So, what happened with the polls?

[Thundercrack85]

Democrat or Republican, Liberal or Conservative, I think if everyone is honest, we can all agree that this election was kind of a shocker. Why were the polls so off? And what we can we do about them?

 

[BestGirl]

Were they really all that off? Last I heard they were predicting a close election and it was.

 

[Thundercrack85]

Were they really all that off? Last I heard they were predicting a close election and it was.

They were so wrong that they very nature of the business of prognostication is in question. Seriously, this is a total embarrassment for all major polling companies.

 

[E.NormusJohnson]

They were so wrong that they very nature of the business of prognostication is in question. Seriously, this is a total embarrassment for all major polling companies.

To be fair a few polls had it in the margin of error. Which meant election day was always going to be a gong show.

 

[The Incredible Hulk]

Most polling right before the election had Clinton as a favorite above the margin of error. Poll aggregators like 538 had her with an at worst 70% of winning coming into the election. Granted that still gave Trump a 30% chance of winning which isn't terrible odds so its not like they were totally off base.

I think according to some the reasons they've been giving was that a lot of the people who came out of the sticks for Trump don't typically vote, so they weren't the database to be polled since polls look for "most likely voters." But I don't know if I buy that?

 

[Thundercrack85]

One common theory is that people simply lied.

 

[Handsome Rob]

One common theory is that people simply lied.

That makes a lot of sense, plus the large number of undecided (who may actually have just been lying, as well).

I mean, I imagine people really don't want to be called a racist, sexist, bigoted, homophobic, transphobic, xenophobic, misogynist, ageist, ableist, etc., person just because they wanted to vote for Trump over Hillary. So, when asked, they probably said they were voting for Hillary or were undecided. Given the willingness of the left to relentlessly go after people they disagree with on social media*, why trust a voice over the phone to keep your answer confidential?

I voted for the candidate (as a write-in) I wanted who didn't make even make it far enough into the primary for me to vote for him then. But, I would have told the pollster it was none of his/her business who I voted for.

Spoiler

*Unnecessary Clarification: My answer is directed toward undisclosed Trump voters, and so the ones who would go after them would naturally be on the left. I am unfortunately well-acquainted with the spats that go on from both sides of the aisle on social media, although I stay away from it, myself.

 

[The Shape]

Basically, the truth is what I always feared, which is that there were a ton of what I called "closet Trump supporters" out there who were unnaccounted for in the polls. These were not the people who would have daily pro-Trump or anti-Clinton Facebook rants, or the people who boasted about their support for Trump to friends or in the workplace, or the thousands of people who attended each of his rallies.

Rather, these were people who were embarrassed to admit that they planned to support Trump -- even to the strangers who conducted these polls. Polling, in general, is rooted in the belief that the people participating in these polls are being completely honest, which is obviously not always the case.

Beyond that, there was the surprise resurgence of the FBI "investigation" into Clinton's emails (which wasn't even really a re-opened investigation). There's no telling how much of an effect that bombshell had on the election that was to take place only 11 days later. Despite the fact that it amounted to nothing and was announced to be a "false alarm" days later, the damage had already been done, it still led to bad headlines for Clinton leading into the election, and definitely might have been the "straw the broke the camel's back" for some people who were on the fence or hesitantly leaning towards Clinton at that point. Trump latched onto it hard, and he wound up having a surge of momentum right through Election Day.

But even when the polls were showing a tight race between the two of them, most people just couldn't even fathom that Trump actually had a chance of winning it. The narrative was always that Clinton had it in the bag, either by a landslide or in a fairly tight race.

 

[The Shape]

I voted for the candidate (as a write-in) I wanted who didn't make even make it far enough into the primary for me to vote for him then. But, I would have told the pollster it was none of his/her business who I voted for.

Great job with that vote.

 

[E.NormusJohnson]

One common theory is that people simply lied.

Could be. I know a few personal examples but I bet there are quite a few out there who picked him, but were never open in support of him... even when polled.

