The polls were very clear that 2016 could go either way. Clinton was favored but it was close. This is equivalent to saying "You're gonna roll the same die that just rolled a six?"

I don't agree at all. Many articles were published about the polling crisis in 2016 and on.

https://www.theatlantic.com/ideas/archive/2020/11/polling-ca...

So there's lots of extreme language in this. What you makes you say this is wise, vs. a rant from some dumbass who doesn't know anything? Do you know anything about statistics and polling? Here's some comments from this article:

> Even with the results of the presidential contest still out, there’s a clear loser in this election: polling.

> This is a disaster for the polling industry and for media outlets and analysts that package and interpret the polls for public consumption, such as FiveThirtyEight, The New York Times’ Upshot, and The Economist’s election unit. They now face serious existential questions.

> The real catastrophe is that the failure of the polls leaves Americans with no reliable way to understand what we as a people think outside of elections

> Without reliable sources of information about public opinion, the press, and by extension, the public, should perhaps employ a measure of humility about what we can and can’t know in politics. As wise as this may be—and even if people manage to act on it—that sort of epistemic humility risks falling prey to the same asymmetrical warfare that has characterized much of the Trump era.

You know what we don't see in this? Any sort of measurement of how good the polls were. And you know what? They were fine. Not a great year for sure. 2 years prior was very good.

If you want an analysis on the efficacy of polling from someone knowledgeable I would suggest looking at Nate Silver's reflections on 2020: https://fivethirtyeight.com/features/the-death-of-polling-is...

Or this one: https://fivethirtyeight.com/features/how-fivethirtyeights-20...

99% of races favored for republicans were correctly called, including 100% accuracy for anything not labeled a toss up. 92% of races favoring dems were called correctly. and 95% overall.

Yeah even the most favorable polls for Clinton gave Trump a 20% chance. The failure in the runup to the 2016 election wasn't polling, but rather most people failing to understand probability.

A 20% chance still happens quite often.

Exactly. It's a 1 in 5. Sure the polls put the odds against Trump, but they were far from overwhelming.

No it isn't I'm suggesting that the expert class is using manipulated dice for gaming social dynamics to get money,power, and influence.

Aka it's not about truth. I think the replication crisis is one and the 2016 election. Manny many polls absolutely said Trump has no chance. I remember everyone I knew truly being shocked Trump won.

It's easier to imagine a nefarious plan, and to lump everything negative together into one big ball of conspiracy, but it's not reality. Realty is far more complicated than anyone's mental model, and each of the things you describe--the replication crisis, your misunderstanding of 2016 polling, the increased concentration of wealth and power--are separately explained in reality, but each is complicated.

Humans are messy and seek simplicity, which ironically is one of the biggest factors in the replication crisis. One might even think that seeking simple answers to complex question is what has led us to this point right now, in which some of choose to reject even the idea of complex answers, insisting that only simple answers can be trusted.

It's probably fair to say experts as a group are failing their duty, and americans are losing trust in them.

How America Lost Faith in Expertise:

https://www.foreignaffairs.com/articles/united-states/2017-0...

There are a lot of failures affecting the public trust going on.

Top economists fail to predict inflation (wall street bets didn't get it wrong)

https://thehill.com/opinion/finance/3259197-nobel-economists...

Pollsters got 2016 so wrong, articles were written about the polling crisis.

https://www.theatlantic.com/ideas/archive/2020/11/polling-ca...

Psychological Science doesn't replicate, and is probably all garbage. https://www.vox.com/science-and-health/22360363/replication-...

No evidence that depression is caused by low serotonin levels, finds comprehensive review (Apparently Tom Cruise was right about psychiatry)

https://www.sciencedaily.com/releases/2022/07/220720080145.h...

https://www.youtube.com/watch?v=eymr9NZpna8

Alzheimers research had some high profile fraud caught recently. https://sciencebasedmedicine.org/fraud-scientific-rigor-and-...

Sure, but you're relying on reporting here, which seems a bit circular, no?

WSB has predicted a spike in inflation six times out of the last one. They also predicted that their wife's boyfriend would ride a rocketship to the moon. In addition, it's by no means settled that the economists mentioned in that opinion piece were wrong: inflation is currently a worldwide phenomenon seemingly tied more to Russia's invasion of Ukraine and supply-chain disruption than any US-specific increase in the money supply. The extent is debatable, but the linked opinion piece is offering a simple critique of a complex situation.

Reporters wrote many articles demonstrating their lack of understanding of statistics and polling after 2016, yes.

Psychological "science" has been largely-garbage since the beginning, which is why it is often described as a "soft science," a charitable way of saying "not really science." The seratonin survey specifically is debated, and seems like another example of reports over-reporting the limited findings of the review.

There will always be examples of failures at almost every level, and I don't think there are more failures now than in decades past, but we are primed in a variety of ways to now believe that it is a new and somehow dangerous thing.

If you're describing "experts as a group" and meaning "reporters," then no question, reporters have been getting things wrong for a very, very, very long time. As a general rule of thumb, have you ever seen any news reporting ever get things about computers right? And if not--I certainly haven't--why should I not assume that doctors feel the same about medical things, lawyers feel the same about legal things, and so on. Reporting is a skill, but that skill is writing and communication. It's not expertise.

Americans are losing trust in institutions for many reasons, but I don't think those institutions are failing more spectacularly than they have in the past. I think the loss of faith is more subtle than that.

And specifically, many Americans are replacing that lost faith in institutions with the most ludicrous faith in bizarre nonsensical theories of everything, seemingly with the reasoning that "if any experts are wrong about anything, then they must all be wrong about everything."

Most of us got most of our general news from one of three or four news readers, and so we mostly shared similar opinions. People shouted around the edges about details, but that was easy to tune out since it wasn't on one of the big three TV networks. Then 24-hour cable news and the internet changed the way any of us consume news, and I suggest that has much more to do with changing attitudes than any specific details of "experts failing."

Again, we all want simple explanations, and it's easier to believe no experts know anything than to accept that life is so complicated that future events are often hard to predict with perfect accuracy, but most of the time, most "experts" are mostly right, so we don't notice that they were only mostly right, not perfectly right.

P.S. Still, great list of links, so I upvoted you despite disagreeing.

It's funny because I remember 538 and company being attacked for saying it's likely Clinton but could go either way; and how that "wasn't a real prediction".

You were not well informed. That's the short of it. I have no doubt that many people took aggressive takes that were not valid. But reputable analysts were very clear on the issue.

Appealing to the majority on nuanced issues is never a good idea.

Worth noting that most of 538's 'secret sauce' was in combining state by state poll results with national polls to take account of the weird Electoral College system. I'm guessing most of the predictions that people remember as saying Clinton can't lose were basically just predicting the popular vote, which she didn't lose, but even then wasn't really that far outside the margin of error.

https://en.wikipedia.org/wiki/Nationwide_opinion_polling_for...

I'm not sure anyone was saying "literally no chance" in regards to the data rather than "odds are very low" — even with low odds there's always a chance. Of course people were shocked; the odds were low! If the odds were always right no one would care about sports.

> I remember everyone I knew truly being shocked Trump won.

A lot of the shock was due to how terrible the outcome was, rather than due to how unlikely it seemed.

I tend to agree here. Exhibit A would be this story which is on HN at exactly the same time as this one here. https://www.nature.com/articles/s41586-021-04198-4

That's not how polls work. They have error bars. Polls showed that election could have gone either way.

They just showed Clinton was the more likely winner. Polling in the US tends to underrepresent actual Republican votes by a few percent.

Could you be more specific about who you don’t trust, and how you determine truth?