Pollster Ann Selzer ending election polling, moving ‘to other ventures and opportunities’
https://eu.desmoinesregister.com/story/opinion/columnists/2024/11/17/ann-selzer-conducts-iowa-poll-ending-election-polling-moving-to-other-opportunities/76334909007/
32 comments
Imagine being so wrong about America it destroys your faith in your profession.
I’d be fine with ending all polling. It’s almost never right and doesn’t serve any real purpose.
Breaking News: Person fucks up and changes to a new job from a dying profession.
She had the single most discussed poll of the election cycle, and one of the worst.
I think going into other non public opportunities may be wise. She ‘ll be wearing this albatross for awhile.
When that Iowa poll came out with Harris leading and everyone was believing in it, I was gobsmacked that people thought it was accurate when it was probably the first poll of the entire election cycle where she was leading in that state, which ceased being a swing state years ago.
Worse, someone on Twitter said it wasn’t accurate and people needed to stop jumping to conclusions that she was going to win in a landslide… and he was met with a mean-spirited body shaming joke that got 100k upvotes on r/MurderedByWords. Unsurprisingly, he was right, the poll was bullcrap.
Literally one poll made everyone celebrate. ONE. After months and months of “don’t listen to the polls” ONE poll made people jump for joy, simply because it was beforehand a very trusted and often accurate poll.
All polling is now useless. Just watch the betting markets which had trump up pretty big. There was post-election debate about how the Dems lost so much of the young vote. While there’s numerous reasons, we can’t ignore the fact that an entire generation of young males who have grown up in an age of online gambling jumped on the betting trend and voted where there money was.
There’s an early ‘big name’ person the history of analytics – George Box – who’s quote I’d like to share.
‘All models are wrong, some are useful’
It’s an impossibility to ‘never be wrong’, she was bound to have this happen one day – it’s a matter of odds over time.
Her method was to basically not weight a sample at all and simply call every sampled number till they picked up. As long as the sample of people who will eventually pick up look like the sample of actual voters, this works. And it did, quite well. For a long time. The alternative is to up- and down-weight the people who respond to polls from groups that are under- and over-represented in answering the phone.
Her polling method was excellent (but also hard to pull off) until it wasn’t, so she was right to call it. She’s a smart person. She knows when to call it.
Bye , Bye
I think the more likely scenario, is that at her age, she was on the way out anyway, so she chose to air a poll that could give democrat voters optimism and drive turnout. She knew she was promoting a bad poll, but for reasons she thought were virtuous.
I will point out that she says in the article that this was always going to be her last poll…she isn’t quitting because of her results, this was her planned exit since 2023.
In all likelihood, it’s the Census data she used to weight responses that’s fucked up, not her methods. Does no one remember how chaotic the 2020 Census was between the political interference and Covid disruptions?
Maybe asking people who thrive on a right-wing media sphere of blatant lies to truthfully respond to a random stranger on a land line isn’t reliable anymore!! Maybe just stop all this nonsense and noise and let people vote and call the winner when ALL votes have been counted.
Way to go out on a high note Ann….
I’m curious what data she used that was so off
Yep
Lmfao
Has she commented on her thoughts on why her poll was so far off? And possibly why all polls were so far off?
But we had the keys! How did we lose when we had the keys!!
I’d like to announced that I will no longer be trying to woo Kate Upton and I’m moving onto other ventures.
Why is American polling so bad compared to the UK
I got discord into oblivion for saying that he steak of accurate predictions was most likely driven largely by luck. If you have dozens of pollsters out there all using equally accurate methods, some are going to have steaks like that purely by chance.
Another one who found out this is *exactly ‘*who we are’
When you look into how much behind the scenes work the pollsters do to make the data they collect more reflective of the demographics that they *think* will be the final layout, it’s easy to see how someone could get disillusioned from listening to them altogether.
She took the L
A swing and a miss…
Me before the election: If Selzer is right and this is a Harris landslide, she will be the most famous pollster of all time.
Me at 9:00pm eastern on Election Day: “If I ever see Ann Selzer again, it will have to be at a Kroger in Iowa.”
i remember when someone posted her poll before the election in all caps like a banshee screeching like it meant anything.
People were so ready for a Harris win because of anything that aligned with their views. I keep telling you guys, you can’t discredit all the other polls. Good thing the polling industry will pretty much be dead now so we can focus on actual candidates that actually represent the working class like Sanders.
Sanders won’t run for president again, but definitely someone like him will succeed.
She sold out her reputation for a fleeting dopamine high with Harris supporters. Not a good trade.
Out with a big fat thud.
Hopefully that dork with his so called “keys” does the same. Still out there making excuses as so why he was so unbelievably wrong
This may be the last real election Americans get, in which case, who needs pollsters?
Comments are closed.