2024 Political Polls

  • Thread starter Thread starter nycfan
  • Start date Start date
  • Replies: 2K
  • Views: 55K
  • Politics 
CallMeTyler said:
Dems leading in multiple state races, but Harris is trailing? Nah.
Click to expand...
I'm not sure about your view as to what is or isn't "realistic." It's not that uncommon for presidential candidates to trail Senate candidates of their own party. In fact, it's vital to Dems' chances to ever hold the Senate, because red states in presidential elections badly outnumber blue ones. Jon Tester won his race in 2012, even though MT went red by a mile. Hopefully he will do that again, though it's not looking great.

For instance, in 2012, Pubs won MO for president but badly lost for Senate. Barack won NV but the Pub won the Senate seat there. He also won Maine, as did the Pub Senate candidate.

Even when the top line isn't different, Senate candidates commonly outperform their party's presidential nominee. A lot of that has to do with incumbency. And right now, the Senate candidates who are running ahead of Kamala are largely incumbents running against challengers, who might not be very well known. As the election nears and people start paying attention to the Senate, you might see movement in the Senate polling in some of these places. Especially MI (where there is no incumbent) and OH (which is really a red state now).
 
The reality is that the race is basically tied and it's remarkable that Harris even has a chance given where things stood a month ago.

I don't understand how half the country would vote for donald trump, but here we are.

We can't celebrate polls that show her up and decry polls that show her down without some actual, legitimate reason. Better to look at the averages over a several-week period, anyways (although we are fast running out of weeks in the campaign left).
 
Huh
Jamison2Carter said:
The reality is that the race is basically tied and it's remarkable that Harris even has a chance given where things stood a month ago.

I don't understand how half the country would vote for donald trump, but here we are.

We can't celebrate polls that show her up and decry polls that show her down without some actual, legitimate reason. Better to look at the averages over a several-week period, anyways (although we are fast running out of weeks in the campaign left).
Click to expand...
I wouldn’t celebrate any poll.
 
lawtig02 said:
And I’m just questioning the reliability of a model that appears to be inconsistent with the observable data. Maybe his model is great. But if it is, it’s picking up on something right now that’s not showing up in the data.
Click to expand...
There are a couple of factors that would explain the move.

First, as you note, very small swings in the popular vote projection can make a huge variance in the model given how close the race is and the built in Republican advantage in the electoral college.

Second, Nate is still applying a convention bounce bias against Kamala, although that will work its way out of the system in another week or so. Meaning that historically (although perhaps not this year) candidates get a polling bump after the convention, which fades after a few weeks. So the model still assumes that Kamala's true polling numbers are lower than her current numbers.

Third, the model considers time to the election. A Trump +4 over Biden in June is not as significant as a Trump +4 in September. It adjusts for the lack of time to change the numbers (although again, that adjustment may make less sense this year given Kamala's recent entry into the race).
 
Jamison2Carter said:
ok. i didn't quote/reply directly to you - not sure why you're responding as if i did.
Click to expand...
This is an open forum. You don’t have to directly address someone to receive a reply.

I’m just suggesting that celebrating any poll is pointless.

Not sure the reason for the hostility though.
 
I will continue to worry until after the election is over. But this is like the Heels being up 4 against dook, with the ball. It's going to take a heckuva performance from dook, and some bad plays from UNC, to lose the game.

Dems appear to be the one galvanizing new voters. The issue here is the numbers.
Dems appear competitive across several states, including ones that Pubs won last election (i.e., North Carolina).
Stock market and economy are doing well. Yes, it could always be better. But with rate cuts imminent, that is a good thing.
Who is more likely to alienate potential voters, Trump or Harris? I'm going with the former.

If you could get an honest answer from the people running Trump's campaign, I bet virtually every one of them would admit that they would rather be Kamala Harris now, as opposed to Donald Trump.
 
TarHeelGrad said:
I will continue to worry until after the election is over. But this is like the Heels being up 4 against dook, with the ball. It's going to take a heckuva performance from dook, and some bad plays from UNC, to lose the game.

Dems appear to be the one galvanizing new voters. The issue here is the numbers.
Dems appear competitive across several states, including ones that Pubs won last election (i.e., North Carolina).
Stock market and economy are doing well. Yes, it could always be better. But with rate cuts imminent, that is a good thing.
Who is more likely to alienate potential voters, Trump or Harris? I'm going with the former.

If you could get an honest answer from the people running Trump's campaign, I bet virtually every one of them would admit that they would rather be Kamala Harris now, as opposed to Donald Trump.
Click to expand...
This is a good analogy except we've all seen dook make up those four points in the blink of an eye.
 
Note on Nate's model. Those state-by-state probabilities look weird, right? How can Trump have a 87% chance of winning the election if he wins NC, but Kamala has a 90% chance of winning if she wins PA? What if she wins PA and Trump wins NC, as in 2020?

The answer, I think, is correlation. Those numbers suggest that splitting PA and NC is unlikely. If Trump wins PA, it means he's having a great night and he probably wins NC too. If Harris wins NC, she's having a great night and she wins PA also.

Still, the correlation between PA and NC seems odd to me. Uh, oh -- I'm using my eyeballs to judge the model! Exactly what I've been saying I shouldn't do. and I shouldn't. But those numbers imply that PA and NC are quite likely to go the same way (I could figure out the probability of that based on that data, but I'm rusty with my Bayes and it would take me longer than it's worth), which is weird considering that they almost never have in the past. I think they have only gone the same way in 16 and 08, and 08 was a landslide so it doesn't really count (just like 84 doesn't count, given that Reagan won everywhere). Of course, the rejoinder is "how often have the polls in both states looked almost identical" and well, that's why we have a model I suppose. Still.
 
You must log in or register to reply here.
Back
Top