• Moderators, please send me a PM if you are unable to access mod permissions. Thanks, Habsy.

OT: American Politics

zeke said:
yes i know you're slavishly devoted to these economic prognosticators, but unless he now admits that the concept of stimulus itself has drastically outperformed and now in his mind superceded his favorite cliche economic theories, i'm going to mock him.
Click to expand...

I actually don’t really pay attention to them… none of them are consistently correct
 
Preston said:
I fucking love polls again.



Depending on what they continue to show
Click to expand...

I'm still #CrosstabMafia4Lyfe

My expectation fwiw, was always that the Dem candidate was going to end up very close to the top end of their MoE based on shitty sampling and poll weightings. So Biden 46 with a 3% MoE means Biden 48.5-49.

The same holds for Kamala. The polling problems haven't fundamentally changed, we're just going to see a more stark reality of it now that someone who can beat Trump by 5+ points in the popular vote instead of 2-3 points is at the top of the ticket.
 
CH1 said:
I actually don’t really pay attention to them… none of them are consistently correct
Click to expand...

going from "Stimulus is the most irresponsible economic tool possible" to "omg this is the greatest economic recovery ever" is more than just being 'incorrect'.
 
zeke said:
going from "Stimulus is the most irresponsible economic tool possible" to "omg this is the greatest economic recovery ever" is more than just being 'incorrect'.
Click to expand...

That’s how prognosticators speak. You should know that first hand in your sports journey
 
MindzEye said:
I'm still #CrosstabMafia4Lyfe

My expectation fwiw, was always that the Dem candidate was going to end up very close to the top end of their MoE based on shitty sampling and poll weightings. So Biden 46 with a 3% MoE means Biden 48.5-49.

The same holds for Kamala. The polling problems haven't fundamentally changed, we're just going to see a more stark reality of it now that someone who can beat Trump by 5+ points in the popular vote instead of 2-3 points is at the top of the ticket.
Click to expand...
I may go full Jonas on you if you're right. Not sure if you want that but just thought I'd let you know.
 
zeke said:
So there's 5 polls where the sample started on monday, after Joe quit:

Rasmussen: Trump +7 (lol)
CNN: Trump +3 (except that this was a re-sample of their previous poll)
Marist: Trump +1
Morning Consult Kamala: +1
Reuters: Kamala +2

there's also one other poll that at least includes monday polling (but starts on friday), and that's Yahoo which has the race even.

so it's tied on average when we rightly ignore Rasmussen.

given that the CNN poll was just a re-sample of a previous Trump +6 over Biden poll, that one might be a bit dubious too.
Click to expand...

and while ME's poll unskewing is admirable, let's not forget that Biden's polling had legit plunged into a dire place.

JUL 19-21 Quinnipiac: Trump +3
JUL 19-21 HarrisX: Trump +6
JUL 19-21 Florida-Atlantic: Trump +8
JUL 19-21 Echelon: Biden +1
JUL 18-21 Rasmussen: Trump +8
JUL 14-21 Bullfinch: Trump +5
JUL 16-18 YouGov: Trump +5
JUL 16-18 CBS: Trump +5
JUL 17-17 SoCal: Trump +6
JUL 15-17 3W: Trump +4
JUL 15-17 RMG: Trump +5
JUL 15-17 Morning Consult: Trump +4
JUL 16-16 Ipsos: Trump +2
JUL 15-16 Emerson: Trump +5

The average had moved into the +5 range - beyond possible polling bias imo.
 
zeke said:
and while ME's poll unskewing is admirable, let's not forget that Biden's polling had legit plunged into a dire place.

JUL 19-21 Quinnipiac: Trump +3
JUL 19-21 HarrisX: Trump +6
JUL 19-21 Florida-Atlantic: Trump +8
JUL 19-21 Echelon: Biden +1
JUL 18-21 Rasmussen: Trump +8
JUL 14-21 Bullfinch: Trump +5
JUL 16-18 YouGov: Trump +5
JUL 16-18 CBS: Trump +5
JUL 17-17 SoCal: Trump +6
JUL 15-17 3W: Trump +4
JUL 15-17 RMG: Trump +5
JUL 15-17 Morning Consult: Trump +4
JUL 16-16 Ipsos: Trump +2
JUL 15-16 Emerson: Trump +5

The average had moved into the +5 range - beyond possible polling bias imo.
Click to expand...
Yeah, I definitely think this is fair - as calls to step down grew the polls seemed to dip beyond that. The last ones were a bit scary.
 
Preston said:
I may go full Jonas on you if you're right. Not sure if you want that but just thought I'd let you know.
Click to expand...

