How were the polls so wrong?

Disenfranchisement amongst the Democrats and lower turnout, and lots of shy Trump voters.

Poor sampling and Trump voters outside of the Southern States tending to hide their true political/economic beliefs for fear of being socially outcasted for openly supporting Trump due to rhetoric about racism.

This.

youtube.com/watch?v=LUjLDX-oJ3k

The polls weren't "so wrong". They overestimated Clinton's lead by 2-3% on average (unfortunately, the overestimate was higher in the swing states, magnifying the effect).

Some of the sites which used polls to provide "probability of victory" numbers were way off due to underestimating the margin of error and/or correlation (or both).

But don't forget how close the result was. An across-the-board 1% flip would have completely changed the outcome. Predicting US presidential elections is like trying to predict a coin flip based upon initial conditions.

One thing which polls can't reasonably account for is respondents lying; primarily about a) who they're planning on voting for and/or b) how likely they are to vote.

Even if you know that such discrepancies exist (e.g. asking people whether they voted in 2012 showed that ~5% of those who didn't said they did), there's a limit to how much you can adjust for such discrepancies before you're no longer polling and just making up numbers to fit a particular conclusion.

> Disenfranchisement amongst the Democrats
While that probably affected the outcome, it has no bearing on polling error. Pollsters only poll people who are actually registered to vote.

Thank you for the interesting and insightful post britanon

But they don't take into account things like deliberately bad locations of polling places, etc

>People who vote Labour were, for whatever reason, easier to get a hold of during office hours.

I'm saying they're mostly on the dole, I'm just pointing out the BPC's findings.

*Not saying.

I ppersonally think their biggest mistake was polling people who didn't vote, like most minorities. Trump didn't get too many more voters than romney, hillary just failed to get as many as obama

Trump also made gains with most demographics, including Latino's, of whom 30% voted for Trump.

This. Plus the EC system. A small shift in popular votes can hugely effect the number of electoral votes. So your polls only need to be slightly off to be completely wrong

> I personally think their biggest mistake was polling people who didn't vote, like most minorities.
Turnout amongst minorities tends to be lower than amongst whites, but it's hardly non-existent. The "likely voter" figures are weighted according to the proportion of voters from each demographic group in previous elections, i.e. they take into account racial disparity in turnout.

However, if they only (or mostly) used the figures for the last two cycles, they'll have overestimated black turnout. Again, this leads into the issue of "polling versus prediction".

Assuming that the effect upon turnout of "Obama is black, Clinton isn't" will be zero is almost certain to be wrong. So what would be a "correct" estimate? Hint: there isn't one.

As a general principle, any number other than zero (for anything) has to come from /somewhere/. Somewhere other than your ass, that is. Otherwise you've given up polling and taken up punditry.

Polling is about "here are our numbers, here is how we got them". Interpretation is the job of the reader.

> including Latino's, of whom 30% voted for Trump.
I'd like to see the breakdown between Cuban and non-Cuban Latinos. Those are almost entirely separate demographics in terms of voting habits.

>Those are almost entirely separate demographics in terms of voting habits.

Really, do you know why? I'm guessing it's got something to do with fleeing a communist dictatorship, right? Also I doubt there's any data to that resolution, still if what you say is true, it'd probably interesting.

They weren't wrong, Trump did not have a 0% chance so it makes sense for him to win.

Clit-ton having the popular vote means that she indeed had more chances of winning in different worlds. In a world where voter turnout was lower she would have won. Trump was pretty lucky, as is the case with everyone who wins with a minority.

Lefties and libtards were calling every person that thinks slighty different nazis, hitler and bigot shit and they are expecting those people to say truth in polls?
They may lie in polls, but they will vote on their own

Lefties mind
>you are not pro immigrant, you are nazi
>you are not prog transgender >some leftist insult
>you are not pro feminist? you are cis white fucking priviliged white male
>you are not pro abortion? you are fucking chiristian white male

etc etc.

leftist thinks that is okay to use such methods because they think that their opinion is the only right opinion.

here is a very rare example of libtard with actual brain
youtube.com/watch?v=GLG9g7BcjKs

Not really Cubans. Hispanics tend to be conservatives in pretty much everything, and idolize strongman authoritarians, the likes of Duerte. Cubans are only like 10% more "Red" than the rest of the country.

nytimes.com/interactive/2016/10/30/upshot/florida-poll.html

Florida did not really diverge much from polling.

>(((Binyamin)))

Christ, they aren't even trying anymore

...

If the latest polling technology is so off, what about historical polls? They must have been way off- fabricated, one might say.

>I'm going to reply seriously, and assume you aren't just ironic shitposting.

