We shouldn't treat the results of polls as gospel truth, but we should trust them more than we trust our intuitive sense of what's gotta be true. A poll showing 90% of Americans are satisfied with their personal life is a pretty strong refutation of claims like "suchandsuch trend has propelled America to unknown depths of misery", even accounting for the possibility of bias.

> We shouldn't treat the results of polls as gospel truth, but we should trust them more than we trust our intuitive sense of what's gotta be true.

Yes and no. Polls shouldn't be treated as gospel truth, but they shouldn't override our intuitions either. If there's a discrepancy, it means you have to dig deeper find the truth.

I think there's a level of discrepancy where they should override our intuitions. If I thought most Americans were unhappy these days, my reaction to seeing this poll would be to question how my intuition went so wrong, not to look for ways the might be biased.

> If I thought most Americans were unhappy these days, my reaction to seeing this poll would be to question how my intuition went so wrong, not to look for ways the might be biased.

That's putting undue weight on quantified data, and letting it shut down attempts to actually find the truth. Polls can be biased or badly designed, metrics can be gamed or not actually measure what they claim to, or the poll could be founded some flawed assumptions. If you assume your intuition is wrong when you see a "strong" poll result, you'll never discover errors of the kind I just enumerated.

> 'If I thought most Americans were unhappy these days, my reaction to seeing this poll would be to question how my intuition went so wrong, not to look for ways the might be biased.'

Your comment reminds me of this anecdote:

> "What I've noticed from users is that a windows user will blame the computer every time even if it is their own mistake, and an apple user will blame themselves every time even if the hardware failed.

> "Apple users know they have the Fancy Name Brand that Cool People(TM) use, so they have multiple levels of psychological aversion to blaming it for anything. They would be less cool by extension. And the brand image doesn't place any value on being capable of using "regular" technology, so they take no hit to their hipster image by blaming themselves.

> "It is really the same as any other name brand item. If a person buys name brand jeans and they tear, they're going to blame themselves for not taking better care of something expensive. Another person who buys a different brand gets mad that it didn't last longer because they only tried to use it in the "normal" way. If they blame it for anything, even its own real failings, then they lose the snob value."

The interpretation of said anecdote as it applies to your situation is left as an exercise to the reader.

Agree. We’re living with every choice being polled and chosen by those polls. There’s nothing worse for any policy than a bad poll. Unless it’s the seeming acceptance of around 43% being a majority for some.

Edit-it’s always worth considering the error bars on polls. News will gladly say A is preferred to B when their difference is well within error bars.

The 2016 poll was not off, what was wrong is the prediction models which did not take account of electoral college sufficiently. Poll and prediction are two very different things.

No pre-election analysis had Trump at 0%. Unlikely outcomes happen.

If I put two blue marbles and one red marble in a bag and ask you to blindly draw one, you probably won't be too shocked if you end up drawing the red one. Yet if it happens in an election everybody starts suddenly claiming all polling and statistics are useless.

> After the 2016 election where all of the polls showed Hillary winning the election

The actual analysis of the polls showed Trump was well within striking distance of Hillary. People just didn't want to hear that.

Why bring mass-insult style politics into this? The polls described large differences in likelihoods, not large differences in numbers. People were not ignoring the polls.

> Why bring mass-insult style politics into this?

I didn't think I was?