One pollster’s explanation for why the polls got it wrong • Hacker News Board

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

#321

Post by jeffbee » Wed, Nov 11, 2020, 2:15 AM UTC

Retric wrote: post #112

orangecat wrote: post #89

Earlier quoted context omitted.

Where does the "elite" factor in? As a rough approximation, almost everyone in your social circles having a college degree.

Except that’s really not what the statistics show. Republicans gain a majority of support from high school graduates, individuals with some college experience, and college graduates. People without a high school degree or with a post graduate degree favor Democrats. An individual’s likelihood of being a Democrat decreases with every additional dollar he or she earns. Democrats have a huge advantage (63 percent) with…

None of those statements is consistent with the Pew Research data on partisan identification among the electorate. E.g. https://docs.google.com/spreadsheets/d/1v58Nt6SW832ZUq7xLI6N... for 2019 demographic breakdowns.

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

#322

Post by dan_quixote » Wed, Nov 11, 2020, 2:21 AM UTC

evo_9 wrote: post #104

Who replies honestly to political question to random strangers? I know I don't. What's the benefit of letting anyone know where you stand politically at this point? None. It's much more likely - no matter which side you're for - that the person asking could have an adverse reaction. Things are just way too polarized now, not sure how that can change, but that's how it is. I mean outside friends/family I simply take a…

Funny, I realized that I've never thought about how I react to polls until reading your comment. I flat out refuse to answer polls regardless of who is asking. I refuse to leave any public artifacts of my political feelings (signs, bumper stickers, facebook comments, etc). The reason is exactly as you describe - I have nothing to gain and potentially something to lose.

I am a problem for pollsters and there's no reason to think I'm unique in this regard.

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

#323

Post by shajznnckfke » Wed, Nov 11, 2020, 2:22 AM UTC

slg wrote: post #271

kjeetgill wrote: post #188

Earlier quoted context omitted.

Do you think the bulk of voters for Trump believe one of these things? He won a considerable number of votes.

To put some numbers to it here are results from a Morning Consult poll[1]: >Republicans, in particular, have been drawn to QAnon, with 38 percent of GOP adults who had heard of QAnon saying they believed the claims were at least somewhat accurate, compared to 18 percent of Democrats and 16 percent of independents. Strangely enough Democrats and Independents were more likely to have an opinion of QAnon than Republican…

I don’t find it too surprising that more Democrats have an opinion of QAnon. There’s a small segment of people who actually have consumed the material and believe in it, and a much larger set who have heard of it and have a vague (and possibly wrong) notion of what “side” it’s on. I think due to the repugnant nature of the movement, most of that commentary is going to come from the left, so mostly left-wing people will form an opinion. Most of the people who are favorable toward QAnon in those polls on both the left and right probably have little idea of what it actually is.

See: https://slatestarcodex.com/2020/05/28/bush-did-north-dakota/

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

#324

Post by hackinthebochs » Wed, Nov 11, 2020, 2:22 AM UTC

flowerlad wrote: post #298

hackinthebochs wrote: post #268

Earlier quoted context omitted.

>Why else would most Trump voters come from rural areas? Because the political and cultural interests of rural areas are different than the interests of urban areas? The economies of rural areas have crumbled. But 50% of the population live in cities and so economic policies that benefit the "majority" of the population will completely miss the issues rural areas face. Trump's support is because he at least pretends…

> pretends to care about them. And that's why I said: Not bad people, but gullible people.

You're missing the point. It's not that they're gullible, it's that one side ignores them completely and the other side is speaking to them. It doesn't matter that the guy telling you what you want to hear is probably lying, its still a better spot than the guy that ignores you because they don't want or need your vote.

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

#325

Post by Green_man » Wed, Nov 11, 2020, 2:26 AM UTC

nxmnxm99 wrote: post #306

snowwrestler wrote: post #229

Earlier quoted context omitted.

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…

What? Trafalgar was one of the only polls calling a close race. In Trumps favour but close nonetheless. IIRC they were the only ones calling Florida going red. You’re being intentionally disingenuous saying they were equally as bad as the rest of the field predicting a blue wave.

Trafalgar almost certainly got more states wrong (counts aren't all finished, but we do know likely winners in all states). Other sites like 538 predicted all states correctly, except NC and FL (50.5%/48.8% and 50.9%/48.4% respectively, both predictions wrongly favoring Biden [1]). Meanwhile, Trafalgar's polls put Trump up in NV, GA, PA, AZ, and MI (from their website, all late Oct / early Nov polls [2]). Once counts are final, we can more accurately compare margins and errors more specifically, but Trafalgar does not have a good record for 2020. Comparing models to a single pollster isn't completely even, but it's important to note that the Trafalgar polls were mostly outliers, disagreeing with the consensus in many states [3].

