Nov 2023 Mod notice:
There may be other, more specific, threads about some aspects of US politics, but this one tends to act as a hub for all sorts of related news and information, so it's usually one of the busiest OTC threads.
If you're new to OTC, it's worth reading the Introduction to On-Topic Conversations
and the On-Topic Conversations debate guidelines
before posting here.
Rumor-based, fear-mongering and/or inflammatory statements that damage the quality of the thread will be thumped. Off-topic posts will also be thumped. Repeat offenders may be suspended.
If time spent moderating this thread remains a distraction from moderation of the wiki itself, the thread will need to be locked. We want to avoid that, so please follow the forum rules
when posting here.
In line with the general forum rules, 'gravedancing' is prohibited here. If you're celebrating someone's death or hoping that they die, your post will get thumped. This rule applies regardless of what the person you're discussing has said or done.
Edited by Mrph1 on Nov 30th 2023 at 11:03:59 AM
One of the flaws in your logic there, Galadriel, is that you're treat all the polls as if they were independent. Pollsters using the very similar methodology in three very similar states is naturally gonna lead to very similar results. So if they're all off by about the same amount, then that's just things working as you'd expect.
And, y'know, 538 has published a thing detailing how often their predictions are accurate
, and it turns out they do pretty well- For example, of the time they predicted something had a 55% chance of happening, their prediction came true 51% of the time. If they number weren't like that, well, that would be how you'd falsify a 'percent chance' model. But in fact that results have largely vindicated their work.
Edited by Gilphon on Oct 23rd 2020 at 9:54:45 AM
Some reassuring news:
https://twitter.com/NateSilver538/status/1319819274668605441
For context, Nate Silver talks about how one of the surest signs Clinton was in trouble were the district polls were showing her losing to Trump. Well tons of district polls show Trump losing badly to Biden, with a pretty large Biden swing to boot.
Of course I should always leave with the caveat: "Don't get complacent."
You can only write so much in your forum signature. It's not fair that I want to write a piece of writing yet it will cut me off in the mid![]()
![]()
Dewey's only half the picture, though - I don't think Trump and Truman are really comparable.
Edited by KarkatTheDalek on Oct 23rd 2020 at 10:13:43 AM
Oh God! Natural light!@Gilphon: I know, all three Midwestern states have similar dynamics. It was just such a big miss that it’s gotten in my head. I’ll wait and see whether the polls are close this year, or if the problem is more enduring.
It boggles the mind that the Republicans, in the last 20 years, have spent 12 years in office despite only winning the plurality of the vote once.
Also: About Biden's 35% chance of winning Texas: I feel like that number needs to be put into context a bit, because it's better than it sounds.
What you have to understand is that according to 538, Texas is the fourth closest state. The only states where the underdog has a better chance of winning are Georgia, Ohio and Iowa. (Although technically, Georgia and Iowa are so neck-and-neck that you can't honestly called either candidate an underdog, so really only Ohio has a better chance of an upset).
The states where Trump has a less than 35% chance of victory include North Carolina, Arizona, Florida, Pennsylvania, Wisconsin and Michigan. So if you think Biden doesn't have a realistic chance of winning Texas, understand that by the same logic, Trump doesn't have a realistic chance of winning the six states I just mentioned.
It literally cannot be understated how badly the Electoral College needs to die.
...And honestly, that should be a bipartisan position - California Republicans are disenfranchised as much as Alabama Democrats right now - and there's really no greater proof that Republicans don't support democracy than the fact that it isn't.
Edited by nrjxll on Oct 23rd 2020 at 9:52:57 AM
Republicans openly say they don't support democracy ("we're a republic, not a democracy"). There's a bit more nuance to that statement than what you might guess at first glance, but it's basically still an excuse to enforce minority rule and make it so that they get to arbitrarily decide what's best for everyone even when it's not the majority of people want or prefer.
Edited by Draghinazzo on Oct 23rd 2020 at 11:56:30 AM
The massive surges in early voting is probably going to have a big impact on the election.
