The US federal election prediction thread

[9-10ths_Penguin]

That's very interesting.

I suspect this election, there will be some kind of controversy regarding mail in ballots, and at least some won't be counted. And, of course, there are my personal ideas, but is there any way polls take these kinds of circumstances into accounts?

Good models look at past elections and the differences between pre-election polls and the final results to determine - and account for - systemic factors, so a typical amount of spoiled/ineligible ballots shouldn't throw off the models.

New factors that haven't affected any other election? Adjusting for that will probably come down to educated guesses.

 

[halbhh]

I figured it might be interesting to create a thread on the various predictions, models, etc., for the outcome of the upcoming US federal election.

Which models/forecasts/etc. do you follow? What do they say? Which ones do you trust? Post a link!

Please note that this thread is in a discussion only forum.

Starting things off with one I follow: 338 Canada. I got familiar with their forecasts for Canadian elections, but they now do forecasts for the US - the presidential race, anyhow - as well.

Their model currently shows the odds of winning the electoral college at 74.2% Biden, 25.4% Trump. Biden's currently down from his high point in the prediction (~83% in July, IIRC).

Why exactly is Trump trying to paint mail in voting as fraudulent?

Even though total mail-in ballot vote fraud in the U.S. has been low (e.g. 0.0025% of votes), Trump is trying to make it look like mail in voting will be fraudulent ahead of time.

With a lot of repeated rhetoric that only a very devoted follower that believes (faith, religious style belief) in Trump would believe.

Last night on the PBS Newshour, a guest (writer from The Atlantic) was interviewed, and roughly I think he was saying that Trump is trying to discredit mail-in voting because he thinks his own supporters will mostly avoid mail in and thus their votes would get counted from election night faster than the mail-in ballots, so that Trump could claim victory before it actually happened, and then when the mostly democratic votes in the mail-in ballots are counted over the next day or more, then claim those are fraud, even though they are not.

See? A way to try to destroy the U.S. election result and take over like a dictator.

Of course, the opposition to that would be intense not only from democrats, but across the spectrum.

But...Trump really is a great salesman, and may have more than 1 sales tactic (trick mean to create a false impression) up his sleeve.

So, if all that paranoid, but highly plausible scenario happens, it could be pretty rough for a time.

After all, why is Trump working so hard to falsely claim mail-in ballots are fraudulent voting?

Why?

This is a pretty plausible scenario that would explain why.

(can slide time slider to about 23:50 in the video)

 

[Heyo]

In 2018, I predicted that the Democrats would pick up 38 seats, plus/minus 2. They picked up 40.

This year, I predict a Biden win, by about 8% in the popular vote, and a minimum of 310 EC votes.
I predict the Democrats hold onto the house, with a slight increase (about 10 seats)
I predict the Democrats take the Senate by just 2 seats.

I'm not so sure about the presidency but I agree that there's a good chance that the dems hold the House and take the Senate (Being that Trump just thinks about himself and the pubs neglect the down ballot.) And without Moscow Mitch Trump is only half as dangerous.

 

[Mindmaster]

I figured it might be interesting to create a thread on the various predictions, models, etc., for the outcome of the upcoming US federal election.

Jury is out until the debate. Whoever does well there will gain enough points to win nearly automatically probably. By Oct. 15th the cards will be down, IMHO. I expect Biden to have a really hard time there. Sure, you can be all Dem, but I am not confident in his ability to perform and I think that matters. It's going to be on display for the nation in just three days, lol.

 

[Evangelicalhumanist]

Jury is out until the debate. Whoever does well there will gain enough points to win nearly automatically probably. By Oct. 15th the cards will be down, IMHO. I expect Biden to have a really hard time there. Sure, you can be all Dem, but I am not confident in his ability to perform and I think that matters. It's going to be on display for the nation in just three days, lol.

No doubt that Trump will make every effort to rattle Biden in the debates -- and you have a point, it might work. Then again, if Beiden is well enough prepared around a finite (but not small) array of talking points, he may not fare badly. And if that is the case, and I hope it is, then my hopes for a Democrat win are even more likely to be realized.

 

[Mindmaster]

No doubt that Trump will make every effort to rattle Biden in the debates -- and you have a point, it might work. Then again, if Beiden is well enough prepared around a finite (but not small) array of talking points, he may not fare badly. And if that is the case, and I hope it is, then my hopes for a Democrat win are even more likely to be realized.

There is no doubt Trump has profited from every debate he's ever stood in, and now he has way more of them under his belt than Biden. Personally, I think it's locking Biden in a cage with Mike Tyson and expecting him to survive. The literal worst Trump could possibly do is hold, but I find that unlikely in regard to his past debates. The difference between how the Republicans kind of 'Darwin' to get their nominee and the Democrats shilling for one is all the difference. Trump beat 17 other people to win his presidency, and that means a lot in that it took a bunch of time for the party to warm up to him. He was seen as an outsider and everyone was hostile to him initially. This was no easy feat. The pundits, press, and everyone else were mocking his performances while the people were coming out in droves and he was winning his primary votes and later on the electorals that he needed. He knows what to do and despite all the banter of lefties everywhere only a fool would underestimate him at this point. Underestimating his ability to win is the absolute dumbest thing anyone can do at this point. Biden, however, it's not the same... He's relatively run uncontested when he was a senator and the opposition was weak, as it was when he was facing off in the VP debates, he's not going to get that luxury.