Not really since, in the present case, we have to ask people and those people often lie, for reasons already stated. When it comes to history you tend not to (just) talk to people, but look at the records and other artefacts from the time, taking the whole thing together leads to a more robust conclusion.

The holocaust never happened.

But it will.

>Lefties and libtards were calling every person that thinks slighty different nazis, hitler and bigot shit and they are expecting those people to say truth in polls?
Well not Nazis, but certainly bigots. Trump is also actually a fascist, and his movement is an actual fascist movement. Are you denying that?

But as we have seen, records can be easily modified. There were Hillary +12 polls the week before the election.

>There were Hillary +12 polls the week before the election.

But again this is asking people on their voting intention which might not be a good reflection of the population, for the reasons mentioned. You're comparing apples and elephants here.

Ignoring you holocaust denier bait
>In addition, some 5 million other persons were killed
It disturbs me how many people forget this part and only talk about muh six million.

even in the primaries, Hillary had a +21.4 polling advantage in Michigan and then Bernie ended up winning by +1.5
realclearpolitics.com/epolls/2016/president/mi/michigan_democratic_presidential_primary-5224.html#polls

How are the alleged Jewish population figures any different?

You gotta remember who's doing the polls. It's all media organizations that stood to gain a lot from Hillary winning.

What did FOX have to gain?

Cozy relationship with Madame president

Rupert Murdoch owns Fox. He's a big-time globalist. So of course he'd want Hillary to win; globalism has been gaining strength and Hillary would continue that.

Well I can see where this is going. Let's stop and review:
>Polling figures underestimated the amount of Trump supporters because of several factors.
>One of which was social acceptability
>This underestimate the Trump vote

Applying this reasoning to the polling of Jews in 1930's Europe would mean that the polls [math] underestimated [/math] the number of Jew in Europe.

Or Jews stopped reporting as Jews, dumbdumb

"no"

It simply had to do with the overwhelming majority of undecideds on election day

fivethirtyeight.com/features/shy-voters-probably-arent-why-the-polls-missed-trump/

That's what he just said moron.

Which is exactly what I said. The population figures would be smaller than the actual population of Jews.

lmao no

Give me a reliable source that proves that

And there is your 6 million killed figure. Don't you have finals to study for, undergraduate?

nytimes.com/interactive/2016/11/08/us/politics/election-exit-polls.html?_r=0

>Latino: 29%

Two point difference from last election.

elections.nytimes.com/2012/results/president/exit-polls

So he made gains with them.
>B..but it was only 2 points

Never said it was a big gain.

A shy tory effect happens when there are more voters than expected in the actual election.

But the ammount of republican voters had only a slight increase, what did change was the huge ammount of votes that the democrats lost in comparaison to 2008,2012.

So it was a mix of abstinence and fragmented vote in some states for 3º parties that actually had a sever impact in the election.

How was the board you posted in so wrong?

census's before WW2 are sparse and inaccurate
The soviets didn't take ethnic census's afterwards

So mostly its fantasy

I suspect most of it is just voter fraud by democrats

What the fuck are you talking about.

demoscope.ru/weekly/ssp/rus_nation.php

ARE YOU KIDDING ME!??

no, normal people were you tired, they were tired of all this bullshit
>muhh safe space
>Professor Called ‘Racist’ for Correcting Black Student’s Paper
>muhh diversity
>muhh immigrants

funniest part is that majority of this statements were made by some spoiled white kids from guarded estates that had no real life experience outside of private schools and they did not work single hours in their lifes.

Normal people were voting trump not because they like him but because it was a fuck you into all this bullshit elites that live in ivory tower and have no connection with normal people life.

>George E. P. Box - "essentially, all models are wrong, but some are useful"

George Edward Pelham Box FRS[2] (18 October 1919 – 28 March 2013) was a British statistician, who worked in the areas of quality control, time-series analysis, design of experiments, and Bayesian inference. He has been called "one of the great statistical minds of the 20th century"

>538

lol

Dunno.

Let's hope Dr. Stein finds out.

>Denies he made a gain
>Ends up proving he made a gain

Total bullshit, the turnout so far has increased. The total turnout as of right now for this election (only counting Trump and Hillary) is 127,079,854 and 2012's was 126,849,299. There was ~1 million decrease in Democrats overall and ~2 million increase in Republicans overall. This is likely almost all in the Rust Belt considering the results.

I am going to correct myself, the gains were across the entire North

>Trump is also actually a fascist, and his movement is an actual fascist movement. Are you denying that?
I think most sane people would deny that.
If you call everyone a fascist, what are you going to call actual fascists? It's like you've never heard of the boy who cried wolf.

