Real time Facebook presidential poll shows a completely different story than mainstream media polls

Whether people want to accept it or not they are being manipulated, the whole point is that they don't realise they are being manipulated, so they think they came up with their beliefs themselves.

It all goes back to Bernays who used all kinds of tricks to manipulate the public in all kinds of ways, from getting women smoking with their "torches of liberty", to getting people to eat bacon for breakfast, and on and on.

One of his favourite tricks was to set up what appeared to be an independent third party organisation to "verify the facts" to the public, like a grand sounding medical establishment to say smoking is not harmful.

With politics part of that would be polls.

With regards to politics he believed:

"the public's judgment is "not to be relied upon," he feared the American public "could very easily vote for the wrong man or want the wrong thing", so that they had to be guided from above."

He said:

"If we understand the mechanism and motives of the group mind, is it not possible to control and regiment the masses according to our will without their knowing about it? The recent practice of propaganda has proved that it is possible"

He spent his life learning how to manipulate the public and putting it into practice - does anyone really think that current politicians have forgotten about this, or don't want to manipulate people to do what they desire?

If you feel that politicians are far more shady today than they used to be in the past, then you can be 100% sure that they are using every dirty trick they can muster.

Since it is patently clear which one the media is backing, and the media is owned by the big money men - the ones who spend $Billions per year lobbying congress - the ones who really own the country and all of you too - well don't believe a word the media says about the one they aren't backing.

Agreed. Help us out as a former pollster and statistician. How are the populations selected to be representative of the voting public and then analyzed? Was I close? Are there statistics on accurate and erroneous predictions by public or private organizations to help us focus on just the reliable ones? Help us understand how fair and logical or flawed they are. I don’t think you were implying that just because two actions are the exercise self-selection represent the same population?

a reply to: darkstar57

a reply to: Greggers

Legitimate polling organizations poll RANDOMLY, with sample sizes that give reliable margin of error.

How do the legitimate polling organizations determine that the random sample size give reliable margin of error? I mean .... ................... I don't know much about statistics.

originally posted by: thethirdsign
a reply to: Greggers

Legitimate polling organizations poll RANDOMLY, with sample sizes that give reliable margin of error.

How do the legitimate polling organizations determine that the random sample size give reliable margin of error? I mean .... ................... I don't know much about statistics.

Exactly, if it's random (with so few voters polled) it could randomly be 100% Trump, or 100% Clinton, so to be scientific for such a small sample size it must be predetermined.

If it's predetermined then it is neither random nor scientific because subconscious bias will influence who is asked.

Polls are fools fodder, the only way to really get any real idea would be to poll say 50% of voters across all regions, genders, ages, race, colour, religion, etc.

Trying to predict or even claim that the poll of 1000 people, or 10,000 people, or 100,000 people, or even 1 million people is accurate when the total voting pool of eligible voters is 250 million ish...

Well, you don't need to be a rocket surgeon to see that any data from such a poll is ridiculously unreliable and is therefore only there to try and sway opinion in one direction or the other.

That is why the Facebook poll of 100,000+ is far better - the news that there is a poll to vote on spreads far and wide, it incorporates lots of people from lots of backgrounds (like me - Im in the UK - even I have heard of it though I haven't voted on it), so it isn't tied to state, town, city, colour, creed, gender, religion, etc.

It encompasses everyone from every background, every place, every political persuasion AND it covers a MUCH larger sample size.

Would you be happier to put YOUR money in some outcome where one poll of 1000 people says X, or another poll of 100,000+ people says Y?

THINK.

This is conspiracy site, yet people seem far more interested in believing we're being ruled by alien lizzard overlords than that the politicians and TPTB might be trying to manipulate public opinion for their own gain.

FFS - WAKE UP!

originally posted by: thethirdsign
a reply to: Greggers

Legitimate polling organizations poll RANDOMLY, with sample sizes that give reliable margin of error.

How do the legitimate polling organizations determine that the random sample size give reliable margin of error? I mean .... ................... I don't know much about statistics.

I posted this a while back. It's a simple online calculator where you can plug in the population size, along with the margin of error and confidence level and get back the required size of the sample: www.surveysystem.com...

Polling organizations all have their own special sauce to get the most accurate mix. For example, most do try to get a representative split on things like gender, but political affiliation they do not because there is no hard data for it and it's viewed as flexible. So, one of the things that gets measured inadvertently is the percentage of people who claim to be a loyalist to each party.

