What Happened? How the Polls got Hillary's Defeat so Wrong.

So what?

 

So in time the lefts wounds will heal.

But not today.

 

The commission investigating voter fraud has barely begun it's work and already 20 (to 44) states are refusing to grant (or fully comply with) information requested of them. Why? What do they have to hide?

Trump voter fraud commission: Here's what to know

Study supports Trump: 5.7 million noncitizens may have cast illegal votes

 

Just because a poll is wrong doesn't mean they have an agenda. Pollster don't like to look like fools and want to be as accurate as possible especially compared to the competition. No pollster wants to be laughed at like they were after the 2016 election no matter their political agenda. Another thing is that both right-wing and left-wing polls were wrong about the election so you can't just blame the liberal agenda for this one.

The biggest problem facing polling today is that people often do not pick up their phones nowadays and not so many have land-line phones so pollsters have to do a lot of demographic adjusting to their data to make it more accurate but that is risky because it involves interpretation.

Polls like realclearpolitics aren't so good at picking up on last-minute shifts because they average polls and it isn't like 10 polls come out right on election day in every swing state and they usually have to average earlier polls because of a lack of them. But this is better than looking at only individual polls because these have a wide margin of error. For example in 2012 the realclearpolitics polls predicted that Obama and Romney were neck and neck in the popular vote but Obama went on to win by a comfortable 4% margin. This is partially due to him having a last-minute surge. Trump also had a last-minute surge thanks to Comey reopening the email investigation.

Another issue is that pollsters aren't so good at predicting swings in voter participation. In 2012 Obama's victory was underestimated in part because he was able to get unexpectedly high voter turnout among minorites and young people like 2008 when many predicted that the passion had died down and it would be more like 2004. Trump got some unexpectely high turnout among poor whites and this helped him win.

You attack realclearpolitics for being inaccurate but it only predicted a Hillary victory by a 272 - 266 margin which is extremely slim and they admitted that 138 were total tossups and could go either way and almost all the states that unexpectedly went to Trump were marked as being tossups. Realclearpolitics predicted the election was a tossup and they were right.

 

Yeah, but the rest of the year they had her winning by huge margins in the EC.

Pollsters don't care that much about their reputations, otherwise, they would not put out polls showing Hillary winning by 15% 3 months before the election. That outlier should have been thrown out or they should have examined their internals. They don't just randomly call people, they decide who to count based on race, religion, region and other factors, and throw out what they believe to be outliers.

 

Its pretty obvious that you don't understand how polls work. Polls before the election do not measure the margin of victory, they simply measure who is leading at the time and that can easily change. And Hillary was not leading by 15% 3 months before the election (early July), she was leading by 4.5%. This crumbled to eventually Trump leading by 1% in late July when the Republican convention started. She then shot up to an 8% lead in early August thanks to those lewd Trump tapes but this crumbled to only a 1% lead in the middle of September as a bunch of leaks started coming out. In the middle of October she gained a 6% lead thanks to the debates but this crumbled to a 1.5-3% lead because of last-minute leaks and Comey reopening the email investigation and she won the popular vote by 2% but thanks to razor-thin Trump victories in a bunch of swing states she lost the electoral vote.

Every professional cares about their reputations because its their careers on the line and bad performance from them can hurt them and their companies. Everyone cares about their professional reputations? Would you like the be known in your company as a screw-up? I thought not and nobody else does either. And why would republican pollsters forge polls to make it look like Hillary was going to win when they hate Hillary?

 

I know how they work and that is why I don't believe them.

 

I really don't think you do.

 

The pollsters are a part of the MSM and the MSM doesn't seem to care much about their reputations as journalists as the word objectivity has been deleted from their vocabulary.

 

so you can't support your claim. Thank you.

 

You can't assume that everyone in the media is dishonest. Pollsters aren't journalists, they are statisticians and are responsible for making accurate polls.

 

This is how polling science works these days:

“Inside scoop on Silver's forecasting model

“IN CASE I, NATE SILVER, DIE, FOLLOW THESE STEPS TO UPDATE THE FIVETHIRTYEIGHT ELECTIONS FORECAST MODEL” writes Tao Yang in McSweeney's:

-"Navigate to these coordinates (40.731014, -74.303500) on the night of a waxing moon. Bring a white goat. Once there you will see a keypad.”

