2024 Pre-Election Political Polls | POLL - Trump would have had 7 point lead over Biden

  • Thread starter Thread starter nycfan
  • Start date Start date
  • Replies: 6K
  • Views: 183K
  • Politics 
So is there a reason why the Dems - with that billion $ war chest - isn’t ALSO FLOODING THE POLLSTER LANDSCAPE by “dropping dozens of polls into the battleground States” to skew the polling averages back to where they obviously are in reality?
If a poll is paid for by campaign funds, then that has to be disclosed when the poll is released.

What Pubs are doing isn't coming from the campaign, it's from party supporters who are willing to work for party success but outside the party structure.

It's really part of the bigger issue in that Dems have taken on the task of both trying to win elections while also upholding the overarching rules and norms of elections. Pubs aren't constrained in that way in that they've created an entire ecosystem built to convince their base of whatever alt-reality they're pushing and they also don't care about a functioning government or electoral process itself.

Pubs gain from pushing false narratives in ways that generally don't benefit Dems to nearly the same amount, so Pubs do so in ways that Dems don't.
 
Rosenberg's calm and confidence helps to keep me somewhat sane -- only somewhat.
I doubt he's right. The reason the Pubs are dumping polls into PA isn't because they think they are losing there. They were going to do it anyway.

Stop worrying about one poll here or one poll there. The NBC poll was just that -- one poll, and I think it was even registered voters.

The poll did have quite an effect on the 538 forecast, though. It went from 53-46 in favor of Kamala to 53-47 in favor of Kamala.
 
About 4 years ago I cut back on Facebook posting and that was highly beneficial. It's still helpful as a way to connect with friends but removing it from my home screen helped

I never post on Instagram so that's not an issue.

Twitter is toxic. I haven't deleted it but I don't post

LinkedIn is a necessary evil. But my God of the apps that give me heartburn it's the worst

I have never done ticket ticks or snap chat.

I feel in a way social media is going through some things as people are becoming very aware of the toxicity
I’ve been able to curate my FB feed so that I only really get soccer stuff, comedy stuff, generic non-political junk, and 99% innocuous stuff from friends. I have seen a somewhat political post from a couple of right wing friends maybe twice in the last year.

Reddit is also easy to curate. Favorite only non-political topics you like and never visit political forums. Easy.

LinkedIn, for any vaguely political post I just use the feedback to say I don’t want to see topics like that. Keeps my feed very apolitical.

I’ve never done Instagram so no idea about that.
 
If a poll is paid for by campaign funds, then that has to be disclosed when the poll is released.

What Pubs are doing isn't coming from the campaign, it's from party supporters who are willing to work for party success but outside the party structure.

It's really part of the bigger issue in that Dems have taken on the task of both trying to win elections while also upholding the overarching rules and norms of elections. Pubs aren't constrained in that way in that they've created an entire ecosystem built to convince their base of whatever alt-reality they're pushing and they also don't care about a functioning government or electoral process itself.

Pubs gain from pushing false narratives in ways that generally don't benefit Dems to nearly the same amount, so Pubs do so in ways that Dems don't.
Et voilà merci beaucoup.
Exactly the same way Dems eat their own (hello Al Franken) but pubs push dickheads like trump down our collective American throats… claiming all things being equal.
 
I doubt he's right. The reason the Pubs are dumping polls into PA isn't because they think they are losing there. They were going to do it anyway.

Stop worrying about one poll here or one poll there. The NBC poll was just that -- one poll, and I think it was even registered voters.

The poll did have quite an effect on the 538 forecast, though. It went from 53-46 in favor of Kamala to 53-47 in favor of Kamala.
He's partially right

They increase them when they are concerned. They would have done them... Just way fewer
 
Polling margin of error vs. the poll's expected voter turnout model.

Maybe someone can explain the difference? I tend to think that if the poll's turnout model is wrong then the margin of error doesn't mean anything. But I could be wrong.
 
Polling margin of error vs. the poll's expected voter turnout model.