[1]https://projects.fivethirtyeight.com/2020-election-forecast/

[2]https://www.thetrafalgargroup.org/

[3]https://projects.fivethirtyeight.com/polls/president-general... https://www.realclearpolitics.com/epolls/latest_polls/

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

#326

Post by ab3rC1te » Wed, Nov 11, 2020, 2:27 AM UTC

It's called cheating, fake news, and whatever else you want to deem it as. That's why the Postal service got a 10 billion stimulus package. Just like COVID is so dangerous that it jumps 26 ft in cold air, punches your plexiglass cause it wants to attack you, and put that mask on, it can't go through that either. Even though it's air borne and been air borne since it was released. Just look up and see what's being sprayed in the atmosphere to rain down COVID 21 on you. When will people stop being idiots to the bullshit lies that they feed you.

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

#327

Post by boomboomsubban » Wed, Nov 11, 2020, 2:28 AM UTC

Lazare wrote: post #291

boomboomsubban wrote: post #220

Earlier quoted context omitted.

So you're defending the person who is trying to explain why polls got it wrong by using polling data to show that their argument belittling the BLM protests was correct. This is ridiculous on it's own, but also ignores things like how wrong it is to blame the protests for violence often instigated by the police, how dumb it is to trust polling data from the Nixon campaign when we know he was involved in all sorts of…

> show that their argument belittling the BLM protests was correct. David Shor has not, to my knowledge, ever made an argument belittling the BLM protests. I welcome any links you have that might demonstrate otherwise. > how wrong it is to blame the protests for violence often instigated by the police Neither Shor, nor the researcher he linked, Omar Wasow did this. > how there is evidence that the protests strongly h…

>You have constructed a series of strawmen.

I was responding to a post that said "It’s conventional wisdom now that failure to condemn violent protests earlier cost Democrats votes, including among minority groups." So if Shor's point was not about such topics, it has been subverted to be used about such topics.

I'll admit, he does not seem to have belittled the protests directly though. I may have went to far with that one.

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

#328

Post by rayiner » Wed, Nov 11, 2020, 2:29 AM UTC

snowwrestler wrote: post #229

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…

Vox euphemized Schor’s departure in a way that was highly relevant to the subject of the article. The issue isn’t that pollsters couldn’t quantify the problem, it’s that they didn’t even realize the problem existed: https://www.washingtonpost.com/outlook/2020/11/02/shy-trump-...

> Trump voters are not liars. As a Democratic pollster and campaign strategist, I feel strange typing these words. But it is required to debunk the “shy Trump voter” myth.

This was written days before the election.

My post isn’t about “Shor’s personal HR problem.” It’s about a culture that caused very smart people to misperceive the facts on the ground, and misallocate funding as a result, in the middle of a high stakes situation.

People in Silicon Valley should understand better than anyone that culture matters and bears on organizational and institutional effectiveness. Often these are pretty theoretical and abstract debates. But here is a case study that elucidates culture that works versus culture that doesn’t work.

I’m not weighing in on the larger cultural issue. As I caveated, this is context sensitive. Political polling, news, academia, and the like are areas where free discourse with respect to sensitive issues is particularly important for the effectiveness of the institution. This is less true if you’re running a food delivery business.

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

#329

Post by jimbobimbo » Wed, Nov 11, 2020, 2:31 AM UTC

systemvoltage wrote: post #314

nxmnxm99 wrote: post #305

Earlier quoted context omitted.

Every single demographic went further red than 2016 besides white men. There’s a reason for that.

You’ve got data to back that up?

https://www.essence.com/news/politics/55-percent-white-women...

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

#330

Post by EGreg » Wed, Nov 11, 2020, 2:32 AM UTC

This is ridiculous!

Here I am telling people that Republicans are distrustful of mainstream media, science, colleges etc. and that pollsters must know what they're doing, getting representative samples is the best we can do etc.

And then I found out what I suspected is true. The most trivial thing of some people being overrepresented in a poll by the mere fact that they agree to take one, apparently, was somehow TOO MUCH FOR THE POLLSTERS TO HANDLE. Are you kidding me?

The methodology should involve getting truly REPRESENTATIVE and RANDOM samples. While each one is going to be biased in some way, the bias should be independent of political affiliation.

For example, stopping people on a street. Yes it excludes people who don't go out on the street, or people who won't stop.

Then you are supposed to note how many people refused to take the poll. And then you are supposed to measure the correlation between that and their beliefs.

If you really can't do this, then have a sign outside the voting areas, as people exit, with Donald Trump or Biden or JoJo etc. and have them sign up for polling in the next four years. This way at least you can make sure going in that your population is balanced as of four years ago.