 

[Evangelicalhumanist]

There is no doubt Trump has profited from every debate he's ever stood in, and now he has way more of them under his belt than Biden. Personally, I think it's locking Biden in a cage with Mike Tyson and expecting him to survive. The literal worst Trump could possibly do is hold, but I find that unlikely in regard to his past debates. The difference between how the Republicans kind of 'Darwin' to get their nominee and the Democrats shilling for one is all the difference. Trump beat 17 other people to win his presidency, and that means a lot in that it took a bunch of time for the party to warm up to him. He was seen as an outsider and everyone was hostile to him initially. This was no easy feat. The pundits, press, and everyone else were mocking his performances while the people were coming out in droves and he was winning his primary votes and later on the electorals that he needed. He knows what to do and despite all the banter of lefties everywhere only a fool would underestimate him at this point. Underestimating his ability to win is the absolute dumbest thing anyone can do at this point. Biden, however, it's not the same... He's relatively run uncontested when he was a senator and the opposition was weak, as it was when he was facing off in the VP debates, he's not going to get that luxury.

I think that much of what Trump relies on is pure bluster -- and bully tactics. Now, the bluster might do for him, unless Biden really has a hard grasp on the facts. But the bully tactics play only to some people -- and those people are already on his side. The demographic leaving in droves appears to be women, and they don't always respond quite as well to bullies as the men who never went to college do.

 

[Mindmaster]

I think that much of what Trump relies on is pure bluster -- and bully tactics. Now, the bluster might do for him, unless Biden really has a hard grasp on the facts. But the bully tactics play only to some people -- and those people are already on his side. The demographic leaving in droves appears to be women, and they don't always respond quite as well to bullies as the men who never went to college do.

Whatever he does works, and at the end of the day that is what matters. His apparently strength + bullying win some people, his apparent unflinching love for the country wins others, and so on. Most people in this country haven't attended college and that's the largest voting block. Whoever appeals to the family is the one that wins. I see nothing in Biden that does where as with Trump that's obviously important. I realize Dems don't largely live in conventional family arrangements so they probably would care less about those things, but indies and conservatives certainly do. Also, it's worth mentioning that he's been doing a lot for the black voters and I think it matters a lot -- the black men especially are really loving what he is doing. Legal migrant Hispanics like him too as they're much more conservative than what the Dems are currently on politically.

But, we'll see... Certainly Biden has a lot of auto-vote that's helping his cause. A majority of Dem voters vote dem all day and don't care what's going on, but we'll see how COVID has changed that situation.

 

[Heyo]

I figured it might be interesting to create a thread on the various predictions, models, etc., for the outcome of the upcoming US federal election.

Which models/forecasts/etc. do you follow? What do they say? Which ones do you trust? Post a link!

Please note that this thread is in a discussion only forum.

Starting things off with one I follow: 338 Canada. I got familiar with their forecasts for Canadian elections, but they now do forecasts for the US - the presidential race, anyhow - as well.

Their model currently shows the odds of winning the electoral college at 74.2% Biden, 25.4% Trump. Biden's currently down from his high point in the prediction (~83% in July, IIRC).

Cenk called it:

 

[9-10ths_Penguin]

Starting things off with one I follow: 338 Canada. I got familiar with their forecasts for Canadian elections, but they now do forecasts for the US - the presidential race, anyhow - as well.

Their model currently shows the odds of winning the electoral college at 74.2% Biden, 25.4% Trump. Biden's currently down from his high point in the prediction (~83% in July, IIRC).

Well, a few weeks later and Biden's odds have improved slightly. As of today, they have:

- Biden: 77.3%
- Trump: 21.7%
- tie: 0.9%

 

[Ponder This]

I like the 13 keys predictor. Allan Lichtman predicted a Biden victory, but I disagree on his evaluation of 3 of the 13 keys. These 3 keys do make a difference in who the 13 keys model predicts. Charisma Key; Long-Term Economy Key; Foreign Success Key.
Trump has had long apolitical national popularity before he ever ran for President. His rallies, even during a pandemic, always have a lot of people. People seem to want to go see Trump.
The economy over the entire term of his Presidency was doing better. The Pandemic has caused economic downturn only during this year - but we have a short-term economic key already in the prediction model.
And his multiple trade deals, his multiple peace deals, and his resounding defeat of ISIS all indicate success in foreign affairs. I just read today about how a third mid-east country (Sudan) is joining the Peace Agreement with Israel!