Maybe someone can explain the difference? I tend to think that if the poll's turnout model is wrong then the margin of error doesn't mean anything. But I could be wrong.
Depends on how they compute margin of error, but my understanding is that's usually a purely statistical estimate of the uncertainty. An uncertainty due to turnout model would be an additional, systematic (i.e., this uncertainty does not average down if they do the poll over and over) source of uncertainty (that they probably aren't reporting).
 
Polling margin of error vs. the poll's expected voter turnout model.

Maybe someone can explain the difference? I tend to think that if the poll's turnout model is wrong then the margin of error doesn't mean anything. But I could be wrong.
if their turnout model is significantly wrong (see 2016) then the whole poll doesnt mean anything. I think they have overcorrected since then.
 
Yeah I'm not concerned about the red wave polls, more about the non-red-wave polls that have had concerning results (like NBC national poll showing a tie).
NBC said this morning that if dems come out strong, then it changes their model to like Harris +3. They admitted (of course) that turnout is the complete unknown.

If Dems (and dem independents) come out and vote, she will win
 
NBC said this morning that if dems come out strong, then it changes their model to like Harris +3. They admitted (of course) that turnout is the complete unknown.

If Dems (and dem independents) come out and vote, she will win
Modeling is inexact by definition, but I’m really not sure why anyone would expect anything less than high Dem turnout. Since 2016, Trump in the news, and especially Trump on the ballot, has always resulted in strong Dem turnout.
 
Modeling is inexact by definition, but I’m really not sure why anyone would expect anything less than high Dem turnout. Since 2016, Trump in the news, and especially Trump on the ballot, has always resulted in strong Dem turnout.
especially post-dobbs

I think they are just scared of repeating 2016
 
Sign Game:
Already been noted that there are way more Harris signs than Biden in 2020 (Honestly this shocks me - I thought more dems would be too wary to put things out, given the insanity of MAGA...I mean I have been). There are been noticeably fewer Trump signs than 2020 and 2016....I admit I thought the sign stuff was a weak measure of enthusiasm, but the turnout proved me wrong.

But in my drive this morning through rural Orange County (yes a very blue county...but the rural areas are still fairly reddish): 17 Harris - 6 Trump (2 Trumps were on these two homes side-by-side that have about 50 signs total between them...but they get 1 vote each). Inside the town limits it was 8-0 Harris (so 9-6 in the rural stretch)
 
Last edited:
1. Can we avoid the cock-based rhetoric, please. It's really not OK.
2. I understand where you're coming from, and when I was younger, I used to feel this way. But let me offer a different perspective, one that isn't necessarily a challenge to your view point so much as a mollifying complement.

It's important for US presidents to maintain some continuity in foreign policy, for a number of reasons. When I say "continuity" I don't mean exactly the same; I mean that the policy should gradually change instead of radically swinging depending on the administration. And one reason for this is that we can't have effective foreign policy if other countries can't trust us to do what we say we are going to do, because the administration might change and everything gets dismantled. A substantial amount of foreign policy goes to treaties, agreements, protocols and the like.

Because of Trump, our credibility is almost shot. For instance, It was a calamity when Trump withdrew us from the Trans Pacific Partnership. That treaty was a dozen years in the making, and required the US to do a lot of cajoling of some of the included nations. It was no small ask for Malaysia or Singapore, for instance, to join a trade partnership that excluded Beijing. And yet with lots of diplomacy, they finally came on board . . . and then it vanished. Those countries gave China a small middle finger and for what end? I doubt they will ever trust the US again. It shouldn't really matter whether you think the TPP is good or bad; the point is that the sudden withdrawal was damaging to the United States long term, and that damage is considerably greater than whatever harms might have come from the treaty (editor's note: opposition to the treaty was rarely based in fact, because it was not what the far left and far right depicted it as).

And of course, what Trump did to the TPP, he also did to the Paris Accords and the Iranian nuclear treaty. And he could do to NATO. Or the UN. Or any other agreement, and then what will happen to our standing in the world?