You may recall during the last election there was a site called unbiasedpolls (I think) that focused exclusively on polls with a 50/50 split with Republicans and Democrats. All those polls showed Romney winning. All those polls were wrong. The reason is that the 50/50 split is not representative of reality. Party affiliation is dynamic, and lots of factors influence whether a person will profess to being D or R at any given moment in time.

a reply to: Power_Semi

What you say I find it to be consistent with what I know ... I mean it's difficult to get a reliable data with marginal error of say 5% if you really take a random sample even if the sample population is known; because the sample can be affected by the environment unless and until they are in a controlled environment. What I am saying is that if the random sample is not in a controlled environment the data obtained should be skewed to a certain target which the FB poll seems to point (unless the FB data is taken from a preconceived/controlled sample) which in effect should be what democracy is. LOL. But frankly I'm not an expert on all these polls and elections or rather statistical data analysis.

a reply to: Greggers

Dammit I should have paid more attention to statistics during my school days. I have no Idea or rather I cannot comprehend the info provided in your link. But it was informative unfortunately I am more confused now..

Why has this thread become so popular?

Very odd.

originally posted by: defiythelie

originally posted by: Rezlooper

originally posted by: introvert

If she wins this election I will be totally convinced that America has been hijacked by the Clinton Cabal and we are all friggin' doomed!

Because of a non-scientific poll conducted on facebook?

That makes no sense.

Scientific? This poll is much more accurate than these phony polls by the media who you have no idea who they are actually calling.

The media hires firms whose whole business is polling, if they were not accurate they would go out of business. Anyways, you are naive if you do not understand how easy it is to game an internet based poll.

And you are naive if you do not understand how easy it is to game a media-controlled poll.

originally posted by: knoxie
a reply to: starlitestarbrite


I live in Colorado, and I honestly don't know anyone voting for trump expect my father (who's from Pittsburgh), and I'm still hopeful he'll vote third party like my mom.

in our neighborhood there are two Clinton signs and two trump signs. most signs are for our congressman who's a democrat.

Colorado? The voting booths are too far when you're stoned, laid out on the couch with a bag of chips and a video game.

Seriously... Colorado would have a lot to lose if a conservative agenda took over American politics. They may put a stop to the weed free-for-all. I may not agree with it, but it's true.

originally posted by: SaturnFX

originally posted by: Bloodydagger
a reply to: SaturnFX

I have over 1000 friends on Facebook. Family, old high school friends, current friends and co-workers etc.

I am an old man (40) and so who I am generally surrounded with in the deep south here are trump supporters.
but
I have good genes (with a bit of prep I can pass for much younger) and my 26y niece is like my sister. She and her fiance are constantly asking me to go out with them, and they have a very wide circle of friends (you know the type..tons of hipster girls and muscle dudes all past their teens but still obnoxiously young) and they are all anti-trump simply because its the in thing.
They dont participate in online polling or any of that nonsense, they aren't locked behind a keyboard because they are typically having one of those stupid social lives.
If I went by stay at home moms and dads that stopped participating in normal social society, sure. it would be a Trump landslide. Thing is, online political activities are most done by old grumps like me, and teens who can't go out past 10 so will yammer on about stuff that is a bit above their knowledge base...the rest are watching netflix during the week and out socializing on the weekend, with occasional visits to snapchat and facebook to discuss trivial issues.

I am trying to spread the good word about how there are 3rd party candidates and you dont need to vote for the lesser demon...nobody listens, nobody cares enough tbh, and frankly they may be right. house/senate can keep a potus in check if they get too crazy.

You're so very wrong to claim twenty-somethings aren't on Facebook. I don't know a one that isn't.

originally posted by: thethirdsign
a reply to: Greggers

Dammit I should have paid more attention to statistics during my school days. I have no Idea or rather I cannot comprehend the info provided in your link. But it was informative unfortunately I am more confused now..

A few pages back I showed that you can accurately measure the entire voting population of the U.S. with a +/-3 margin of error and a 95% confidence level with a sample of only 1069 people.

originally posted by: Greggers

originally posted by: thethirdsign
a reply to: Greggers

Dammit I should have paid more attention to statistics during my school days. I have no Idea or rather I cannot comprehend the info provided in your link. But it was informative unfortunately I am more confused now..