Other Predictors:
Polls are probably the worst predictor. We know polls are manipulated strongly up until it gets close to an actual election and they were certainly unreliable in the previous election. Already, the more reliable polls are narrowing as we approach election.
Rasmussen has it down to a three-point race:
Biden 49% Trump 46% Some Other Candidate 2% Undecided 2%
NYTimes has it at Biden 50% Trump 41%

Betting markets were also wrong in the last election, but we know that historically the betting markets are a better predictor than polls. So it should be interesting to see what happens there. Will they get it right this year? Predict It has: Trump 41 c Biden 64 c

Voter Registrations are an interesting predictor. It looks like Republican registration is higher than Democrat registration, which suggests that chances will swing towards Trump (not away from Trump). So... yeah...

I'm looking forward to seeing how the Halloween mask predictor goes and I'm curious if anyone knows of any other fun or silly predictors. Flip a coin, maybe you'll be right half the time!

 

[Heyo]

I like the 13 keys predictor. Allan Lichtman predicted a Biden victory, but I disagree on his evaluation of 3 of the 13 keys. These 3 keys do make a difference in who the 13 keys model predicts. Charisma Key; Long-Term Economy Key; Foreign Success Key.
Trump has had long apolitical national popularity before he ever ran for President. His rallies, even during a pandemic, always have a lot of people. People seem to want to go see Trump.
The economy over the entire term of his Presidency was doing better. The Pandemic has caused economic downturn only during this year - but we have a short-term economic key already in the prediction model.
And his multiple trade deals, his multiple peace deals, and his resounding defeat of ISIS all indicate success in foreign affairs. I just read today about how a third mid-east country (Sudan) is joining the Peace Agreement with Israel!

Other Predictors:
Polls are probably the worst predictor. We know polls are manipulated strongly up until it gets close to an actual election and they were certainly unreliable in the previous election. Already, the more reliable polls are narrowing as we approach election.
Rasmussen has it down to a three-point race:
Biden 49% Trump 46% Some Other Candidate 2% Undecided 2%
NYTimes has it at Biden 50% Trump 41%

Betting markets were also wrong in the last election, but we know that historically the betting markets are a better predictor than polls. So it should be interesting to see what happens there. Will they get it right this year? Predict It has: Trump 41 c Biden 64 c

Voter Registrations are an interesting predictor. It looks like Republican registration is higher than Democrat registration, which suggests that chances will swing towards Trump (not away from Trump). So... yeah...

I'm looking forward to seeing how the Halloween mask predictor goes and I'm curious if anyone knows of any other fun or silly predictors. Flip a coin, maybe you'll be right half the time!

Pat Robertson said that god told him that Trump will win and Pat Robertson is a very reliable predictor - he never ever got it right.

 

[Tambourine]

My prediction: Physical ballots will portray Trump as the winner, mail-in ballots will portray Biden as the winner.
Since Republicans have for a long time seeded the idea that mail-in ballots are intrinsically untrustworthy, they will use this to justify declaring Trump the winner. Chaos ensues.

The Republican-dominated Supreme Court will suppress any attempts at a recount.

 

[Heyo]

My prediction: Physical ballots will portray Trump as the winner, mail-in ballots will portray Biden as the winner.
Since Republicans have for a long time seeded the idea that mail-in ballots are intrinsically untrustworthy, they will use this to justify declaring Trump the winner. Chaos ensues.
The Republican-dominated Supreme Court will suppress any attempts at a recount.

And that's when the riots start.

 

[Terrywoodenpic]

The heavy betting money is saying Trump will lose.
I wonder if Trump is betting on himself to lose or for Biden to win.

 

[9-10ths_Penguin]

My prediction: Physical ballots will portray Trump as the winner, mail-in ballots will portray Biden as the winner.
Since Republicans have for a long time seeded the idea that mail-in ballots are intrinsically untrustworthy, they will use this to justify declaring Trump the winner. Chaos ensues.

The Republican-dominated Supreme Court will suppress any attempts at a recount.

I'm a bit confused by your description: mail-in ballots are physical ballots. In the US, the only non-physical ballots are cast in person.

Other than that, I'm not sure your prediction really lines up with how the (admittedly wonky and inconsistent) American presidential electoral process works. Here's one quick explanation:

Election night marks the end of one phase of campaign 2020 – and the start of another

Short version: for most potential irregularities in the presidential election, it's the newly-elected Senate and House that get to decide, not the Supreme Court.

Edit: of course, none of this would stop Trump doing whatever he wants to spur his supporters to action - whatever that action is - after the election.

 

[Twilight Hue]

Pat Robertson said that god told him that Trump will win and Pat Robertson is a very reliable predictor - he never ever got it right.

Be careful he might run again. ;0]

 

[Heyo]

Be careful he might run again. ;0]

... and totally ruin the economy and get followed by Lisa Simpson.

 

[Stevicus]

... and totally ruin the economy and get followed by Lisa Simpson.

If Pat Robertson ever became President, that would be the least of my worries.

 

[Ponder This]

New Rasmussen poll out today:
Trump 48% Biden 47% Some other Candidate 3% Undecided 2%
first time Trump has been ahead since mid-September.

PredictIt Betting Market:
Trump 41c Biden 63c