I’m deliberately starting this thread on election night, before the final results are in.
But the NY Times is currently showing Trump leading in ALL the battleground states.
except Arizona and Minn.
So it looks like a Trump win.
When all the polls told us it was 90% sure to be Biden.
And the pollsters promised us that this time they are fixing the mistakes they made in the last election.
“Oh, well, last time, we just sorta underestimated the white male vote, so this time we’ve corrected it by changing the rules of our statistics.”
What excuse will they have for the next election? “We only got it wrong by 70% with Clinton, and 90% with Biden, but this time, we’re gonna get it right.”
If Trump wins the polls will have been off even more than last time in his favor. So yeah, 538 can go take a flying leap. Silver tried to temper things by saying Trump still had a chance but we’ve now lost both 3-1 and 9-1 projections.
The simple fact is that this is what you needed to steel yourself for when the country embraced early voting. Pennsylvania hasn’t even started counting them yet. Michigan and Wisconsin haven’t posted any significant early votes yet.
The map looks exactly as it looked in 2016 with a few exceptions, notably Arizona, so don’t start tripping balls about it.
Incidentally, as it stands now assuming Florida does in fact go to Trump he still only has a 1 in 3 chance of winning per 538. But you don’t trust them anymore, so we’ll just have to wait and see. Common sense should tell you that a big lead won’t stand up in any of those states, polling or no, simply because they were razor-thin last time.
Could it happen? God help us, yes. He could be re-elected. But that was always a possibility no matter how little sense that actually makes to you and me. But for now, just relax.
I don’t think it’s 538’s fault, it’s that the polls that we and 538 rely on seemed to have had a systematic bias towards Biden
I don’t get it. Pollsters have one job: try to get a representative sample of people who will likely vote. How they screwed up in 2016 and 2020, and both times biased in the same direction, is very strange
What I wonder about is internal polling, on which candidates base the entire national strategy. If the public ones are off, what does that mean for the internals?
It looks like there’s a better than average chance that Biden will very narrowly squeak this out. Arizona and Nevada are very close to in the bank. Georgia is not lost yet but it will be razor thin either way. Wisconsin and Michigan are looking close but positive and Pennsylvania is a bit of a black box, but no reason to think it’s going Trump necessarily. The win in Omaha very possibly could be the decisive vote.
So, if it does play out to be Biden with narrow wins in WI, MI and PA. 538 and their ilk will proudly proclaim victory, conveniently ignoring the fact that they released a million polls saying that Biden would win every contested state on the map going away. I’ve gotten into many Reddit fights over the previous week telling these people that their victory laps were premature and the polls could not be trusted. I got shouted down every time.
Should people trust polls? Never, ever again. This was completely obvious after 2016 but people are stupid and the media companies are lazy creatures of habit.
But, don’t blame the pollsters necessarily. Somehow Trump and his dark web trolls seem to have found a way to convince a huge swath of the populace that pollsters are the enemy. They are actively lying about their intentions to undermine the process. It’s active measures, hell, even this could be a fog or war tactic from Russia. We don’t know, but clearly the methodology has been hacked. They aren’t bad at their jobs, unless you count them being slow to understand that the landscape changed underneath them and the sky is no longer blue.
The internal polls are almost certainly no better. They use the same people and the same methods. Biden and Harris were going to Texas and Florida in the last week, they drank the koolaid too. They weren’t going to Georgia and North Carolina.
Nate Silver can go suck a big one after this, even if Biden does manage to eek out a victory. 538 has pretty much been coasting on its Obama-era predictions, but seems to be woefully unprepared for polling in the Trump era. I’ve read reports of Trump supporters lying to pollsters about who they were voting for, just to fuck with them. I wonder if that’s something the polls took into consideration.
Maybe the constitutional amendment which abolished the electoral college can also have a provisions allowing for wrong pollsters to be publicly flogged?
Nate Silver, while never giving Trump a high chance of winning, did emphasize repeatedly that there was a high(er) chance for the race to be interesting like it is now. Also yesterday, they showed that even if the polling error is the same, or even greater, towards Trump in 2020 as it was in 2016, Biden would still win, but very narrowly.
But yeah, the results could look better for them right now though. We’ll see in the end.
Why do you? I keep saying I’ve never been polled once. Yeah, yeah, I’m one person, but it’s always, ohhhhh, Nate Sillllllver (wank, splooge)
Please. I was told and chided for saying both WI and PA would go red. Guaranteed. I think I’m right. Hopefully, I’m totally wrong, but some sports geek who owns a website is as trustworthy as Rush Limbaugh.
Who are they asking in these polls?? Who is telling the truth? I mean, if I personally polled 100 million registered voters, I’d at least figure on 20 million lying to me.
I still don’t understand internal polling. .
I started a thread about internal polls a couple weeks ago,
But it didn’t really answer my question: how and why are internal polls different than the public polls?
They are carried out by the same companies (Gallup, etc), using presumably the same techniques.
Why do we often hear that the internal polls were more accurate than the public ones?
If it is so easy to be accurate for internal use by a specific party, why can’t the pollsters be accurate for the public?
Yes, the internal poll might ask “secret” questions that are designed to expose a weakness in the party, which they don’t want to reveal. So they order (and pay for) the poll. (maybe, I dunno, targeting very local issues or ethnic groups).And then the party might try to improve in that specific area, without letting the public know.
But most of the polls we see are done by large,objective organizations such as news agencies, Gallup, etc…And those agencies have as much of an interest in getting to the truth as any political party which orders an internal poll.
The polling agencies make their reputations–and their money!- by being accurate…So even if they are under contract with a political party and have a non-disclosure agreement, they could still run their own polls, at their own expense, and try to get the maximum accuracy possible. It’s in their own interests.
For example: right now it looks like the famous “90%” chance of winning for Biden is so totally wrong that the public may give up all interest in polling at all. I’m certainly not going to believe them in the future.So if the polling companies want to avoid bankruptcy, they’d better improve their results.
(or maybe our society has changed enough over the past 5-10 years that it is now impossible to do accurate polls. (maybe because so many people refuse to answer phone calls these days?.)