A few pages back I showed that you can accurately measure the entire voting population of the U.S. with a +/-3 margin of error and a 95% confidence level with a sample of only 1069 people.

A well-constructed survey, employing a simple random effects OLS (ordinary least squares) model can, as described above, use a sample of 1069 respondents to give a pollster a clear, unambiguous result that represents the whole of 320M people. Again, as mentioned, the confidence interval of 95% referred to is solid and the degrees of freedom afforded with said sample size is rigorous and would be a well-accepted statistical result among those who understand this type of analysis. This, of course, is all predicated on a well-constructed survey and very dependent upon the methods used to reach said analysis. This is all true. Also true is the fact that many (maybe all) polling organizations freely provide their methodology for verification and validation.

Speaking of verification and validation, many on here have whined about the over-sampling of democrats; these surveys are bias and skewed to represent a higher than expected spread between the candidates and underhandedly making attempts at rigging the election by constantly presenting polls that can't be trusted because they're bias. I hate to break it to you, but these people (pollsters) are professionals and know what they're doing and fully stand behind their results because any 'joe six-pack' could go online and read the appendix and find their methodology and run the numbers themselves -- a chance to pull back the curtain and show the people what the polls are really doing and ruin Hillary's night by showing a fraudulent poll, hence, reinforcing the idea that the election is rigged. Well, as I said, these people are professionals and even if there is an over-sampling of democrats in a certain poll (some polls are random, some are comprised of panels of paid survey-takers, some of them are pre-screened to reach a certain demographic, etc.) there are creative little techniques/tools statisticians/econometricians/etc use to account for that sampling demographic - the easiest of these techniques is weighting. The are many ways to weight the independent variables so that the statistical power of said survey is still powerful and credible. It's a shame not everyone had to spend their young twenties in graduate school learning this stuff, but for me, I did and it paid off. There's a lot more to it and Greggers has been relatively spot on, but the drivel coming out of some of these people's mouth is beyond hilarious, it's a bit concerning because I've seen some - what I at least consider - intelligent posters say some idiotic things and did little to think critically, letting their 'vision' or worldview render them brain-dead.

Self-selection bias and the inability to understand how pollsters weight their coefficients based on sampling demographics/make-up are the two biggest issues I have noticed in reading EVERY response. People, please take a minute to understand the dynamics of a poll and understand that no one single poll will ever always be correct, but an aggregation (sample) of those poll results does have the ability to narrow those margins of error and increase the confidence interval due to several reasons particularly, increased degrees of freedom.

I recognize I replied to your post and rehashed some of what you had already said, but it appears brute force of reminding those of how statistics (e.g. political polls) work is the only way I can contribute and see it happening.

I guess, Greggers has more patience than me, has consistently established the rules of the game (statistical protocols) and the same nonsense continues to pop up. What's with that? Learning something new has its on merits even if it's using an example that doesn't please your personal sensibilities.

That's a shame. This site is about "Deny Ignorance". How about we all do something of that, myself included (although this thread isn't one in which I would be denying ignorance outside of pointing out the painfully obvious false assumptions and information being tossed around -- mainly driven by visions/worldviews and not constructive debate tools and/or using math to deny ignorance). Please, anyone feel free to ask questions/comments/concerns and allow me to do my best to help in continuing the denying of ignorance. Good luck and I hope we can continue with "Deny Ignorance" as our impetus to contribute to this thread, and any others as well, of course.

Cheers.

a reply to: conscientiousobserver

You realize Bill gates has "some college or less" right?

You've already been shown to be wrong. Stop doubling down. You are a bigot and a racist.

People from the south are inbred rednecks and education is a white thing. Your words. It didn't take much to expose you for what you are.

Now we can add "liar?" to it because you just completely made up your numbers. Here are the real numbers.

Voters with higher levels of education are more likely to support Democratic presidential nominee Hillary Clinton, while older voters who didn't attend college lean toward Donald Trump, a new poll shows.


According to a Morning Consult poll released Thursday, Clinton leads the Republican presidential nominee by 17 points, 51 to 34 percent, among voters with postgraduate degrees.
Among those with bachelor's degrees, Clinton leads Trump, 45 to 40 percent. Clinton's biggest lead comes from those between the ages of 18 and 34 who have graduated college: 54 percent said they support Clinton compared to the 30 percent who support Trump.

