The One-Stop Thread for Polling

1. 75

   Valeantposted 2 months ago

   Matthew Dowd, former Bush strategist, is thinking North Carolina is going to go to Harris.  His data to back that up is that more GOP is voting early, and yet Harris in exit polling is leading 50-47%.  Meaning, that either Independents are breaking heavily to Harris or there is more GOP crossover than Dem crossover.

   replyreport
   1. 86

      My Esotericposted 2 months agoin reply to this

      Sure hope that is reality.
      
      I heard from two different sources that Trump's hate-fest at MSG is hurting him with Latino's in at least Nevada and Pennsylvania.

      replyreport
2. 86

   My Esotericposted 2 months ago

   CNN had a great update which ought to make Harris feel a little bit good.  I will try to summarize.
   
   * In late Summer, Harris led Trump by 6 in WI, by 5 in MI, and tied in PA
                    Today, Harris led Trump by 6 in WI, by 5 in MI, and tied in PA
   Looking at the most motivated voters, it doesn't change much -
                               Harris led Trump by 5 in WI, by 6 in MI, and by 3 in PA
   
   * In MI, Harris gains with Kennedy in the mix
   
   * 6% of voters in WI are undecided, 7% in MI, and 8% in PA (which is good since PA is more  iffy.
   
   * On trust with the Economy: Harris trails Trump by 3 in WI and by 4 in MI (she has basically pulled even with him in those two states).  In PA she trails by 8.
   
   * On trust to protect Democracy: Harris holds an 8 point advantage in WI and MI, but only a 4 point advantage in PA
   
   * On honesty and trustworthiness Harris holds a massive lead of 17 and 16 points over Trump in WI and MI, respectively, but only an 8 point lead in PA
   
   * In putting America's interest over their own: Harris beats Trump by 10 in WI and by 8 in MI. She leads by only 1 in PA.
   
   * They care about me: Harris ahead by 8 in WI and MI, and by 3 points in PA
   
   * More people who are backing Harris say they are doing so in opposition to Trump than Trump voters being in opposition to Harris. This makes a lot of since given how many Republicans are rejecting Trump.
   
   * White voters without a college degree (Trump's main base): Trump has lost a lot of ground  in this metric.  In 2020, Trump scored 54 - 43 in WI, 54 - 37 in MI, and 61 to 34 in PA. 
     That has shrunk to 48 - 48 in WI, 48 - 47 in MI. Trump has a wider margin in PA.
   
   * For responders who have already voted, they broke 60% - 38% in WI and 61% - 35% in MI.  In PA it is a smaller spread of 57% to 40%. That would be good news for Trump until you realize that back in 2020, mail-in ballots were bad, so Trump suppressed that vote.
   
   https://www.cnn.com/2024/10/30/politics … index.html

   replyreport
3. 75

   Valeantposted 2 months ago

   NBC News is tracking the ballots returned by party registration.  It's an interesting picture:
   
   https://www.nbcnews.com/politics/2024-e … early-vote

   replyreport
   1. 86

      My Esotericposted 2 months agoin reply to this

      Quite a disparity when broken down by age group, surprising.  I would have thought the 65 and over would be higher.

      replyreport
      1. 75

         Valeantposted 2 months agoin reply to this

         When I went to early vote in the middle of the afternoon, I was expecting someone to yell out, 'Bingo!'

         replyreport
         1. 71

            Ken Burgessposted 2 months agoin reply to this

            JUST IN: The most accurate poll of 2020, AtlasIntel, releases their final battleground poll showing Trump winning all 7 battleground states.
            
            Arizona: Trump +5
            North Carolina: Trump +4
            Nevada: Trump +4
            Georgia: Trump +2
            Pennsylvania: Trump +1
            Michigan: Trump +1
            Wisconsin: Trump +.3
            
            If this is how the election plays out, Trump would win 312 electoral votes.
            
            But of course, there are all those late arriving "mail in ballots" to be found after election day in those swing states... so who knows?
            
            

            replyreport
            1. 75

               Valeantposted 2 months agoin reply to this

               Yeah, AtlasIntel had likely voters going for Walker at 49% to 45.8% for Warnock in the 2022 midterms.  Pretty sure they missed post-Roe polling pretty badly.  They ended up undercounting the Dems by 4 points.

               replyreport
            2. 86

               Sharlee01posted 2 months agoin reply to this

               AtlasIntel gained attention for its prediction accuracy in the 2016 U.S. presidential election, where it was one of the few polls to forecast Donald Trump's victory.  Also right n in 2020.
               
