Scrutable

lpm wrote: Mon Oct 26, 2020 12:58 pm

lpm wrote: Tue Oct 20, 2020 4:11 pm Jesus, Biden is really slumping in the polls today.

Not clear if it's just a freak day or the start of disaster.

He's plummeted from +10.7 in the 538 average to +10.4.

The slumping has continued to +9.1.

Despair, everyone. DESPAIR.

Mind you, before the 1st Debate and Trump's little visit from a coronavirus, Biden was +7.1 and fairly steady in the 6.5 to 7.5 range. So any reversion to mean type thing hasn't really happened. Trump had a sane 2nd debate and got better, but Biden retains part of his bump.

Almost no recent live caller polls, just panels and IVR polls. With IBD in particular being 4-5 pts more Trump than the other polls, that means the average drops.

State polls, apart from Trafalgar ones, show Biden's lead stable or slightly increasing.

dyqik wrote: Mon Oct 26, 2020 11:36 am On 538, it looks like they've removed Trafalgar from the polling averages.

This was after Trafalgar came out with a poll of Michigan that was 15 points more Trump than a slightly right leaning pollster using the same surveying method. Trafalgar also released their "crosstabs" for the first time, before deleting them, which showed nonsense like 30% of self-identified Democrats voting Trump and 25% of self-identified Republicans voting Biden, and seriously flawed push polling questions.

That's interesting - I was just looking at them this morning and wondering why they were so far out.

There's still Rasmussen, though. Look upon them and wonder wtf.

lpm wrote: Mon Oct 26, 2020 12:58 pm

lpm wrote: Tue Oct 20, 2020 4:11 pm Jesus, Biden is really slumping in the polls today.

Not clear if it's just a freak day or the start of disaster.

He's plummeted from +10.7 in the 538 average to +10.4.

The slumping has continued to +9.1.

Despair, everyone. DESPAIR.

Mind you, before the 1st Debate and Trump's little visit from a coronavirus, Biden was +7.1 and fairly steady in the 6.5 to 7.5 range. So any reversion to mean type thing hasn't really happened. Trump had a sane 2nd debate and got better, but Biden retains part of his bump.

State polls have been a bit more consistent, too. Both were around +7 before the 1st debate and Walter Reed stay, national polls rose to almost +11 then fell back to 9ish, state polls slowly rose to be consistent with a 9 point national race, and right now the two largely agree.

In the 2 hours since I posted +9.1 it's slumped to +8.7.

The trend is clear. At this pace it will be even by first thing Wednesday morning, Trump will be +10 by Friday and at the start of election day Trump will be +29.

lpm wrote: Mon Oct 26, 2020 3:09 pm In the 2 hours since I posted +9.1 it's slumped to +8.7.

The trend is clear. At this pace it will be even by first thing Wednesday morning, Trump will be +10 by Friday and at the start of election day Trump will be +29.

All it needs is Rasmussen to release ten polls showing Trump up a day.

El Pollo Diablo wrote: Mon Oct 26, 2020 2:12 pm There's still Rasmussen, though. Look upon them and wonder wtf.

They retweeted the alt-right GatewayPundit recently, if anyone is still wondering their motives

EACLucifer wrote: Mon Oct 26, 2020 3:30 pm

El Pollo Diablo wrote: Mon Oct 26, 2020 2:12 pm There's still Rasmussen, though. Look upon them and wonder wtf.

They retweeted the alt-right GatewayPundit recently, if anyone is still wondering their motives

And Trump is now reading off their morning poll numbers to his rally.

dyqik wrote: Mon Oct 26, 2020 3:40 pm

EACLucifer wrote: Mon Oct 26, 2020 3:30 pm

El Pollo Diablo wrote: Mon Oct 26, 2020 2:12 pm There's still Rasmussen, though. Look upon them and wonder wtf.

They retweeted the alt-right GatewayPundit recently, if anyone is still wondering their motives

And Trump is now reading off their morning poll numbers to his rally.

I gather he has staffers who work to present him with only the most flattering poll numbers in order to avoid setting off his tantrums, so he may genuinely believe Rasmussen are actually representative, rather than a bunch of bigots with a poor track record as a pollster.

The shape of the 538 model is interesting. They seem to have the highest peak on the rolling average at about 420 electoral college votes for Biden. Surely that’s too high?

It's 413-125 if Biden wins Texas, Iowa, Georgia and Ohio. All currently very close per polling averages.

335-203 if Trump keeps all four.

Could all remain pale red, could all turn pale blue.

Makes perfect sense that 413 is the top of the range, anything beyond that is a bigger stretch and not supported by any polls. The next sizeable state in view is South Carolina which was +14 to Trump in 2016.

Grumble wrote: Mon Oct 26, 2020 4:38 pm The shape of the 538 model is interesting. They seem to have the highest peak on the rolling average at about 420 electoral college votes for Biden. Surely that’s too high?

