winner in november, it will be because they figured out how to make the official vote count reflect something other than the clear will of the people.

*that*, we don't deserve.

[yellowcanine]

15. Yep. Win now with GOTV and then fix it.

[MatthewG.]

12. Comey, and, ironically, Trump's inexperience decided it

Comey’s announcement very likely swung the election.

Apart from that, Trump ironically benefitted from not having had a political record, which allowed people to believe whatever they wanted about him. A big part of Trump ‘16’s appeal to moderates was that he would be a centrist Republican independent of conventional party dogma. Conservatives, on the other hand, assumed (correctly) he’d govern as a conservative.

[Klaralven]

16. The difference is that now Trump is the least liked candidate.

People are really tired of Trump. The Presidential Apprentice show is over, and the ratings are tanking.

[yellowcanine]

3. Looks as if Georgia and Iowa are getting ready to drop for Biden and Texas not far behind.

[Norbert]

5. 25 Days can't get here soon enough

[malaise]

8. It will be 95% before Halloween

Take that to the bank

[Tribetime]

13. So after Texas it's S.Carolna Arkansas and Missouri

It would be great to see one of those states go to Joe. At 10% it's possible.

[philf99]

14. Alaska

I think is one of those states that could flip in a mega blue tsunami

[Johnny2X2X]

17. Bumped for 86-14 update

From reading 538s methods on their forecast, it is my understanding that the polls get weighted more as we get closer to the election, but not a smooth transitions, but increments at milestones. One of those milestones is 25 days, so tonight or tomorrow the polls will be weighted more and that 86 should turn to 90 by my guess.