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

OT: American Politics

CH1 said:
I see these Presidential polls as largely entertainment and geek value. They’re useful for trend spotting (ie Kamala >> Biden) but their margin of error is greater than what we are likely to see decide those important swing states in November
Click to expand...

Averages can definitely give high quality polls even more accuracy though. It's not the philosophy or the science of it all that's the problem. It's how this is put in practice and then the media using is as a way to judge the horse race, and they do it completely uncritically of the quality of the polling.
 
And remember 5-6 weeks ago when I said “it’s still very early”

Well, it’s still early.

Rich Tau specializes in swing voters and this is what he had to say the other day.

>>Rich Thau, who runs focus groups among key voter segments, convened nine Hispanic voters in battleground states who said they disapproved of both candidates in late August. Eight hadn’t watched Harris’s convention speech, and none had watched Tim Walz’s. “The people who are disproportionately likely to determine the outcome of the election are those paying the least attention,” he says. And among this group, Trump was the candidate most commonly identified as the change agent, not Harris. As far as her campaign messaging trying to reframe the race goes, Thau says, “a lot of this stuff is not penetrating yet.”

TL;DR people like us following every twist and turn are outliers.
 
Preston said:
In other words, unless Kamabla absolutely dominates and the absolute numbers we see from the polling are historically heavily skewed to favor Trump, they were fucked with Biden. He was gonna lose.
Click to expand...

It's kind of unknowable tbh. The polls over the last 2-3 weeks before Biden dropped out were influenced heavily by the DNC drama about forcing him out. A few weeks prior, Biden's polling was in pretty good shape and didn't flinch after the debate disaster.
 
MindzEye said:
It's kind of unknowable tbh. The polls over the last 2-3 weeks before Biden dropped out were influenced heavily by the DNC drama about forcing him out. A few weeks prior, Biden's polling was in pretty good shape and didn't flinch after the debate disaster.
Click to expand...
😬
 
MindzEye said:
Averages can definitely give high quality polls even more accuracy though. It's not the philosophy or the science of it all that's the problem. It's how this is put in practice and then the media using is as a way to judge the horse race, and they do it completely uncritically of the quality of the polling.
Click to expand...

I’m not saying it’s not a problem. In fact, covering elections to make them seem like a horserace (which over 95% of the media does) has been a longstanding beef with me.

It undermines democracy in a variety of ways because it emphasizes personal quirks and campaign momentum over policy.

Trump’s viability is the most awful and extreme example of the horse race dynamic
 
Last edited:
Preston said:
😬
Click to expand...

Meh, say what you want. The polls were what they were, the disaster debate happened, polls didn't move. Polls started to suffer 1-2 weeks after the party turned on him. The polls the day before he left and a month before he left had a 3-4 point spread and it wasn't the debate that changed things with voters
 
zeke said:
Let's remember that all the polling averages - even Nate's - have the Dems winning. That's even including all the possibly low quality polls.

It's only a couple of projection models that have them losing.
Click to expand...

Yeah. I am not worried.

Though I thought Nate recently gave Trump a thin electoral college win
 
From what I've seen, the methodologies used by a lot of these pollsters are mindblowing in a hilarious way. There will naturally be confounders in all of these but there are ways to lessen them. Without looking too much into it, I question whether many of them are actually data people.

All I see are small samples and poor methodologies. And if there are a few good pollsters, that's great but they'll still all suffer from small samples and some bias that is nearly impossible to adjust for. I agree with ch1 in that the biggest things we can take from them are the trends.


And as for Nate, yeah beyond using the poor pollsters, he's doing the convention bounce adjustment thingy. Which likely doesn't make sense in this cycle. But we'll see.
 
  • Like
Reactions: CH1
I have no idea how any of these pollsters will be able to capture all the young Dems that are registering in large numbers for the first time in order to vote Kamala

it doesn’t happen under low ceiling Joe and I don’t see it happening with Trump
 
MindzEye said:
Meh, say what you want. The polls were what they were, the disaster debate happened, polls didn't move. Polls started to suffer 1-2 weeks after the party turned on him. The polls the day before he left and a month before he left had a 3-4 point spread and it wasn't the debate that changed things with voters
Click to expand...
I know I know, I'm mostly joking.


Mostly. The debate was an honest to god catastrophe, as are the majority of his public appearances. He's cooked. No one was ever gonna be enthusiastic about getting out to vote for him which could be the difference.
 
Preston said:
I know I know, I'm mostly joking.


Mostly. The debate was an honest to god catastrophe, as are the majority of his public appearances. He's cooked. No one was ever gonna be enthusiastic about getting out to vote for him which could be the difference.
Click to expand...

It was a trade off with Joe. I agree with the "low ceiling" bit. But his biggest demos are two massive voter demos, olds and organized labour, his third is black women. I haven't looked at the crosstabs in a bunch of weeks because Kamala has made them matter a lot less to my personal level of zen, but I do remember seeing a big shift in olds between Joe and Kamala, which was something I was concerned about during the few weeks the knives were out. For obvious reasons Kamala is doing just fine with black women, and she's done an awesome job catching any stray labour votes over the last month as well (the Walz pick helps a lot there). She's made up the lost olds with younger voters and then some (a demo Joe struggled with by democrat standards)

As for the debate...it was a much bigger media disaster than it was anything else. The "he's too old" polling numbers didn't move, his national numbers didn't move, nada. imo, it was baked in to voter expectations. They knew he was ancient and to his coalition, it didn't matter much.
 
You must log in or register to reply here.
Back
Top