I think sometimes people don't really understand what international cooperation does for us. For instance, the US has been able to cajole most offshore tax havens into policies that make it much harder for tax cheats and money launderers to stash new money there. We got Switzerland to open its banks to US law enforcement after decades of cajoling and diplomacy. It's not in Switzerland's interests to do so, but the US was able to use carrots and sticks to make progress on that front. Not just the US, but you get the point. Banking regulations are coordinated between international countries, and financial policy in particular. One of the reasons for the 2008 financial crisis is that Bush basically stopped caring about regulatory harmonization, and thus so did Europe, and thus the financial crisis was largely birthed out of a UK subsidiary of AIG that was following UK law but engaging in ridiculously risky practices.

So broadly speaking, our foreign policy needs to be predictable. It doesn't necessarily have to be fixed, but it can't be a thing where a new president takes office and does for foreign policy what s/he typically does for domestic policy. That's something to consider when evaluating Obama.

3. Now I understand that I'm talking about concerns that don't necessarily get to the heart of your complaint. After all, I take you not to be talking about the WTO but rather Allende and Mobutu, Saudi Arabia and Iraq. And those situations do present more of a challenge to the continuity I outlined above. The cost benefit is different. And where there's no real risk to America's reputation, quick shifts aren't as dangerous. There are still other reasons why continuity should be the norm, but let's stay with this one and analyze the Vietnam War. Would it have been a problem if Nixon had decreed in 1969, we were pulling out of Vietnam immediately? I don't think so. We didn't have war partners to speak of. Nobody except the SV kleptocrats wanted us there. The spectrum of world opinion there ranged from "you should get out" to "we don't give a fuck really," but nobody wanted us in.

Let's also take the second Iraq War. Bush 43 did assemble an international coalition, but other than the UK, I don't think anyone really wanted to participate. It was more of a "if you insist, we will grudgingly contribute." Some of the "assistance" offered by some allies was comically symbolic. There were countries, IIRC, that sent less than 10 soldiers as their "contribution." So if Kerry had won in 04 and stopped the Iraq War, nobody would have minded about the reversal (which is different from a decision as to whether a sudden reversal would have been a good idea on the merits, given that we broke it and thus had a responsibility to "fix it", which of course we didn't).

But on other issues, where our national interest more closely dovetails with those of our allies and thus our allies (and non-allies for that matter) rely on us, sharp deviations from policy are highly suboptimal. Yes, if you evaluate individual policies on their own merits, some of them were gross -- but that's not the same thing as saying we should just reverse them. In foreign relations as in law, sometimes consistency is more important than getting it right.
Super - can you explain the point on the AIG subsidiary re the 2008 crisis. Not super familiar with that and if you got color, would appreciate it. (If we can do so without derailing the thread).
 
Yeah I'm not concerned about the red wave polls, more about the non-red-wave polls that have had concerning results (like NBC national poll showing a tie).
Oh yeah I hear you. Definitely understandable to be concerned. But I think it’s important to not pay attention to any individual polls, even ones that are good news for Harris. Like last week, the New York Times/Siena poll, which has not been the best one for Harris/Democrats, was really good for Harris. But it’s only one poll. The NBC one is only one poll (I know you know all of this, I’m not preaching at you or anything!). Overall in the aggregate, polling is favoring Harris, and as I said yesterday, virtually every single possible fundamental intangible is significantly in her favor and significantly against Trump. Of course, that does not guarantee that she will win or that he will lose, but it’s just to say that right now she is a solid favorite, and you would much, much rather be the Harris campaign than the Trump campaign right now, which was not the case back when it was the Biden campaign versus the Trump campaign.
 
Sign Game:
Already been noted that there are way more Harris signs than Biden in 2020 (Honestly this shocks me - I thought more dems would be too wary to put things out, given the insanity of MAGA...I mean I have been). There are been noticeably fewer Trump signs than 2020 and 2016....I admit I thought the sign stuff was a weak measure of enthusiasm, but the turnout proved me wrong.

But in my drive this morning through rural Orange County (yes a very blue county...but the rural areas are still fairly reddish): 17 Harris - 6 Trump (2 Trumps were on these two homes side-by-side that have about 50 signs total between them...but they get 1 vote each). Inside the town limits it was 8-0 Harris (so 9-6 in the rural stretch)
Northern Orange County is very red...
 
Back
Top