               AtlasIntel had a notable performance in the 2022 midterm elections, where they correctly predicted the outcomes in various races. They avoided the "red wave" narrative that many other pollsters had suggested, instead highlighting a more balanced outlook for the Democrats. In particular, their polls indicated that key Senate races would remain competitive, and they proved accurate in forecasting that Democrats would retain their majority in the Senate.
               
               They project significant Republican gains in the Senate. Like this... They are holding out on a prediction on the presidential race.​  Stay tuned.

               replyreport
               1. 86

                  My Esotericposted 2 months agoin reply to this

                  Atlas Intel, while one of the more credible pollsters has some problems with state level races.  According to a search by ChatGPT
                  
                  Atlas Intel had a mixed record in state races for the 2020 and 2022 U.S. elections. In 2020, while their national predictions were close to the final outcome, certain state polls were less accurate, with some higher-than-average polling errors in battlegrounds like Wisconsin and Florida. For the 2022 midterms, Atlas Intel continued to face challenges with state-level accuracy in a few close Senate and gubernatorial races but remained competitive overall. Their strongest performance has often been on national-level predictions, though state-specific accuracy has varied.

                  replyreport
                  1. 88

                     wildernessposted 2 months agoin reply to this

                     Have you ever noticed that no one ever quotes a non-credible source?  Everyone only uses the "more credible" source to provide their data!
                     
                     And now we have biased AI systems that we can use to agree with us on what source is "more credible" - all we have to do is use one with our own bias!  Wonderful the way things work out, isn't it?

                     replyreport
                     1. 86

                        My Esotericposted 2 months agoin reply to this

                        Who  says there biased other than those who disagree with the facts and sources they produce.  Think about it, if they were truly biased, why would anyone use them?  Does my using ChatGPT for work mean anything I produce from that input is biased?  I think not.
                        
                        You would probably love Musk's AI program because he actually programs it to be biased and give out false answers.
                        
                        I asked ChatGPT if "Trump is a good guy?"  Here is the "biased" answer it gave me.
                        
                        "Whether Donald Trump is perceived as a "good guy" is highly subjective and depends on individual values, political beliefs, and perspectives on his actions and policies."
                        
                        To me, that is the definition of "unbiased".

                        replyreport
                        1. 88

                           wildernessposted 2 months agoin reply to this

                           " Think about it, if they were truly biased, why would anyone use them?" 
                           
                           Now that answer is as obvious as it can be.  People use them because they give answers/comments that they agree with; as such they must be correct, right?
                           
                           "You would probably love Musk's AI program because he actually programs it to be biased and give out false answers."
                           
                           Your evidence for this?  A copy of the program maybe?  Not that I doubt it is biased, but what can you offer as proof that I cannot offer for ChatGPT?
                           
                           I do commend chatgpt for the answer.  Particularly as I got the same basic answer for Biden and Harris.

                           replyreport
                           1. 86

                              My Esotericposted 2 months agoin reply to this

                              That may be why you use them, but I, and anybody that I know who uses then (GA for example), use them to get honest answers quickly.  It saves an enormous of amount of time scouring the web on our own.
                              
                              The only caveat is one must make sure the answer passes the laugh test.  Several times, especially on math problems, it gives what I know to be the wrong answer. When I point that out, it apologizes and comes up with the correct answer.
                              
                              In a recent case, however, I was asking about rate of inflation and rate of wage growth.  It said something counterintuitive like the price of hamburger went up 19% while wages went up 24%. Then it said that wages hadn't overcome the increase in hamburger prices. That, of course, made no sense since 24 > 19.
                              
                              So I asked why it said such a counterintuitive thing.  It came back with a perfectly reasonable answer that while wages indeed went up more than hamburger prices, those wage increases are spread over a much wider ranges of things than just hamburger.
                              
                              If the model was as biased as you say it is, it would have figured out the answer I wanted and given me that.  Instead, it gave an unbiased answer.
                              