-Enter the probability of a sample mean being two standard deviations from the population mean at a 95% confidence level.

-To enter the chamber where the model resides you must pass three trials: shouting out certain baseball metrics; find four stone tablets with a scatterplot etched on each and press the one with a correlation of 0.6; and then find an old Compaq desktop in the final of three chambers, turn it on and boot up to PokerStars with a $10 bankroll you turn into $10,000 to enter the final chamber.

Finally, select 'Update Model' and find an altar appearing behind you. "Chant the new poll results while performing a blood sacrifice with the white goat on the altar.”
VANITY FAIR/POYNTER, THE PRESS BLAMES ITSELF FOR TRUMP’S WIN, As election polling blows it, press self-flagellation begins. BY JAMES WARREN, CHIEF MEDIA WRITER, POYNTER.ORG, NOVEMBER 10, 2016.
http://www.vanityfair.com/news/2016/11/the-press-blames-itself-for-trumps-win

 

Pollsters measure popular views, not electoral dissections. The polls predicted nationally that more voters supported Clinton, and they did.

 

Still didn't stop a lot of media giants and others from trying though.

 

If 3 different articles on the subject is your definition of not supporting my claim. You are welcome.

 

Statisticians, Climatologists, whatever the profession, if they have a political agenda, they won't let science get in the way of their determination to sway public opinions.

 

They were wildly inaccurate. You should read the OP.

 

Statisticians belong to all sort of professions and news organizations both right-wing and left-wing. Statisticians are afraid of their numbers coming out as inaccurate especially if it is shown to be a result of sloppy work. I know because I studied to be one. Statisticians are big math nerds and we see democratic and republican statisticians, Hillary supporters and Trump supporters.

 

Yet another objective open minded liberal.....

I love how when it suits you use the gay thing as a slur.

You are all such hypocrites...

 

I'll agree, but in the case of the ones working on these polls, they got it terribly wrong. They were still expecting a Hillary Clinton win at 8pm EST.

 

No, the polls predicted a tossup. There was no clear leader in the swing state polls.

 

Please don't try to re-write history.

The vast majority of polls and pundits predicted an easy Clinton victory. Only a few disagreed. They are the ones everyone is looking at now.

PRESIDENT TRUMP made absolute fools out of Hillary and everyone who laid at her feet. That included the democrats, the mass hysteria (I mean media) and many on this website--many of whom seem to have dropped out.

Comedian Steven Colbert went so far as to host a live Clinton victory party on Showtime last election night. He and others were laughing about a Trump defeat right up to the moment Trump won Florida.

It's a kick to see pompous asses get egg in their face.

Colbert has been leading the sore loser brigade ever since.

 

This is from 2014, but it shows how widespread the problem is: "A quick calculation, as a point of reference. This local reporter found 94 illegally registered voters in one small region using one narrow verification method. If you extrapolate his number over Florida's 67 counties, that's nearly 6,300 people." And these were the voters Democrats weren't using to cheat with! https://townhall.com/tipsheet/guybe...-dozens-of-illegal-voters-in-florida-n1811547

From 2016: "Project Veritas and its founder, James O’Keefe, have released an undercover video of Alan Schulkin, the New York Democratic commissioner of the Board of Elections, confirming there is widespread fraud. In the video, he is heard disclosing that organizers use buses to haul people from poll to poll to vote." http://www.wnd.com/2016/10/democrat-confirms-rampant-voter-fraud/

And from this year about the November 2016 election: "William Baer, who stated that during his time as a poll watcher, he saw people being bussed in to polling stations. If Baer’s remarks were to be proven true, they would give credence to President Donald Trump’s concerns about voter fraud in the 2016 election, which Democrats have been very quick to denounce." https://milo.yiannopoulos.net/2017/06/cnn-edit-voter-fraud/

I think it's pretty obvious that when Democrats oppose voter ID laws, their concern isn't blacks being disenfranchised, their concern is having more difficulty in cheating. Because if they can't cheat, they can't win.

 

I read it. The polls had Clinton a point or two ahead on election day, which equates to the actual vote.

 

Sure, nationally. But they always favored Clinton for 18 months. And as I pointed out, they had it wrong in 8 battleground states that President Trump won by as much as 10%. The national polls had Clinton leading by huge margins in the weeks and months leading up to the election, not that the national polls mean anything, because we live in a republic, not a democracy.