People really didn't like Comey reopening the investigation.

I do not take the time to be polled so they never get my type of demographics.

They made their predictions based on turnout from Obama/Romney. Which is dumb because they should have used historical black turnout and such.

>I think most sane people would deny that.
I am sure most people would deny it.

>If you call everyone a fascist, what are you going to call actual fascists?
I am well aware of repercussions of being dismissive of political ideologies. As well the problems with calling everyone racist/nazi.

When I say Donald Trump is a fascist, I mean it as a factual observation, not in a derogatory sense. Not fascist as in "mass murder", or "white nationalist", or "Nazi".

His movement/philosophy is rebirth myth anti-liberal ultra nationalism, decrying societal decay, as well as emphasizing traditionalism, masculine values, and "energy". The only thing that is missing is dismissal of ineffective monarchy/church.

Maybe he succeeds in making America great "again". His movement nonetheless is a fascist movement.

Trump is a liberal.

Fascist is a loaded word. It's impossible to call someone a fascist without the implication that they are like a Nazi. That's why it shouldn't be used except for people who could legitimately be compared to Hitler.

In the same way conservatives claim the Hitler was a liberal.

liberal bias

You are an idiot. Trump is just parroting the politics of the ~80s that he grew up with. Hell a bunch of the stuff he said came out of Bill Clinton's mouth in the 90s. Modern times are so liberal everything looks fascist to you, and you are too young to have any kind of actual historical perspective.

And for the record, there is literally nothing wrong with white nationalism, nazism, and fascism. Trump would have gotten even more votes if he was an actual fascist.

This sounds very PC to me.

Connotations are very important in the English language. You wouldn't trust a doctor who told you you had "dick flu," would you?

>And for the record, there is literally nothing wrong with white nationalism, nazism, and fascism. Trump would have gotten even more votes if he was an actual fascist.
You got me friend. My head is actually spinning with this racist/not a racist thing.

You are figuratively making my liberal head explode.Too many mix signals are hurting my liberal sensibilities.

>massively funded computerized voting polls with 2016 technology are very wrong

>paper sheets produced at a time of literal chaos in 1945 are 100% indisputable fact and you will be jailed if you disagree

>there is literally nothing wrong with white nationalism, nazism, and fascism

USA should do what they did for Brexit. Paper ballots with the counting done broadcasted live.

Glad someone else agrees.

Are you aware of the irony and just trying to play the level headed role?

its called shy tory factor in britain

liberals make all the ebil nazis feel so bad they just pretend to vote for the "nice" person then actually vote how they want to

kill yourself

Political hacks desperately trying to convince the sheep to herd up behind the party approved choice by any means they can.

It Wasn't the Polls That Missed, It Was the Pundits
realclearpolitics.com/articles/2016/11/12/it_wasnt_the_polls_that_missed_it_was_the_pundits_132333.html

Anybody who actually read the polls could have predicted Trump's victory against Clinton back during the DEMOCRATIC PRIMARIES. The Dem party and its pundits deliberately suppressed coverage of polls that showed Bernie beat Trump by ten extra points compared to Hillary. The same shit happened all through the general election as well, and even with pollsters deliberately tweaking their sample methods to give Hillary an edge ( reuters.com/article/us-usa-election-poll-reutersipsos-idUSKCN10910T ) there were still plenty of results showing that it was a close race toward the end.

The polls didn't lie, it was the "data science" faggots like Nate Silver, and the pundits who lied to you.

>Registered democrat who voted in the primaries
>Get multiple calls from polling niggers
>"IT SAYS HERE YOU'RE UNDECIDED. ARE YOU GONNA VOTE FOR TRUMP OR HILL?
>Rinse and repeat for nearly a month

>1.6% trump

well it wasn't 0, so it wasn't wrong.

Nate Silver was defending his 30% chance of a Trump victory constantly in the days leading up to the election. He is one of the good guys.

Just because something is extremely improbable doesn't mean it won't happen

30% was obviously still a bullshit stat since it turned out that Trump won, meaning giving him anything less than a 50% chance was wrong.

Statistics is science.

>tfw there are people who actually think like this

Face it, Nate's prediction was wrong. If his 30% chance is right BECAUSE IT STILL ALLOWED THE POSSIBILITY OF TRUMP WINNING, then the predictions giving Trump a 1% chance of winning were also right.

You have to go back

Delusional liberal detected. Please, explain how Nate's totally arbitrarily-designed algorithms are scientifically and statistically "true" when they can't even predict the winner out of two possible candidated.

Shoo, back to your hugbox. We don't want you here.