                              As to Grok, I found this on Wired - "Rozado says that shortly after he posted his results on X, he was contacted by a member of the xAI team who wanted to know more about his experiments. Not long after that, Musk waded in to question his methodology. “This test does not seem accurate. Some of the questions are outright ridiculous and many lack any nuance,” Musk wrote. But a few days later, he confirmed in a reply to another X post that xAI would try to make Grok less politically biased in future. "
                              
                              Then there is the fact that most of Grok's training material comes from X, itself, a hardly unbiased platform

                              replyreport
            3. 86

               My Esotericposted 2 months agoin reply to this

               I guess 538 is another site you don't visit.  There poll analysis has Atlas at #22, not #1
               
               Are you sure those are the final numbers? The ones I see for Oct 30 or Oct 31 are different in several cases.
               
               Also, what was the margin of error with those out of context numbers?  If you look at that then each of those races is a statistical tie.

               replyreport
4. 75

   Valeantposted 2 months ago

   In the 2022 midterms, the polls underestimated the Dems in MI, WI, and PA by 4 points.

   replyreport
   1. 88

      wildernessposted 2 months agoin reply to this

      Was it because of the millions of indiscriminately mailed ballots, or maybe because they were counted after the legal time frame?  I don't know - just throwing out possibilities.

      replyreport
      1. 71

         Ken Burgessposted 2 months agoin reply to this

         It was because most people don't care about mid-elections... and they failed to recognize how big the backlash was going to be for reversing Roe.
         
         Its the same reason why Harris will win this election, because they have convinced enough women that their rights and freedoms in general and Abortion in particular are on the line.
         
         You can also blame men for being apathetic and not voting... the people most effected in a negative way over the last 4 years are the least likely block to vote... middle and lower class men.
         
         The people most likely to lose jobs to migrants... middle and lower class men.
         
         The people most likely to be thrown into the meat grinder when they crank up their WWIII efforts... middle and lower class men.
         
         The people least likely to vote in this election... middle and lower class men.
         
         
         ----
         
         And then... of course... even should the voters choose Trump... there are always those Mail In Ballots... unverifiable... unaccountable... that come in AFTER the election is over.

         replyreport
         1. 86

            tsmogposted 2 months agoin reply to this

            turning page in history

            replyreport
         2. 88

            wildernessposted 2 months agoin reply to this

            *gasp*  You mean that votes received after the legally appointed time frame are counted is some areas!?!?  Tell me it isn't so!  Tell me that the fine upstanding Democrats would never attack our Democracy by violating election laws!
            
            Or...you could just tell me there is no fraud anywhere in our elections, and hope that I believe you as Democrats think I believe their tall tale about it.

            replyreport
         3. 86

            My Esotericposted 2 months agoin reply to this

            The thing is, with very few exceptions, people don't lose jobs to migrants.  If you really think they do, you should be able to find a some studies that prove it.

            replyreport
         4. 86

            My Esotericposted 2 months agoin reply to this

            You probably need to catch up to reality. Mail-in Ballots ARE verifiable and accountable, even if they come in after election day in those states that allow it.
            
            Prove me wrong.

            replyreport
            1. 71

               Ken Burgessposted 2 months agoin reply to this

               No... prove that BS right.
               
               Its not the voters that count... its the people counting the votes that matter.

               replyreport
               1. 86

                  My Esotericposted 2 months agoin reply to this

                  Didn't think you would since there are no reports of widespread (even smallspread) problems with mail-in balloting.
                  
                  But since you deny yourself real information by avoiding mainstream news, here is a video for you.
                  
                  https://abc7.com/vote-by-mail-in-ballot … d/6428995/

                  replyreport
      2. 75

         Valeantposted 2 months agoin reply to this

         So, you're throwing out unproven conspiracies.  Very on-brand.

         replyreport
      3. 86

         My Esotericposted 2 months agoin reply to this

         Since investigations over the years have repeatedly found there is nothing wrong with mail-in ballots (which Trump now agrees is true), what is the point of your criticism?

         replyreport
5. 86

   My Esotericposted 2 months ago

   Some more polls dropped and they moved the needle for the moment.
   
   By the way I calculate things (similar to Silver and 538), she "flipped" four battleground states: MI, NV, PA, and WI.  (AZ and GA seem to be lost to her.)  By "flipped", I mean that my numbers reflected that Trump held a slight lead in all four states using measures that included all polls, and 3 of 4 using just the most reliable polls and now he doesn't, at least using the most accurate polls.
   
   As I mentioned before, I keep four tallies per state. One uses all polls. the second uses polls with accuracies > 0.5, the third use 538's top 20 polls, and the fourth uses just the top 10.
   
