Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

One pollster’s explanation for why the polls got it wrong

vox.com

Post reply on HN
Print view View original

221–230 of 565 posts

Re: One pollster’s explanation for why the polls got it wrong

#221

Post by flomo »

legitster wrote: post #7
0x53 wrote: post #3

One of the biggest issues I think is that the polling is still done by just randomly calling people. I think I agree that there is a large skew introduced by the simple fact of only reaching people willing to answer a call from an unknown number.

This isn't strictly true though, there are any number of polling methods that pollsters use these days including online polling and essentially pre-approved call databases. One of the examples in this article was an in-person poll with 70% response rate. It's probably more accurate to say it's harder to effectively weigh or average responses in a setting where there are 4 or 5 different channels you are collecting re…

Public radio once had a guest describing their online polling methodology - it was running "Win a Prize" type banner ads. So probably not very representative either.

Re: One pollster’s explanation for why the polls got it wrong

#222

Post by dragonwriter »

bpodgursky wrote: post #32
daxfohl wrote: post #16

Earlier quoted context omitted.

Before 2016, polls were much more accurate. And paper ballots were the norm. So there's that.

? A huge percent of ballots this year were mail-in, and all paper. And the polling error was worse in 2020 that 2016.

> And the polling error was worse in 2020 that 2016. reply

The votes haven't even been finished counting in 2020, and that there are shifts over time in the ratios is a well-noted phenomenon. While we're well past the point where that realistically effects the outcome, assessments of polling errors that don't effect the overall outcome may easily be premature (and those started getting made before the race was called, based on the then-current state of the count as if it was final.)

Re: One pollster’s explanation for why the polls got it wrong

#225

Post by bryanlarsen »

SV_BubbleTime wrote: post #190
bryanlarsen wrote: post #95

Another pollsters opinion: Nate Cohn says the story is about Latino voters who are a notoriously hard demographic to poll. Source "The Daily" podcast.

Did Nate Cohn's predictions this election pan out? “The final polls more or less comport with how we already viewed the race. Mr. Biden ends the race up by more than eight points nationwide — the largest lead a candidate has held in the final polls since Bill Clinton in 1996. He’s up by at least five points in states worth more than 270 electoral votes, the number needed to win. Beyond that, he’s got at least a nomin…

Somebody who got it wrong and actually admitted he got it wrong is often a credible source on why he got it wrong.

Re: One pollster’s explanation for why the polls got it wrong

#226

Post by vlovich123 »

vlovich123 wrote: post #52
bryanlarsen wrote: post #36

Earlier quoted context omitted.

People 2020 said that in 2016, yet the average income of Trump voters was higher than Clinton voters. It'll be interesting to see if that changes in 2020. Although there is something to what you say; the income gap between Republican and Democratic voters did narrow substantially in 2016.

If the 1% class leans strongly right wing, the average income doesn’t really tell you much about the highly paid middle class (or even “lower paid” upper class) as incomes are not normally distributed & you’ve got significant skew (1 billion will make more difference on the mean than $0).

Not sure why I'm being downvoted. Look at dragonwriter's post where he explains more clearly why averages are meaningless in this discussion.

Re: One pollster’s explanation for why the polls got it wrong

#227

Post by philjohn »

jonplackett wrote: post #19

When and where was the last poll that predicted an election accurately? Polling is starting to feel more like a weather forecast - presented with absolute certainty, only right about half the time.

2018 midterm polls were pretty spot on.

Re: One pollster’s explanation for why the polls got it wrong

#228

Post by flomo »

gruez wrote: post #197
ballenf wrote: post #166

Earlier quoted context omitted.

How do you know it's a pollster calling?

...because they identify themselves as such? I guess it's possible that someone impersonates a pollster, but that'd fall under the second part of my comment, which I haven't see evidence of happening.

Years ago back when I had a landline, I was excited when "Gallup" had finally called me to participate in a poll.

It turned out to be a "push poll" for a particular issue (which I agree with) where they were trying to identify personal info about myself and sign me up for something.

I suspect this was not an uncommon campaign technique, and it only needed to happen once before ppl stop answering calls from "Gallup".

Re: One pollster’s explanation for why the polls got it wrong

#229

Post by snowwrestler »

rayiner wrote: post #116

> I reached out to the smartest polling guy I know: David Shor, an independent data analyst who’s a veteran of the Obama presidential campaigns who formerly operated a massive web-based survey at Civis Analytics before leaving earlier this year Before being fired for pointing out that violent protests alienate people: https://nymag.com/intelligencer/2020/06/case-for-liberalism-... Schor’s firing is a tremendous examp…

Sadly typical of HN today that a topic that is interesting technically and politically neutral (the challenge of accurate polling) is derailed up top with an inaccurate culture war comment (there is no correlation between when a Dem condemned violent protest and whether they won their election).

FWIW, Shor’s theory about poll respondents is shared by Robert Cahaly at Trafalgar, but Trafalgar polls missed badly too in 2020 but in the opposite direction. So either that’s not actually the critical factor, or it’s not possible to control for this factor in any sort of reliable way. I find that far more interesting than Shor’s personal HR problems.

Re: One pollster’s explanation for why the polls got it wrong

#230

Post by drilldrive »

noncoml wrote: post #209
drilldrive wrote: post #206

Frankly, this has an obvious answer: The people who answer polls are not representative of the population at large. Yet this was known decently enough since 2016. But more at the heart of the issue here is pollsters are providing a service not to inform but to persuade (both the public and donors); internal-polling is actually what matters. And instead of reformulating the service with more accurate questions for tot…

> it is clear they just should be abolished I am not sure I understand what you mean. Abolished how?

What I mean to say is polling as a major industry for political campaigns is done after this election. The current people running these public polls have no place of confidence anymore: donors should not take them of value, and public opinion should ignore polling. They had their chance this time around for survival as an industry but roundly threw that away.
Post reply on HN