Polls useless?

So, in a post Trump world… Are polls useless?

I had an argument in this group many years ago. It was about things that happen in the black community. Folks who disagreed with me were mainly 1. Not black and 2., really trusting of polls.

So. Do we still trust polls?

Because of the failure of poll aggregators after the election, a lot of Trump supporters have essentially decided that polls are fundamentally wrong. Never mind that the actual polls themselves were within error margins for the most part…

Ahh. I’ve checked out of politics. I’m in a fucking weirded out crises about Trump. There should be a support group.

And not because I think he’s gonna fuck things up. Just because my whole idea of presidents and who can get in is shaken on a very deep and fundamental level.

Polls are still fine. They’re reasonably accurate, although subject to some kinds of flaws.

No one would be wise to bet his life on them…and no one would be wise to ignore them entirely.

Looking at the OP from a different point of view - I wonder whether enough people are refusing to answer unrecognized calls (based on Call Display and selective call blocking) to cause an impact on the validity of telephone polls? i.e. does that significantly skew the demographics of a typical poll?

I think there should’ve been a poll included with the OP.

That’s my thought. The polls don’t seem to take the “Fuck off and stop calling me” numbers into account. And those people have a major overlap with “disillusioned and pissed off” which makes them unpredictable.

They are as good as ever and as bad as ever. There is always a margin of error in any study or poll; we just usually forget to see what margin they claim and add to it depending on the subject and who released it. Its a good starting point for thought but hardly a complete thought in and of themselves.

Trump got 63 million votes and became President. Had he gotten only 62 million votes he’d not be President. Is it those extra million votes specifically which weird you out?

It’s incumbent on the news media reporting the polls to scrutinize the polls for scientific legitimacy. And who can trust the news media?

Polls are just polls. Accuracy has been highly overrated at times. A poll can readily tell you if a candidate is going to win by a wide margin. The closer the split of the actual voters the less accurate they are. Polls are also based on models of voting trends by demographic groups. Usually those trends and groups are useful for making predictions, but as in the last election they become useless when there’s a major turn in the political road. You can’t rely that much on a sampling of a tiny percentage of the public. And voting polls can be the most accurate because elections come down to simple choice. When pollsters ask about opinions, approval, stances on issues and the like people are less reliable in their answers because people may not have a strong opinion and less detail. Ask them how they feel about a social program and they may say it’s a waste of money and should be eliminated. Ask them how they feel about losing the benefits they get from that social program and they’ll say they’ll never again vote for the politician who cuts that program. It’s about how you ask.

The question asked makes a big difference. Long ago a group of seminary students were polled and asked if it was allowable to smoke while praying. The poll showed they overwhelmingly thought it was not. Asked again if was allowable to pray while smoking and the poll showed they overwhelmingly thought it was.

Remember when “anyone can be the president” was a promise, not a threat? Good times. :frowning:

Honestly, it’s all 63 million. There are 63 million people in the USA who looked at the choice between Hillary Clinton and Donald Trump, and for whatever reason, decided that the latter would make a better president. My home county went for clinton… by like 2%. That means just under half of the people in the county I grew up in are fucking crazy. I’m not going back there until this fucking mass psychosis lifts.

I’m with you, Nzinga. There’s a post in a thread today (cannae be arsed to find it) which mentions a new poll about something from that 538 site and all I thought was “yeah, that’s the one that said it was virtually impossible for trump to win”. So I just don’t think of them as being valid anymore.

Not true. The day before the election, 538 gave Trump 29% chance of winning. Which basically meat it was a toss up. Earlier they gave him as low as 10% chance, but that was with the understanding that public opinion may shift between then and the election.

Some other aggregators had much lower chance for Trump, but I believe that was because they treated the polls’ margins of error as independent errors, and did not consider the possibility of a systematic bias. 538 did consider it.

The polls weren’t all wrong. Most of them were correct to within their stated margins of error. The problem was that they were ALL slightly off in the same direction.

It really shouldn’t have even been close.

I would say approximately…26 million of those votes really weirded me out. There were about 136,600,000 total votes; by the 27% crazification factor, a major party candidate as bad as Trump* should have gotten something less than 36 million votes.
*Assume here that the Republican Party primary electorate are all just batshit crazy, a disturbing enough proposition.

I mean, I understand that in a country with over 300 million people, you’ll never get a situation where one side gets all the votes, no matter how much I personally may think voting for the other guy is complete lunacy. And obviously I’m joking about the mathematical precision. But, given that a major party nominates a candidate as clearly and objectively god-awful as Donald J. Trump, that party deserves to get blown away in the biggest landslide in the history of the Republic. (I can totally see that not all 73% of the “Oh, dear God, no, President DONALD TRUMP?!? Are you INSANE?!?” vote will necessarily go to Hillary Clinton. I would certainly expect to see a very, very unusually large number of third party votes in such a bizarre situation as a Trump nomination. The party that nominates Donald Trump to the position of President of the United States really deserves to come in third, or maybe lower, even if that party is a traditionally “major party”.)

Yes, this was my point. It weirded me out dramatically that 63 million people voted for this preposterous showman, and 62 million would have weirded me out just as much.

Once we knew there were 60+ million stupid/ignorant (it’s an exaggeration to call them “psychotic”) voters, his winning wasn’t weird — it was all a drunken crap-shoot at that point.

The vote demographics are so disheartening. Among white males, Trump beat Hillary more than 2 to 1. :eek: Among 18-24 year olds and those with post-graduate college — groups we were counting on —, Hillary won, but only about by 8 to 5.

Wow, is this ever WRONG. 538 was among the few popular sites to CALL IT RIGHT. They’re the ones that screamed, week after week, that there was a HUGE chance that Trump could win.

Don’t blame 538 because so many people can’t understand that “between a 1 and 4 chance and 1 and 3 chance of something happening” means, roll the dice, many many times in life, it WILL HAPPEN. Not partially happen. WILL HAPPEN.

And so it did.

Yes. Many, less numerate people see “29%” and think “very small chance.” (Maybe they think that’s the portion of the vote he’s likely to get!) But (while not really a “toss up”), in a context like this 29% is a huge chance — the signature of a close election.

If I were about to get on an airplane and the pilot assured me that the chance of a crash was only 29%, I would not board! Yet that’s what America and the world were faced with on that tragic Eighth of November.

This whole post got me cracking up. But you right. My inability to understand polls didn’t help matters.

But beyond my own ignorance, it was watching pundits and talking heads do spit takes and say, “WHUUUUH?!!!” that made me feel unsafe with polls.

Right. The analogy I used on this board was this: If you have three cartons of milk in the fridge, and you know one of them is sour and will give you a retching stomachache if you drink it, but you don’t know which one, do you think “oh, there’s no way that could happen,” and just grab a carton and start gulping down?