   Remembering that '+' means Harris is up and '-' means Trump is up. In that order, the results are:
   
   MI - -0.1; -0.3; +0.1; +0.7
   
   NV: -1.0; -0.8; +1.0; +1.0
   
   PA: -0.4; -0.3; +0.5; +0.8
   
   WI: -0.1; +0.2; +0.2; +0.4
   
   1) All the numbers are within the MOE
   
   2) As the polls become more accurate, the edge goes to Harris.

   replyreport
6. 86

   My Esotericposted 2 months ago

   Trump has a problem, several of them actually. I'll just stick to the election polling, however.
   
   * The latest Marist poll has Harris leading Trump by 16 points while Trump leads with men by only 9 points (and shrinking)
   
   * There are more women voters and they are more likely to vote than men
   
   * It can't help that the AZ attorney general is investigating whether Trump criminally broke any AZ laws by suggesting to his followers that Liz Cheney should be shot, apparently by a 9-gun firing squad.
   
   * Among PA independents, Harris is now BEATING Trump on who can handle the economy better. In late October, she is ahead 55 to 40!!!  In September, Trump was ahead 49 to 45.  WOW!
   
   * A Washington Post poll nationally showed Trump's margin shrink from 9 points to 5 points (which is statistically tied).
   
   https://www.politico.com/live-updates/2 … y-00184095

   replyreport
7. 75

   Valeantposted 2 months ago

   A+ rated Selzer poll has Harris up by three in Iowa (47%-44%).  Iowa.
   
   https://www.msn.com/en-us/news/politics … ngNewsSerp
   
   The crew at The Bulwark give some deeper info into the poll and its history:
   https://www.youtube.com/watch?v=yTU0ZWFNmzs

   replyreport
   1. 58

      Willowarborposted 2 months agoin reply to this

      WOW.   Farmer's revenge on Trump's tariffs?  But also , women are coming for Trump... whether he likes it or not..
      and bringing their men with them.  I really think she will take North Carolina but flipping Iowa?   I'm here for it.

      replyreport
      1. 86

         My Esotericposted 2 months agoin reply to this

         Damn, you guys beat me to the punch.  I was in a play when I saw the news.

         replyreport
8. 86

   My Esotericposted 2 months ago

   ABSOLUTELY AMAZING!!! In Sept, it was Harris 43 and Trump 47.  In Oct. it IS Harris 47 and Trump 44.  Where is this?  Wait for it ----- IOWA!!!! This, according to the Des Moines Register

   replyreport
9. 86

   My Esotericposted 2 months ago

   NYT/Senia and Atlas dropped another set of polls. In all cases, the NYT polls were neutral (2) or favorable to Harris. Without exception, the Atlas polls were favorable to Trump.
   
   538 rates the NYT at a -1.5 while Atlas gets a -.0.8 (the more negative the number, the more accurate the poll).

   replyreport
10. 86

    tsmogposted 2 months ago

    Iowa poll sends shock waves through the nation. As others have posted it has Harris at 47% and Trump at 44%. Wow!
    
    Read what Nate Silver says it means
    
    A shocking Iowa poll means somebody is going to be wrong the Silver Bulletin (Nov 2, 2024)
    Either Ann Selzer and the New York Times, or the rest of the polling industry.
    https://www.natesilver.net/p/a-shocking … s-somebody
    
    "Her new poll? It shows the state trending even bluer, with Harris leading in Iowa 47-44.
    
    Releasing this poll took an incredible amount of guts because — let me state this as carefully as I can — if you had to play the odds, this time Selzer will probably be wrong. Harris’s chances of winning Iowa nearly doubled in our model from 9 percent to 17 percent tonight, which isn’t nothing. Polymarket shows a similar trend, moving from 6 percent to 18 percent after the survey. But that still places Harris’s odds at around 5:1 against.
    
    The poll has a reasonable sample size: 808 likely voters. Still, the margin of error for the difference separating the candidates in a poll of that size is ±6.6 points. That means in theory, in 95 out of 100 cases, the “real” number should be somewhere between Trump +3.4 and Harris +9.6 if Selzer had surveyed every single Iowa voter instead of just an 808-person sample.
    
    The theory, however, doesn’t capture all possible sources of polling error — especially the fact that the overwhelming majority of people who pollsters attempt to contact never complete the survey. And we ought to be good Bayesians here: even if Harris has a very good night on Tuesday, winning a state you lost last time by 8 points is a big ask.
    