It's a pretty noisy distribution, so looking at the mode might not be that useful, even with the filtering.

Grumble wrote: Mon Oct 26, 2020 4:38 pm The shape of the 538 model is interesting. They seem to have the highest peak on the rolling average at about 420 electoral college votes for Biden. Surely that’s too high?

It's the most likely single outcome out of the range of different nmbers they are projecting, because there's a wide range of swing in which it happens. It starts at a slight swing towards Biden from current polling, giving him Texas and Georgia and the rest of the close states, but then there's a gap, and swing has to be quite big to start giving him states like Missouri and Kansas. This means all their their simulations with swing in a range from about +2 to +8 towards Biden come out with the same result.

This is different from saying it is the likeliest actual outcome, which is more like the low to mid 300s for Biden. Mode vs Mean, basically.

Video via twitter link

The two minutes leading up to the point Trump throws his toys out of the pram and flees from a relatively gentle interview. The thing the pathetic coward was whining about was that Stahl was being professional and refusing to broadcast stuff that couldn't be verified. The paranoia - treating Stahl's polite questioning as rude and hostile because she didn't fawn over him and applaud him for lying - is genuinely disturbing. Prior to the last few years, writing a character this way in a film would have got one laughed at as ridiculously over-the-top.

lpm wrote: Mon Oct 26, 2020 3:09 pm In the 2 hours since I posted +9.1 it's slumped to +8.7.

The trend is clear. At this pace it will be even by first thing Wednesday morning, Trump will be +10 by Friday and at the start of election day Trump will be +29.

In the five hours since you posted this, it's back up to +9.4.

Biden will start producing new citizen voters to vote for him, having won every registered voter sometime around Saturday morning at this rate.

Well the Dow's fallen 650 points today and both the S&P index and the Nasdaq have fallen too on the back of rising cases and a failure to get a stimulus package agreed.

Plus he's got the head of an advisory council, who he appointed, handing in his resignation over Trump's executive order that could remove protections from civil servants.

Way to go Donald.

What horrifies me more than anything else is that for all his shockingly obvious faults, his lies, corruption and his downright inability to manage things (default is just to pick a fight with anyone) - there's still a chance he'll get re-elected.

What also horrifies me is that the main reason he's trailing in the polls is the pandemic. If it hadn't been for Trump's mishandling of covid, he would probably be favourite to win.

AMS wrote: Mon Oct 26, 2020 9:36 pm What also horrifies me is that the main reason he's trailing in the polls is the pandemic. If it hadn't been for Trump's mishandling of covid, he would probably be favourite to win.

I don't think that's true. Biden had a 5-6 point lead on him before the pandemic. Now, maybe the laptop nonsense etc. would have had more effect without the pandemic, but I'm not sure about that.

Do we know how many of the Conservative Supreme Court Justices are planning to be going to at this 'celebratory swearing in' ?
Just thinking out loud really.

Gfamily wrote: Mon Oct 26, 2020 9:53 pm Do we know how many of the Conservative Supreme Court Justices are planning to be going to at this 'celebratory swearing in' ?
Just thinking out loud really.

Roberts seems to be skipping it, and Thomas is administering the oath.

It'd be a shame, etc...

dyqik wrote: Mon Oct 26, 2020 9:48 pm

AMS wrote: Mon Oct 26, 2020 9:36 pm What also horrifies me is that the main reason he's trailing in the polls is the pandemic. If it hadn't been for Trump's mishandling of covid, he would probably be favourite to win.

I don't think that's true. Biden had a 5-6 point lead on him before the pandemic. Now, maybe the laptop nonsense etc. would have had more effect without the pandemic, but I'm not sure about that.

Covid didn't make a huge difference. Trump's disapproval rating is basically the same as a year ago. And the economy was starting to get lacklustre.

At least several Texas counties, and Maricopa County in Arizona, now have early voting levels past 2016 total turnout.

EACLucifer wrote: Mon Oct 26, 2020 10:36 pm At least several Texas counties, and Maricopa County in Arizona, now have early voting levels past 2016 total turnout.

Urban or rural counties? That’s where the big divide is after all.

EACLucifer wrote: Mon Oct 26, 2020 10:36 pm At least several Texas counties, and Maricopa County in Arizona, now have early voting levels past 2016 total turnout.

My town is almost at 2016's total turnout, even though there's almost nothing at stake here (Biden is 34 points up, Markey is 34 points up, all other races are uncontested).

EACLucifer wrote: Mon Oct 26, 2020 10:36 pm At least several Texas counties, and Maricopa County in Arizona, now have early voting levels past 2016 total turnout.

Today's Siena/NYT poll has Trump +4 in Texas. Thatsca top rated pollster.

Biden doing well with white suburban, but badly with Hispanic males. Apparently this divide helps certain Texas state rep districts overall, which is good, but it's disappointing to be worse than Clinton with Hispanic voters.