My ex is a data scientist who literally designs polls and teaches experimental research. Because of the difficulty in reaching people sort of randomly these days, it's statistically impossible to reliably control for error (according to her). Almost every poll out there is using some sort of novel formula to try and statistically control for the error, and because they are in the poll business, they will all claim that their methods work. But the fact that no one has a landline anymore, and people don't answer unknown numbers on their phones or respond to solicitations via text means that most participation is subject to statisticall uncontrollable bias. I don't know how this election is going to go, but I don't think the polls are telling us. Based on what we've seen in previous elections over the last couple of years, I'm going to guess it skews towards Harris, but there's really no way to know. However, it might turn out to be an unexpected trouncing one way or the other
The only people that are answering political polls are weirdos. Like you said, who answers unknown numbers on their cell phone in 2024?
Have the GOP flipped the script with early voting? Maybe. I'm sure they've improved. But this is like night and day..
Yeah I've been reading a lot about the polls and how they are trying to not have a repeat of 2020 where they were off by about +4 in Biden's direction. The polls are trying hard to reach more Trump voters and also applying weights to increase the Trump poll result. I just hope they've gone too far in the Trump direction in an attempt to avoid what happened in 2020. The one thing I keep coming back to is the massive difference in fund raising between Harris and Trump. Harris outraised Trump $378 million to $160 million in September alone. That's got to mean something.
That's interesting. The polling has overwhelmingly shown a toss up of a race. I wonder if they're herding their data towards toss up to try and mitigate the likelihood of being wildly wrong.
A great example of why I come to Too Hot. You just never know the unique insights that can be found here. This post is very subtlety quite informative.
My father used to be a ORSA for the military and he keeps saying the same thing. The people running these polls are having to make too many assumptions so they can’t possibly control for everything they need to. I listened to one pollster talk about how they were attempting to control for things it was by using a combination of demographics and door knocking for data.
Close elections, huge shifts in voting methods and timing, unusual candidates, and polling error issues. Might as well flip a coin.
There are polling firms that exist now just to try to move the aggregators … you gotta start asking yourself what the point is anymore. I realize people want to know things before they happen but there seems to be diminishing returns in the polling business. Candidates pay for their owns polling anyway so it just seems like entertainment and content generation now.
I put more stock in this than in any poll. In every state across the country, more people donated to Vice President Kamala Harris than to former president Donald Trump. https://www.washingtonpost.com/elec...ip-code-map/?itid=hp-top-table-main_p001_f005
When you look at how badly the 2020 polls undercounted Trump voters, it is really hard to be positive about Harris chances 2020 Presidential Election Polls: Biden vs. Trump - 270toWin 2020 presidential election results The national votes was about 4-5 pts undercounting Trump and 2.5-3.5 undercounting in key swing states. Biden was +8 nationally and barely won the EC in a few key swing states. Now the polls have Harris and Trump neck and neck. I am sure the polls have tried to correct for prior misses and God knows how much that is factored in. The cause for hope is that since 2022 the D’s have greatly overperformed expectations, and the repeal of Roe probably has a lot to do with that.
Agree. Donations from small donors are one of the better metrics of measuring enthusiasm or likelihood to vote. In 2016 and 2020 Trump received more donations from small donors than either Clinton or Biden and ended up outperforming the polls in both of those elections. The reverse is true in 2024. I guess we will find out if that metric is still predictive when the votes are counted this time.
I was called in either 2008 or 2012. They went through a whole list of questions. The very last question they asked was do I shop at Walmart, and offered answers of never, once a month, or once or more a week.