Urban75 Home About Offline BrixtonBuzz Contact

US election 2020 thread

Ha, I was mildly concerned about Nevada given everything else but Las Vegas has reported and suddenly it's all fine again
 
The 538 explanation will be a piece of writing for the ages. Yet still he'll be the gospel in 2024 on Trump (Ivanka) vs Clinton (Hilary)
tbf it's not just Silver - it appears to be almost the entire polling industry, without which he has nothing else to go on other than vibes

The vibes people appear to have been slightly more accurate at this stage
 
tbf it's not just Silver - it appears to be almost the entire polling industry, without which he has nothing else to go on other than vibes

The vibes people appear to have been slightly more accurate at this stage

High turn out fucks polling models, we saw the same here.

Armies get criticised for preparing to fight the same war as the last one they fought, polling companies do the same thing though.
 
Significantly more people voting for trump than did four years ago i just cant get head around that. They didn't suddenly turn racist.
 
High turn out fucks polling models, we saw the same here.

Armies get criticised for preparing to fight the same war as the last one they fought, polling companies do the same thing though.
Yes exactly. Pretty sure I made the same point waaaay upthread somewhere - they fix the last problem. And encounter new ones. And high turnout being universally good for Dems was an assumption that held water until it didn't...
 
tbf it's not just Silver - it appears to be almost the entire polling industry, without which he has nothing else to go on other than vibes
I'm not sure people should be writing the polls off completely (except for FL that was a miss). GA still looks to be a toss-up, the margin in NH was only slightly too Biden, polling for NE-2 looks to be pretty good. Could end up being very wrong but it is also possible that by the time the final, final tallies are in there is less error than the earlier data indicated.
 
I'm not sure people should be writing the polls off completely (except for FL that was a miss). GA still looks to be a toss-up, the margin in NH was only slightly too Biden, polling for NE-2 looks to be pretty good. Could end up being very wrong but it is also possible that by the time the final, final tallies are in there is less error than the earlier data indicated.
Could be. Margins can be tight - I think the miss on the national vote share last time was half a point too high for Hillary and 1.5 points too low for Trump. They weren't that far off, whatever people's perception
 
The 538 explanation will be a piece of writing for the ages. Yet still he'll be the gospel in 2024 on Trump (Ivanka) vs Clinton (Hilary)
To be fair, if Georgia does go blue (just two points in it now - 350,000 votes left to count and 120,000 behind), and Biden wins the remaining swing states that he's favoured in, then it's a pretty comfortable win for Biden - 307 electoral college votes to 231. That would make 538's predictions pretty bang on. Even if Trump does pull it out of the bag, 538 always said that there was a 12% chance of the polls being wrong and Trump winning. There are certainly a few other pollsters and modellers that will have questions to answer though...
 
The polls appear to have been 7 or 8 points off in Ohio, which doesn't bode well for Pennsylvania.
Ohio counts votes up until the 14th. I'm unsure how different news outlets are accounting for those late votes in their % complete tally but it is quite possible that they are excluding them so there could be a movement there.

EDIT: Though OH still odds on to be a bad miss I agree
EDIT2: And polling for MN looks to have been pretty decent.
 
Last edited:
To be fair, if Georgia does go blue (just two points in it now - 350,000 votes left to count and 120,000 behind), and Biden wins the remaining swing states that he's favoured in, then it's a pretty comfortable win for Biden - 307 electoral college votes to 231. That would make 538's predictions pretty bang on. Even if Trump does pull it out of the bag, 538 always said that there was a 12% chance of the polls being wrong and Trump winning. There are certainly a few other pollsters and modellers that will have questions to answer though...

Yep, fair point. I just think it renders his whole industry irrelevant or fundamentally unfit for purpose. Its the 'sell' that they had the worlds best data scientists working on this (which is very real - you can tell by the visualisations if nothing else). But they were way off. Again. I guess they only reporting/aggregating/weighting polls, which in essence are relying on what people tell them (which were all fairly aligned). Similar to the rating agencies post 08 crash.
 
I spent some time in non urban michigan and this is a reminder of the world view they have.


When i was there i tried to find a bureau de change. No one know what I was talking about. I even used the term money exchange.l It wasn't a thing banks even thought of. I was living with my girlfriend who worked in a bank as a loan supervisor and even she did not recognise a money exchange as a thing I might be interested in.
I will remind you this state shares a border with fucking canada.

I know the terms Bureau de change wheshel and cambio. even though the euro has rendered most of those terms moot.
american is amazingly insular.

don't even get me started on the racism
 
Wtf is all the count pausing for? They have the completed votes, no? Have you all got something better to do or not want to lay on rotating staff? (I think postals can still be received for another few days in PA?)
 
Back
Top Bottom