How The Polls Did In 2024 | 538 Politics Podcast

[@zog382]

The title of this video should have been "Please don't fire us"

[@itsallfunandgames723]

AtlasIntel and Rasmussen Reports blew it out of the water, total accuracy. All of 538's favorite pollsters, terrible. And RCP again wins the race among the aggregators, there really is no reason to use anyone other than them.

[@haroldji7240]

You can just smell 538's bias from the breath of those podcasters.

[@FreeSpeech1959]

Rasmussen were very accurate.

[@User-r5g5f]

And they have been removed from liberal pollster aggregates 😂

[@fergalhennessy775]

they were removed for being heckin problematic and not towing the party line, ok chud? freedom of speech is not freedom from consequences!

[@WagesOfDestruction]

As was the betting market, clearly, the punters did not take 538 seriously.

[@ckq]

The polls were accurate just 3 pts off uniformly and the models incorrectly assessed FL and NJ. AtlasIntel was essentially spot on

[@platexproductions]

We know, Generally the “low quality polls” nailed it, and Ann Selzer is quitting the industry. :)

[@jasonhenry8067]

Don’t let the door hit her on the way out

[@fergalhennessy775]

she needs to be investigated and 538 needs to be investigated for platforming her. What are the odds of missing reality by 3 margins of error?

[@faithgrins]

I think it's the wrong takeaway to look at polling in 2024 and focus on the fact that the error was smaller than the average over the last couple of decades. Pollsters saw the polling error in 2020 and decided that they had to avoid another miss of that size at any cost, including incorporating some methods that they themselves did not believe to be good practice. (Thinking here of weighting by recalled vote.) They did manage to shrink the margin that they underestimated Trump, but they still underestimated him. All of that is deeply worrying.

[@jordandowland7256]

Case study in completely missing, but gathering your coworkers together to pat yourself on the back and continue to gaslight the public based on your own personal biases. 👏

[@NawiTheCore]

Some pollsters did really, the so-called "right-leaning" ones: Rasmussen, Atlas Intel, Trafalgar, Insider Advantage, etc. But most of the A+ pollsters, like the NTY/Siena, Morning Consult, ABC/WaPo, etc. were wrong by 5 or 6%. On average, they were wrong by a few points *THANKS TO THE RIGHT-LEANING POLLSTERS*

[@cliffshavage9534]

Marist college the worst wrong by five

[@jamestepera3356]

This website literally picked the loser to win. Enough patting each other on the back..... you were wrong in the end. Like Alan Lichtman....

[@Sitbear]

They didn’t do that

[@Chaparro585]

The poor and the uneducated are not going to be sitting at a computer. Waiting for an email from a poster to answer 50 question that will take them 5-10 minutes. Find a better way to reach this group of people. I think it's the way to start

[@cliffshavage9534]

Atlas intel ,Emerson and rasmussen nailed this election

[@mikewilliams6025]

Industry With Proven Poor Record with Math Does Self Assessment: According to Our Math, Our Math Looks Good

[@rbu2136]

Everyone I speak to at my office in DC has a very, very negative view of polling. Whenever I bring it up I am met with scowls and very negative comments. Beyond untrustworthy.

[@yagodarkmoon]

This nobody says: Hi I'm here.

[@jamestepera3356]

All these folks have to do is look at right wing pollsters. Rasmussen and Quantis. Atlas did great (again) but Im yet to understand how theyre considered right wing. By the way Marist is terrible and if 538 rates them A+ again then the joke is on Elliott and Co.

[@mourka01]

Yall didn't wanna talk about how people actually thought texas was going purple?

[@Sitbear]

It really doesn’t matter how well you did, people will always say you did terrible because they like narratives and vibes rather than intellectualism and empiricism.

[@haroldji7240]

Yeah. It is just a matter of expectation. As a student who literally failed 2016 AND 2020, getting a C minus in 2024 is SIGNIFICANTLY good🤣

[@mathesonrittenhouse]

1.6k views & 200k subscribers... Hmmmm... Edit: and < 30 comments.

[@0o0ification]

I now believe Trump's promise on reversing any TikTok ban was quite the game changer. I had no idea that he had made such a promise there (i.e. old and square), but I do believe that it's single-issue political pandering to a sizeable, otherwise-disparate, and motivated bloc of voters.

[@Account.for.Comment]

Kidding right? Trump base don't care about Tiktok. Tiktok is used mostly in America by females. Trump did not gain much vote. Biden/Harris lost them on their own trying to flip Trump voters by producing Trump policies, trying to be Republican-lite.

[@damocles5968]

This is the definition of Cope lmao

[@tylerhackner9731]

Atlasintel and Rasmussen did it

[@cliffshavage9534]

Pollsters are playing a long long game 24.00 this video is an excuse for saying we are underpolling trump on purpose but after he retires from politics we will bring accurate polls

[@WagesOfDestruction]

You were wrong until just before the election when things suddenly improved. It's very suss. Even though the last result was better, you can still ignore all the wrong results you released before that. 538 purposely ignored polls that showed Trump ahead. The most reasonable explanation is that the 538 indicated they were politically motivated to help Harris.

[@moniquemaciel1059]

Polls did great?? Hahajahaha shut up they cheated big time.e that s why Donal teump used Elon musk or whatever his name is😂😂😂😂