SIGforum
One year ago the reliability of opinion polling was made starkly evident.

This topic can be found at:
https://sigforum.com/eve/forums/a/tpc/f/320601935/m/7860088234

November 08, 2017, 01:40 PM
joel9507
One year ago the reliability of opinion polling was made starkly evident.
<IMG class="inline_image" SRC="https://s1.postimg.org/33bohfqenj/Wellesley_2016_election_night_sign.jpg"> <BR><BR>A night to remember, indeed, though not in the way they anticipated at Wellesley when they put that sign up. <img src="https://sigforum.com/groupee_common/emoticons/icon_smile.gif" alt="Smile" width="15" height="15"><!--graemlin::)-->With the benefit of a year to reflect, why do you think the polls were so wrong?In the era of violence creeping back into politics and with the current state of privacy, people not favoring Clinton were reluctant to admit it to strangersSampling missed Trump voters and oversampled othersPollsters knew the story but shaded their stories to try to affect the resultThe tea leaves were there but were being honestly misreadOther (please explain)
November 08, 2017, 01:44 PM
Deqlyn
It was quite obvious a purposeful misreading. When you looked at the polls they were weighted incorrectly. Hillary is +20 in OH but the mice type showed they polled 20 more democrats than republicans or indys. They didn't both mentioning it but several polls showed that in the articles. I stuck to my guns on the USC poll as it has the best record and sample size. You can look in the locked trump thread and see me referencing it.



What man is a man that does not make the world better. -Balian of Ibelin

Only boring people get bored. - Ruth Burke
November 08, 2017, 01:45 PM
ryan81986
I honestly think it's a mixed bag of reason. I'm sure there were a large amount of Trump supporters that simply kept their mouth shut. That combined with sampling from specific locations only and polls that were skewed in favor of a certain result.




November 08, 2017, 01:46 PM
Lord Vaalic
I think a lot of Trump supporters just didn't want to admit it and hear it from others, and others maybe didn't want to admit to themselves that they were going to vote for Trump.

In the end, it was "fashionable" and cool to be for Hillary but they just couldn't bring themselves to actually cast the vote for her.

And many others of the conservative bent just don't like to air their business and are private people




Don't weep for the stupid, or you will be crying all day
November 08, 2017, 01:56 PM
oddball
I think a mix of 2 & 3.

Heading towards election day, I was convinced that Trump was going to win because the sampling of the mainstream polls were so skewed to favor liberal demographics, they were not only wrong, but corrupted. Exit polls that day were controlled by the liberal media.

Like Snopes, when leftists control the information, there is a good chance someone is lying through their teeth.



"I’m not going to read Time Magazine, I’m not going to read Newsweek, I’m not going to read any of these magazines; I mean, because they have too much to lose by printing the truth"- Bob Dylan, 1965
November 08, 2017, 01:59 PM
TSE
The national results were actually pretty close. They had Hillary up about 4% and she got about that in the popular vote.
The state results all had higher margins of error and in many cases the margins of error were greater than the spread between votes.
It is likely that next time around both parties will treat the close states differently with respect to polling sample size etc.


Calgary Shooting Centre
November 08, 2017, 02:21 PM
parabellum
I was saying it long before November, 2016- ignore polls entirely. Polls are easily manipulated and should never be trusted.

If a pollster calls you, just hang up on them. Don't give them misleading information. Give them nothing at all.
November 08, 2017, 02:52 PM
HRK
One of the polls called me and I told them I was voting for Bernie....

Screw um, its none of their beeswax and figured if I told them to piss off they'd just mark it one way or the other anyway.

Wonder how many others do that.
November 08, 2017, 03:54 PM
bobtheelf
Polling did a terrible job of predicting turnout and enthusiasm. Same thing that happened in Virginia yesterday.
November 08, 2017, 04:05 PM
JALLEN
“My forecasting is none too good, particularly when it pertains to the future.” Lawrence Berra

They rely on statistics, 67.3% of which are misleading, and the rest flat wrong.

Statistics are real good for flipping fair coins, judging probabilities of closed systems, like cards in a deck, but as one moves into imponderable human actions, there are many problems.

Models based on the past work until they don’t. Look up Long Term Capital Management for the story of half a dozen geniuses burning through ~$6 billion or so under a statistically sophisticated trading algorithm that should have not happened in the life of thr universe times 7.




Luckily, I have enough willpower to control the driving ambition that rages within me.

When you had the votes, we did things your way. Now, we have the votes and you will be doing things our way. This lesson in political reality from Lyndon B. Johnson

"Some things are apparent. Where government moves in, community retreats, civil society disintegrates and our ability to control our own destiny atrophies. The result is: families under siege; war in the streets; unapologetic expropriation of property; the precipitous decline of the rule of law; the rapid rise of corruption; the loss of civility and the triumph of deceit. The result is a debased, debauched culture which finds moral depravity entertaining and virtue contemptible." - Justice Janice Rogers Brown
November 08, 2017, 04:23 PM
Rightwire
First, I can't believe you did a poll on the unreliability of polls... Razz


I think the people polled, and the questions asked we're slanted. I think it was done on purpose to drive an agenda and prey on people that they hoped would vote for the candidate that it appeared would win.




Pronoun: His Royal Highness and benevolent Majesty of all he surveys

343 - Never Forget

Its better to be Pavlov's dog than Schrodinger's cat

There are three types of mistakes; Those you learn from, those you suffer from, and those you don't survive.