can we trust the polls?

Page 1 of 212>
September 12th, 2018 at 3:49:32 PM permalink
reno
Member since: Oct 24, 2012
Threads: 57
Posts: 1043
In final days of the 1980 Presidential campaign, both the national polls and the state polls incorrectly showed a tight race that had Carter and Reagan neck & neck. As it turned out, the polls were off by over 8 points, which is unusually large by historical standards. Reagan won the electoral college 489 to 49.

In the 2004, 2008, 2012, and 2016 Presidential races, the national polls were all off by just 2 or 3 percent. But since Presidential elections are based upon the electoral college, state polls are more important than national polls. That's unfortunate, because state polls are almost always less accurate than the national polls (the most recent exceptions were in 1996 and 1984.) In 2016, the state polls were erroneous by an average of 5.2 percent.

Statistician Nate Silver argues that the general public had gotten spoiled by the unusual accuracy of the 2004, 2008, & 2012 polls which made the 2016 results all the more surprising.

Quote: Nate Silver
As we tried to emphasize before the election, it didn’t take any sort of extraordinary, unprecedented polling error for Trump to defeat Clinton. An ordinary, average polling error would do — one where Trump beat his polls by just a few points in just a couple of states — and that’s the polling error we got.

...the error was more consequential in 2016 than it was in past years, since Trump narrowly won a lot of states where Clinton was narrowly ahead in the polls. By contrast, in 2012, the polls somewhat underestimated Barack Obama’s numbers in several swing states as well as in the national popular vote. (National polls were actually a bit more accurate in 2016 than in 2012.) But it didn’t usually change the winner in these contests — Obama just won them by a clearer margin instead of a narrower one.


The night before the 2016 election, Silver gave Clinton a 71.4% chance of winning the White House. As I write this, Silver is forecasting an 83.1% chance that the Democrats gain control of the House, and a 66.2% chance that the Republicans keep the Senate. Is he right?


September 12th, 2018 at 4:03:10 PM permalink
Dalex64
Member since: Mar 8, 2014
Threads: 3
Posts: 2715
Well let's see. You need to hold the election in millions of parallel universe, record who gains control, and see how close the frequency of who wins compares against the predicted percentage.
"Everyone is entitled to his own opinion, but not to his own facts." Daniel Patrick Moynihan
September 12th, 2018 at 4:55:52 PM permalink
Dalex64
Member since: Mar 8, 2014
Threads: 3
Posts: 2715
Oh and I think a properly conducted poll is plenty accurate.
"Everyone is entitled to his own opinion, but not to his own facts." Daniel Patrick Moynihan
September 12th, 2018 at 5:20:57 PM permalink
Evenbob
Member since: Oct 24, 2012
Threads: 121
Posts: 15782
Quote: Dalex64
Oh and I think a properly conducted poll is plenty accurate.


So why were 99% of the properly
conducted polls in 2016 so wrong,
then. This happens every election
cycle, they talk themselves into
believing election polls, when they
know they can't be trusted.
If you take a risk, you may lose. If you never take a risk, you will always lose.
September 12th, 2018 at 5:54:54 PM permalink
Dalex64
Member since: Mar 8, 2014
Threads: 3
Posts: 2715
Your willfull ignorance is not my problem.
"Everyone is entitled to his own opinion, but not to his own facts." Daniel Patrick Moynihan
September 12th, 2018 at 10:56:39 PM permalink
Fleastiff
Member since: Oct 27, 2012
Threads: 54
Posts: 6502
Dewey Defeats Truman.... banner headline Chicago Tribune poll.

Peace in the Pacific.... banner headline on the most well respected political commentary magazine in the USA: World Report. Date: December 7, 1941.

People who sign petitions for the Save The Skeets Foundation vote in opinion polls as do those who sign petitions to urge Canada to counter the effects of global warming by air conditioning the meeting place of the Canadian Parliament, the National Igloo.
September 13th, 2018 at 11:26:23 AM permalink
Evenbob
Member since: Oct 24, 2012
Threads: 121
Posts: 15782
Quote: Dalex64
Oh and I think a properly conducted poll is plenty accurate.


The 'properly conducted' poll in 1980
had Carter evenly tied with Reagan
the day before the election. Reagan
crushed Carter, got over 400 of the
EC votes. The Lib press was stunned,
a senile clownish cowboy actor was
now president. They never got over
it for 8 years.
If you take a risk, you may lose. If you never take a risk, you will always lose.
September 13th, 2018 at 3:07:23 PM permalink
petroglyph
Member since: Aug 3, 2014
Threads: 18
Posts: 3249
Quote: Dalex64
Your willfull ignorance is not my problem.
From link

"Example: The Catholic church slowly progressing from a concept of “God decides what happens next. We have to propitiate him.” to “God created the laws of physics and logic.” Which allows them to think in terms of logic and physics, while still not calling their old concepts bad, and so come to harmony with reality again."
Everyone gets thrown from the plane to maintain altitude
September 13th, 2018 at 4:37:54 PM permalink
Dalex64
Member since: Mar 8, 2014
Threads: 3
Posts: 2715
Here is a report covering Gallop polls throughout the past 70 years or so regarding church attendance.

https://news.gallup.com/poll/232226/church-attendance-among-catholics-resumes-downward-slide.aspx
"Everyone is entitled to his own opinion, but not to his own facts." Daniel Patrick Moynihan
September 14th, 2018 at 4:09:46 AM permalink
Fleastiff
Member since: Oct 27, 2012
Threads: 54
Posts: 6502
Do you think that the people responding to such polls might have a reason to adopt a posture of "I'm a good Catholic so I don't attend any of these new-fangled English-language services"?
Page 1 of 212>