Meanwhile, the poll shows Trump leading Clinton, 49 to 33, percent among voters over the age of 65 who did not attend college.

Of all surveyed, Trump led Clinton by 1 point among those who did not attend college, 41 to 40 percent.

Trump made headlines in February when he said he loved the "poorly educated."

You fell for the blog spin of "some college or less".

When you actually look at the poll being referenced you can see there is a section for "did not attend college". As you can see they are tied in that field.

Some great spinsters decided to merge the categories into "some college or less" to greatly increase Trump's lead in that category and make his followers appear to be uneducated. Apparently it worked on very weak minds...

As we can see though, Hilary has just as much support as Trump among those with high school or less.

originally posted by: BeefNoMeat

A well-constructed survey, employing a simple random effects OLS (ordinary least squares) model can, as described above, use a sample of 1069 respondents to give a pollster a clear, unambiguous result that represents the whole of 320M people. Again, as mentioned, the confidence interval of 95% referred to is solid and the degrees of freedom afforded with said

There are 150 Million registered voters (slightly less, actually), which is the population size for the survey. Otherwise, I agree with your comments and thank you.

a reply to: Greggers

True about the population size, my bad, I was doing back-of-the-envelope math and used the entire population because I knew that number and failed to remember not everyone can vote -- plus I didn't know the entire population of eligible voters and the sample size of 1069 would've captured the same voting split with 320M voters.

I could google it easy (probably will), but do you have an idea of the voter participation rate in Presidential versus mid-term elections? How about the cohort split (e.g. the number of persons over 18 that constitute the voting population) of eligible versus ineligible voters? Again, I can always google it, but I enjoy the conversation so it's more an attempt at extending the discussion...thanks in advance.

originally posted by: Rezlooper

originally posted by: SaturnFX

originally posted by: Bloodydagger
a reply to: SaturnFX

I have over 1000 friends on Facebook. Family, old high school friends, current friends and co-workers etc.

I am an old man (40) and so who I am generally surrounded with in the deep south here are trump supporters.
but
I have good genes (with a bit of prep I can pass for much younger) and my 26y niece is like my sister. She and her fiance are constantly asking me to go out with them, and they have a very wide circle of friends (you know the type..tons of hipster girls and muscle dudes all past their teens but still obnoxiously young) and they are all anti-trump simply because its the in thing.
They dont participate in online polling or any of that nonsense, they aren't locked behind a keyboard because they are typically having one of those stupid social lives.
If I went by stay at home moms and dads that stopped participating in normal social society, sure. it would be a Trump landslide. Thing is, online political activities are most done by old grumps like me, and teens who can't go out past 10 so will yammer on about stuff that is a bit above their knowledge base...the rest are watching netflix during the week and out socializing on the weekend, with occasional visits to snapchat and facebook to discuss trivial issues.

I am trying to spread the good word about how there are 3rd party candidates and you dont need to vote for the lesser demon...nobody listens, nobody cares enough tbh, and frankly they may be right. house/senate can keep a potus in check if they get too crazy.

You're so very wrong to claim twenty-somethings aren't on Facebook. I don't know a one that isn't.

Oh, they are, I was suggesting they aren't deeply involved in the daily political theater. generally speaking they aren't on facebook to find out the truth in politics and stuff..they are on it to share pics and see who is going where tonight.

a reply to: BeefNoMeat

I'd say it could easily be 5% margin of error in each category of voters in the area. Also it will be area-depended and the participant-input number has to be much higher when there are significant differencies between many areas.

Nobody played one of those Black-Jack systems to scalp casino bonus? Your wins and losses will be around 50% but you need to play maybe 100,000 games to get near 50% You can easily lose 1000 bucks even when you play with 2 dollars per game.
When I have a chicken and you don't have a chicken we have a half of a chicken each so shut up and eat

No way a thousand and sixty nine or two thousand people can tell you anything but an informed guess.

a reply to: PapagiorgioCZ

Are we asking for more than an informed guess?

originally posted by: and14263
Why has this thread become so popular?

Very odd.

What's odd when it's popular among internet people?

People feel that the Internet has a huge potential. Like to make the best polls or even to make some third party candidate viral, crowdfunding his/her campaign and after a hundredth monkey effect takes place to give the world an alternative. It can be done in seven days. Just saying. I still kind of like Trump's program.