 

[Donner]

Polls are ridiculous. You would need to know the method of polling (telephone, mail, internet survey, etc), the time of day people were polled, how many were polled, the ages of those that were polled, and so on to even start to form an opinion about the reliability of the poll. They're a crapshoot at best and a waste of time at worst.

 

[Handsome Rob]

Great job with that vote.

Well, considering I value my conscience more than the opinions of other people, I'd have to agree with you.

 

[E.NormusJohnson]

Remember some of these peeps were of the same ilk that projected Hillary blowouts back in the primary which ended up being W's in the Bernie column... that should have been a pretty big indicator they were not reading the pulse correctly right there.

 

[Darepool]

Well, considering I value my conscience more than the opinions of other people, I'd have to agree with you.

 

[DJ_KiDDvIcIOUs]

Seems like every election polls keep showing their age and that they are becoming less and less reliable. It's always been a guestimation game at best anyway. Only poll that matters in the grand scheme is the one that ended in the wee hours of this morning.

 

[The Incredible Hulk]

538 has a decent summary. Still a lot to figure out.

http://fivethirtyeight.com/features...-we-asked-pollsters-why/?ex_cid=story-twitter

 

[terry78]

Your best friend probably voted for Trump and you didn't even know it.

 

[Paradoxium]

I am not certain if I said it here, or somewhere else. But this is NOT hindsight.

There were massive oversampling of Democrats. Some oversampling is fine like a D+3/4, but D+8-13 is ridiculous. This was the norm in all the polls. Obama didn't get these ridiculous levels of oversampling. They pollster actually thought more Democrats would turn out for Hillary than Obama. After pissing off Bernie fans no less too. Are you kidding me?

Secondly, if any of you paid any attention the Republican primary, you would know Trump outperform in all forms of polling except online ones. The worse ones are land line polling.

 

[Paradoxium]

All the oversampling data are in the fine text. Feel free to go back and see for yourself. They started to tighten up oversampling closer to the election day. That's why it seemed like Trump was "catching up". HE WASN'T.

 

[TheVileOne]

Polls are nonsense and political polling is nonsense. You can twist them to tell you what you want to hear regardless. At the end of the day, it's just sampling. It doesn't give you the whole pie.

 

[fancyarcher]

I'm guessing the majority of Trump supporters were of the "silent kind".

 

[Webfoot Hero]

I'm guessing the majority of Trump supporters were of the "silent kind".

Many were. I was watching video of a conservative pundit who said people would come up to them and whisper that they were voting for Trump because there were afraid of what others may think of them for doing so.

 

[Thundercrack85]

Yeah, that's not an indication that you might want to reconsider.

But hey, at least it means that on some level they still know shame and understand what they were doing was wrong.

 

[SV Fan]

Were they really all that off? Last I heard they were predicting a close election and it was.

The average of polls predicted Hillary would win the popular vote by 3%, she won it by 0.2%(and that number might increase slightly).

The polls were within the margin of error.

It should be noted in 2012 the polls predicted Obama would win by 1% and he won by 3.9%(once again within the margin of error) so they basically were off as much in 2012 as they were in 2016.

Simple fact is nobody can tell what exactly the turnout model will be for any given year.

 

[DeadPresident]

Why are people surprised by this? The whole concept behind representative democracy is that you vote for the person you think will protect your interests? A lot of people in America aren't in a financial position where they can make a moral choice and hope for improved circumstances simultaneously. They were taken in by Trump's isolationist rhetoric thinking it will benefit them.

Some disenfranchised and unemployed steel workers aren't suddenly closet KKK members because they took a (admittedly bad) gamble on Trump hoping he'd take care of them. This idea that voting is a moral issue for many people is so naive it's not even funny. Most of us on the hype probably have the luxury of material circumstances where making a moral point with your vote was an option. A lot of the people who voted for Trump likely did so because they chose the candidate whose election they thought was most likely to deliver them a job.

If it isn't racist for the democrats to have captured minority voters it isn't racist for the republicans to have captured white votes, people vote for whoever will represent them. While it may morally be correct for the unemployed steel worker to vote for Clinton I don't think having a moral high ground pays his bills.