Pollsters got it wrong in last presidential election

Rasselas

Moderator
Feb 2010
71,597
48,715
USA
Her staff needed a leader that was better strategist.
Possibly. I don't think she was a very strong candidate. But do we want every pol well qualified to actually hold office to also be a master electoral strategist? It's up to handlers and political operatives to lead the candidate to the right decisions. They were campaigning in Arizona and Texas because they had misjudged the electorate--but then EVERYONE misjudged the electorate. Even Trump. That man never wanted to be president and was unable to make sure he didn't win (while doing well enough to go on being a public pain in the ass, which is what he really wanted).
 
  • Like
Reactions: NightSwimmer
May 2007
5,283
2,749
your place
You are right that the polls consistently underestimated Trump's support. Here are some likely reasons for this:
1: Response rates to pollsters are rapidly declining, leading to less and less accurate polls. This is turning polls into self-selecting samples and pollsters will make assumptions about demographics about the people who will vote that can easily be very wrong. If pollster assumed the demographics would be like 2012, but it was more like 2004, then the polls will underestimate Trump's support. If Hillary supporters were more likely to respond to polls, then they will overestimate her support.
2: Trump shame. Many people in the middle personally didn't like Trump and both Trump and Hillary had only about a 40% approval. Many of these people were just undecided and didn't want to publicly say that they would actually vote for Trump over Hillary. So when these undecided people walked into the ballot box and actually had to make a hard choice, they did the unexpected.
3: Last minute Trump surge. At the very end of the election, Trump was enjoying a big surge as his sex-harassment bragging video was wearing off, Hillary wikileaks were coming out, and Hillary's email investigation was re-opened. Most of the polls were not done right on election day, and were done a few days before, which means they couldn't take into account anything that happened between their closing and the election.

I hope we can both agree is that even if a poll is wrong about how much a candidate won a state, what matters is whether they got the end-result right, because that is what really matters. Polls were correct about 77% of swing states, but unfortunately the election was so close that this accuracy wasn't good enough. Another consideration is that the media was predicting a much bigger Trump victory than the polls were saying. The polls were saying the election was a complete toss-up while the media was predicting a 2008-like victory for Hillary.

There have been quite a few posts on this thread where I was told I didn’t understand what margin of error means and that the polls were right when they got it wrong but were within the margin of error. Now you are telling me margin of error doesn’t matter as long as the poll predicted the winner correctly.

Also, the polls were not indicating a “complete toss up”. Clinton was expected to win because of the polls. The media reported the pollster results. Clinton was a huge favorite in Las Vegas also and that was because of the polls. The pollsters even fooled Clinton to believe she had a lock on states and as a result she skipped visiting those states.
 
May 2007
5,283
2,749
your place
Above are video clips from ABC and CBS news talking about how the polls got it wrong
 

Rasselas

Moderator
Feb 2010
71,597
48,715
USA
And she lost the rust belt which has been a Dem stronghold forever. She did not pay attention to a natural Dem base.
Because she couldn't connect to that base on a personal level. She's a cold patrician at heart and it shows from every pore.
 
  • Like
Reactions: bajisima

Rasselas

Moderator
Feb 2010
71,597
48,715
USA
There have been quite a few posts on this thread where I was told I didn’t understand what margin of error means and that the polls were right when they got it wrong but were within the margin of error. Now you are telling me margin of error doesn’t matter as long as the poll predicted the winner correctly.

Also, the polls were not indicating a “complete toss up”. Clinton was expected to win because of the polls. The media reported the pollster results. Clinton was a huge favorite in Las Vegas also and that was because of the polls. The pollsters even fooled Clinton to believe she had a lock on states and as a result she skipped visiting those states.
Indeed. She was considered a winner, though not overwhelmingly and certainly not overwhelmingly at the end. The polls were showing a significant rightward swing in the last couple of weeks. Again, most baseball games don't begin with a homerun by the first batter--that's a pretty unusual occurrence. But are we shocked when it happens? Even the "odds of winning" numbers were never more than 80%. Someone's being left-handed is about half as likely--but are you floored when you see someone writing with their left hand?
 