Some of the numbers on early voting I'm seeing are completely bonkers. Like it's not inconceivable that we'll have over 50% of 2016's turnout before November 3rd. We might already be there since not every state releases comprehensive early voting numbers and I believe we are already well over 30%.
In addition we have the pandemic angle to consider. A surge in COVID-19 cases, which is exactly what we are seeing, this close to the election is much more likely to hurt then help Trump.
Edited by Falrinn on Oct 23rd 2020 at 7:55:50 AM
It's hard to read into the early voting numbers, because there's partisan split in them that hasn't been present in past elections. It's possible that they're a sign of good things for Biden, but it's equally possible that we'll just be looking at high turnout for both parties.
Edited by Gilphon on Oct 23rd 2020 at 11:06:52 AM
So, Faux News talking head Lou Dobbs of all people is telling his viewers from South Carolina to not vote for Lindsay Graham. And Dobbs has been a vicious racist (especially on immigration) for 20 years and a devoted Trump follower since Cheeto entered politics.
Politics is the skilled use of blunt objects.For not being servile enough with regards to Trump. The tipping point appears to be Graham not dragging the heads of Facebook and Twitter before the Senate to answer for their "crimes" against Trump.
Politics is the skilled use of blunt objects.
x2 According to The Wrap: Graham is saying that Twitter and Facebook are going to willingly testify to Congress over 'Censorship' of Right Wing Misinformation, when Dobbs (and other Right Wing Extremists) want Republicans to outright Sue them
.
![]()
Incredibly petty, yes, but if it helps Jamie take the seat, I'm all for Republicans eating each other.
Edited by DingoWalley1 on Oct 23rd 2020 at 11:55:22 AM
I know the Lincoln Project will betray us in that sudden but inevitable way as soon as the election is over but I must admit, they truly are the masters of negative campaign ads and attacks. They are the "They go low, we go lower" and it is beautiful with their recent Time Square ads designed purely to piss Trump's kids off (everyone in NYC is voting Biden already).
Edited by CharlesPhipps on Oct 23rd 2020 at 10:25:23 AM
Author of The Rules of Supervillainy, Cthulhu Armageddon, and United States of Monsters.I did get some massive cognitive dissonance when the Lincoln Project put out their ad based on the "admission" that Trump wants the SC to strike down Obamacare from the 60 minute interview.
Because they said "And there it is. Trump wants to take away your healthcare"
And I'm thinking - how many of you people were involved in the fight to stop Obamacare in the first place?
Edited by singularityshot on Oct 23rd 2020 at 11:35:24 AM

I don’t trust the polls, and not because of “percentage chance to win” prediction. (I ignore the “percentage chance” models. They’re unfalsifiable and the 538 one swung wildly during the Democratic primaries, to the point where it was basically just converting the ‘conventional wisdom’ into numbers. A pundit defining their assumptions using numbers rather than words doesn’t make them inherently more valid.)
But back to the topic of polls. Clinton was up +5 or more in three Midwestern states in the polls, and she lost all of them. That’s a huge miss. If the polls had her up, say, +2 and she lost them, well, that’s passable, though miscalling all three would still be an issue. If they missed one and called the other two fairly correctly? Okay, that’s a wide margin of error, but OTOH the margins of victory were really narrow.
But +5? Not just in one swing state, but three? With that kind of error margin, the polls aren’t really telling you anything useful. The point of state polls are to give you a sense of what’s happening in swing states - which ones lean R, which ones lean D, and which ones are really toss-ups. If they can’t do that, what’s the point in looking at them?
And yes, I’ve watched the 538 video, and I know they’ve fixed the education-weighting issue, but I don’t trust the state polls after that big of a mistake. So I’m just looking at the national ones and reminding myself that mathematically you can’t win the popular vote by 9 points and lose the electoral college. Because based on 2016, PA, MI and WI are all still too close to call.
Edited by Galadriel on Oct 23rd 2020 at 9:25:58 AM