It's largely what happened to Trump in 2016 when everyone was saying that the polls were "wrong". They weren't wrong per se, in that they all were wrong within their MoE, but the weird thing was that they were all wrong in the same direction. Pollsters had something fundamentally wrong in their sampling and polling weighting, so Trump "out performed" everywhere by 2-3 points. We've seen the same thing happen but to the left in the last few US elections, the French Elections, UK Elections, etc. Would be shocking if we didn't see similar here. I just think that fundamentally pollsters don't know how to model likely voters anymore.
 
zeke said:
and while ME's poll unskewing is admirable, let's not forget that Biden's polling had legit plunged into a dire place.

JUL 19-21 Quinnipiac: Trump +3
JUL 19-21 HarrisX: Trump +6
JUL 19-21 Florida-Atlantic: Trump +8
JUL 19-21 Echelon: Biden +1
JUL 18-21 Rasmussen: Trump +8
JUL 14-21 Bullfinch: Trump +5
JUL 16-18 YouGov: Trump +5
JUL 16-18 CBS: Trump +5
JUL 17-17 SoCal: Trump +6
JUL 15-17 3W: Trump +4
JUL 15-17 RMG: Trump +5
JUL 15-17 Morning Consult: Trump +4
JUL 16-16 Ipsos: Trump +2
JUL 15-16 Emerson: Trump +5

The average had moved into the +5 range - beyond possible polling bias imo.
Click to expand...


In the end it worked out for the best, but that was largely a function of Biden's party turning on him publicly imo.
 
CH1 said:
That’s how prognosticators speak. You should know that first hand in your sports journey
Click to expand...

Larry Summers is a former Treasury Secretary, one of the most respected economic voices in the world.

If he has actually suddenly realized that large stimulus can be the best possible solution, after a lifetime of austerity bullshit, a complete flip flop on the most significant philisophical divide in economic and fiscal policy over the last century, then that is a monumental thing which could impact worldwide economic policy for the foreseable future.

If he's just saying "oopsie, nevermind", then i'll mock away, thanks.
 
zeke said:
Larry Summers is a former Treasury Secretary, one of the most respected economic voices in the world.

If he has actually suddenly realized that large stimulus can be the best possible solution, after a lifetime of austerity bullshit, a complete flip flop on the most significant philisophical divide in economic and fiscal policy over the last century, then that is a monumental thing which could impact worldwide economic policy for the foreseable future.

If he's just saying "oopsie, nevermind", then i'll mock away, thanks.
Click to expand...

Never suggested you aren’t free to mock
#firstamendment
 
I would just say that I agree that if a poll shows a virtual toss up but also weird cross tabs which seem to buck historical election trends, then I feel like you can probably push it in that direction. Like one yesterday had Trump getting double the black vote he had the last two times against Kamala.
 
MindzEye said:
It's largely what happened to Trump in 2016 when everyone was saying that the polls were "wrong". They weren't wrong per se, in that they all were wrong within their MoE, but the weird thing was that they were all wrong in the same direction. Pollsters had something fundamentally wrong in their sampling and polling weighting, so Trump "out performed" everywhere by 2-3 points. We've seen the same thing happen but to the left in the last few US elections, the French Elections, UK Elections, etc. Would be shocking if we didn't see similar here. I just think that fundamentally pollsters don't know how to model likely voters anymore.
Click to expand...

but remember that last Biden v Trump election, Biden's lead in the polls at this time was very large, and the actual election results were much closer than his polling lead.

I was prepared to believe that the polls had overcompensated, but for the polls to go from a significant biden skew all the way to an equally signifianct trump skew is still a bit of a stretch to me. I can except a small trump skew, but once it started going solidly over the MOE it was too much imo.
 
hockeylover said:
I would just say that I agree that if a poll shows a virtual toss up but also weird cross tabs which seem to buck historical election trends, then I feel like you can probably push it in that direction. Like one yesterday had Trump getting double the black vote he had the last two times against Kamala.
Click to expand...

Or polls showing that 41% or respondants didn't vote in the 2020 election, etc, etc, etc.

I've seen some super weird shit buried in the #crosstabs since I started doing this.
 
MindzEye said:
In the end it worked out for the best, but that was largely a function of Biden's party turning on him publicly imo.
Click to expand...

sure, but i'd still argue there was a notable 2ish point shift immediately post-debate. and then when his party started giving up on him the polls plunged sharply. but you could also argue that the polls were being held artificially close by everyone toeing the party line.
 
MindzEye said:
It's largely what happened to Trump in 2016 when everyone was saying that the polls were "wrong". They weren't wrong per se, in that they all were wrong within their MoE, but the weird thing was that they were all wrong in the same direction. Pollsters had something fundamentally wrong in their sampling and polling weighting, so Trump "out performed" everywhere by 2-3 points. We've seen the same thing happen but to the left in the last few US elections, the French Elections, UK Elections, etc. Would be shocking if we didn't see similar here. I just think that fundamentally pollsters don't know how to model likely voters anymore.
Click to expand...
Any decent model oughta account for some of that stuff. Seems like 538 is the most on the same page as you. By the end they had Biden slightly behind Trump.
 
You must log in or register to reply here.
Back
Top