    To give us a little more perspective, there was also a second Iowa poll out tonight from Emerson College that showed Trump leading by 9 points, close to the margin from 2020. Emerson is a firm that does a lot of herding, so you ought to account for that — they virtually never publish a survey that defies the conventional wisdom.1 However, for what it’s worth, their margin of error runs from Trump +15.7 to Trump +2.5.
    
    Get Nate Silver's latest scoop on polling analysis at
    Silver Bulletin 2024 presidential election forecast
    https://www.natesilver.net/p/nate-silve … olls-model

    replyreport
    1. 86

       My Esotericposted 2 months agoin reply to this

       I agree with Silver but need to point out that I think Selzer has been wrong only once. This may be her second one, although I hope not. If ever she needed to be right, this is it.

       replyreport
11. 86

    tsmogposted 2 months ago

    Emerson College most recent poll published today, Nov 3, 2024. Getting to the nitty-gritty they have Trump = 49% and Harris = 49% through Nov 2nd.
    
    
    
    "
        Independents break for Trump over Harris, 50% to 44%.
    
    Fifty percent (50%) of voters expect Trump to be the next president, while 49% expect Harris to win. Eighty-three percent say they decided who to support over a month ago, while 8% made up their mind in the past month, 6% in the last week, and 3% have not made up their mind.
    
    Vice President Harris has a slight edge on Trump when it comes to favorability ratings: 50% have a favorable view of Harris while 48% have a favorable view of Trump. Instead, a slight majority (52%) have an unfavorable view of Trump, while 50% have an unfavorable view of Harris. President Joe Biden holds a 40% job approval and 53% disapproval rating."
    
    A little further along . . .
    
    "Harris and Trump voters differ greatly in their trust of elections: 93% of Harris voters have a good amount or great deal of trust that elections are fair, compared to 40% of Trump voters,” Kimball said. “Trust in public opinion polls also varies by presidential choice: 54% of Harris voters have a great deal or good amount of trust in public opinion polls compared to 25% of Trump voters.”

    replyreport
    1. 86

       My Esotericposted 2 months agoin reply to this

       There is a reason for this which I will post in the Presidential Election forum.

       replyreport
12. 86

    tsmogposted 2 months ago

    The most recent newsletter
    
    The polls are close, but that doesn’t mean the results will be by Eli McKown-Dawson Nov 03, 2024 ∙ Paid at the Silver Bulletin Nate Silver's website newsletter.
    A normal polling error in either direction could give either candidate a decisive victory.
    https://www.natesilver.net/p/the-polls- … hat-doesnt
    
    The leading first two paragraphs . . .
    
    "It’s been quite the weekend for high-profile poll releases. Last night, Ann Selzer’s new poll showed Kamala Harris leading by 3 points in Iowa. And this morning we got polls of all seven major battleground states from NYT/Siena.
    
    But our forecast has been hovering right around 50/50 since mid-September. Donald Trump gained ground in mid-October, and Harris has regained just a little bit now, but it’s always remained comfortably within toss-up range. So if you believe the polls, we’re coming up on the end of the closest presidential race in 50 years. Harris leads by about 1 point in our national average — though our estimate of the national popular vote, which is mostly not based on national polls, shows a slightly wider margin than that1 — and the battleground states are even closer. Donald Trump has a 0.3-point lead in Pennsylvania, while Harris has small leads in Michigan (D +1.1) and Wisconsin (D +0.9)."
    
    Silver Bulletin 2024 presidential election forecast
    https://www.natesilver.net/p/nate-silve … olls-model
    
    "Last update: 7:45 p.m., Sunday, November 3. A fair bit of polling in the afternoon, some of it pretty middling for Kamala Harris, but also mostly from mediocre and/or Republican-leaning polling firms that tend to herd toward numbers like Trump +1 or Trump +2 and don’t impact the forecast much. A bit more worrisome is a YouGov national poll that showed the race tied — they’ve usually been one of Harris’s better pollsters. Still, Harris got plenty of strong polling from high-quality firms this weekend too, so there’s no real change in the state of the race."

    replyreport
13. 58

    Willowarborposted 2 months ago

    Ann Selzer weighs in on her Iowa polling..
    
    https://x.com/PettyLupone/status/1853169502147449005

    replyreport