  • Like
Reactions: Babba
May 2007
5,283
2,749
your place
Indeed. She was considered a winner, though not overwhelmingly and certainly not overwhelmingly at the end. The polls were showing a significant rightward swing in the last couple of weeks. Again, most baseball games don't begin with a homerun by the first batter--that's a pretty unusual occurrence. But are we shocked when it happens? Even the "odds of winning" numbers were never more than 80%. Someone's being left-handed is about half as likely--but are you floored when you see someone writing with their left hand?

If 30 percent of the population is left handed and you are out and about in the world taking note, then it is a certainty that you will see some people writing left handed. However, that isn’t an applicable analogy.

The polling methodology was wrong. Clinton supporters were less enthusiastic and Trump voters were more reliable in showing up. The polling “experts” generally missed these factors and as a result the odds based on their analysis were wrong. Not all pollsters got it wrong but most did. Below is a video of a pollsters explaining why he got it right. Anybody who was perceptive enough could have picked up on the fact that Clinton was not making an emotional connection with voters. Her campaign speeches sounded canned and she repeated things over and over that were not connecting on an emotional level. Trump did a better job of energizing his supporters.


 
Last edited:

Rasselas

Moderator
Feb 2010
71,597
48,715
USA
If 30 percent of the population is left handed and you are out and about in the world taking note, then it is a certainty that you will see some people writing left handed. However, that isn’t an applicable analogy.
It's not 30% left handed, it's 10%. The point is that when something is a 90% certainty (which was never true of the 2016 election) then we aren't surprised when the 10% sometimes comes up. We shouldn't have been so surprised in November 2016/

The polling methodology was wrong. Clinton supporters were less enthusiastic and Trump voters were more reliable in showing up. The polling “experts” generally missed these factors and as a result the odds based on their analysis were wrong. Not all pollsters got it wrong but most did. Below is a video of a pollsters explaining why he got it right. Anybody who was perceptive enough could have picked up on the fact that Clinton was not making an emotional connection with voters. Her campaign speeches sounded canned and she repeated things over and over that were not connecting on an emotional level. Trump did a better job of energizing his supporters.


You can't tell about turnout. It's nearly impossible to gauge who will be there on Election Day, since some people make that decision the day of the election. That's why rain or sunshine is so important--people are actually discouraged from voting by a bit of rain.

That doesn't invalidate polling, though more than one poster here has explained why it's HARDER to be a pollster today than it wasn't 20 years ago. People want from pollsters what they often want from scientists--for them to know everything. They don't. The only real problem here is when pollsters start making claims that they can't support. I don't think that's happened. You should read polling with skepticism--just like everything else.
 
May 2007
5,283
2,749
your place
It's not 30% left handed, it's 10%. The point is that when something is a 90% certainty (which was never true of the 2016 election) then we aren't surprised when the 10% sometimes comes up. We shouldn't have been so surprised in November 2016/

You can't tell about turnout. It's nearly impossible to gauge who will be there on Election Day, since some people make that decision the day of the election. That's why rain or sunshine is so important--people are actually discouraged from voting by a bit of rain.

That doesn't invalidate polling, though more than one poster here has explained why it's HARDER to be a pollster today than it wasn't 20 years ago. People want from pollsters what they often want from scientists--for them to know everything. They don't. The only real problem here is when pollsters start making claims that they can't support. I don't think that's happened. You should read polling with skepticism--just like everything else.

The percentage of people that is left handed is irrelevant. The fact is that if you run a retail business customers have to sign a credit card receipt, if you do a lot of business then over time it is a certainty that you will get some left handed customers.


It is weird that so many people want to defend the pollsters yet make excuses for them being wrong at the same time. If a non-expert conducts a poll, then they will most likely make mistakes in the way they conduct the poll. They will also make mistakes in identifying likely voters. That is what happened in the 2016 election. The pollsters, in general, did not get it right. They failed to recognize that there was less enthusiasm among Clinton supporters.

We as non-experts shouldn’t be required to interpret polling data. That is the job of polling experts. The overall point I am putting forward stands. The pollsters got it wrong. That is something that was reported by ABC and NBC in the YouTube news clips I provided.