On the idea that the polls got the election wrong

[ASUG8]

Yeah JSO, nobody in the media thought she'd win. Enjoy.

phpBB [video]

[Ibanez]

Finally got around to looking directly at what polls were saying immediately before the election and, as I expected, the perception that the polls "got it wrong" is false. Any reasonable interpretation of the body of polling would have led to the conclusion that the election was too close to call.

See the RealClearPolitics poll summary from just before the election at http://www.realclearpolitics.com/epolls ... e_map.html . You can see by clicking on "Virginia" in the table that the last poll it includes was taken November 1 - 6. The RealClearPolitics people projected Clinton getting at least 203 electoral votes and Trump getting at least 164 electoral votes. But 171 electoral votes were up for grabs. Toss up.

The site had 18 jurisdictions indicated as being "solid" for one candidate. The candidate indicated as favored won all 18 of those jurisdictions.

It had 15 jurisdictions indicated as "likely" for one candidate. The candidate indicated as favored won all 15 of those jurisdictions.

It had 6 jurisdictions indicating as "leans" to one candidate. The candidate indicated as the one the State was leaning to won 5 of those jurisdictions.

When you see that margin of error thing with a poll, that's a 95 percent confidence interval. You expect the truth to be outside of the 95 percent confidence interval about 1 in 20 times over a large number of polls. The polls in aggregate were only "wrong" with respect to 1 jurisdiction out of 38. And, really, the 95 percent confidence thing probably only applies to the "solid" category. "Leans" is not a confident assertion.

Meanwhile, you can go to the RealClearPolitics average of polls for the popular vote just before the election at http://www.realclearpolitics.com/epolls ... -5952.html. It has Clinton winning the popular vote by 3.3 percentage points and she won it by 2.1 percentage points. The difference is well within what one should expect it to be.

Basically, if you were trying to anticipate what would happen based on polling results available on the night of November 7 you would say Clinton would win the popular vote but the electoral vote was too close to call.

And it WAS very close. Trump won Wisconsin, Pennsylvania, and Michigan by 0.8, 0.3, and 1.2 percentage points. It's not reasonable to expect polls to be able to pick the winner in States that close.

John, it's been over 2 months. Trump has been inaugurated. You need to let it go. The polls were showing a Clinton victory. They were wrong. Does it matter if they were wrong by .3% or 15%? No, they were wrong. At this point, what difference does it make?

[CAA Flagship]

It was the economy.

[JohnStOnge]

Yeah JSO, nobody in the media thought she'd win. Enjoy.

phpBB [video]

What people in the media do is not the polls' fault. I've criticized the way the media interpret polls before. You know, acting like the fact that a candidate has a 4 point lead in the poll estimate one week then has a 2 point lead in the poll estimate the next week means something has changed.

But to think the polls "got it wrong" this time is to live in an alternative universe. I actually missed one in my first post. The RealClearPolitics people, who presumably know how to interpret the results, identified favorites for 39 jurisdications at three different general levels of confidence (solid, likely, leaning). The favorite won in 38 of 39 of those jurisdictions. That's 97% of jurisdictions that were not identified as "toss up" called correctly. And the one such jurisdiction the favorite didn't win in involved the lowest confidence level (leaning).

It's just ridiculous to say "the polls got it wrong" in this case. It's not even really debatable.

The problem isn't the polls. The problem is the media people you're talking about interpreting them. I also think that maybe there was some bias introduced by the fact that it's hard to believe any thinking person would have voted for Trump. Just hard to believe enough people would do something that stupid to get him elected.

[BDKJMU]

Yeah JSO, nobody in the media thought she'd win. Enjoy.

phpBB [video]

What people in the media do is not the polls' fault. I've criticized the way the media interpret polls before. You know, acting like the fact that a candidate has a 4 point lead in the poll estimate one week then has a 2 point lead in the poll estimate the next week means something has changed.

But to think the polls "got it wrong" this time is to live in an alternative universe. I actually missed one in my first post. The RealClearPolitics people, who presumably know how to interpret the results, identified favorites for 39 jurisdications at three different general levels of confidence (solid, likely, leaning). The favorite won in 38 of 39 of those jurisdictions. That's 97% of jurisdictions that were not identified as "toss up" called correctly. And the one such jurisdiction the favorite didn't win in involved the lowest confidence level (leaning).

It's just ridiculous to say "the polls got it wrong" in this case. It's not even really debatable.

The problem isn't the polls. The problem is the media people you're talking about interpreting them. I also think that maybe there was some bias introduced by the fact that it's hard to believe any thinking person would have voted for Trump. Just hard to believe enough people would do something that stupid to get him elected.

Maybe not in terms of the national polls, but in terms of the 11 (identified by Politico & others) swing states, they did. And the only polls that really mattered were the swing states. You have to pick one of 2 options:
1. The swing state polls got it right.
2. The swing state polls got it wrong.

I don't see how you can pick #1. Which means you're left with #2.

[JohnStOnge]

John, it's been over 2 months. Trump has been inaugurated. You need to let it go. The polls were showing a Clinton victory. They were wrong. Does it matter if they were wrong by .3% or 15%? No, they were wrong. At this point, what difference does it make?

Good grief Ibanez. The polls were not showing a Clinton victory. They were showing too close to call. How do you interpret a situation in which the polls were showing neither candidate having locked up enough electoral votes to win as "showing a Clinton victory?"

And BTW I was posting on this forum to that effect on the night before the election. I was posting that the polls did not provide sufficient evidence to say Clinton was going to win.

You need to separate what some pundits who don't understand survey sampling were saying from what the polls were saying. The polls were NOT saying that Clinton was going to win. They were saying there was not sufficient evidence to make a call either way.

[JohnStOnge]

Maybe not in terms of the national polls, but in terms of the 11 (identified by Politico & others) swing states, they did. And the only polls that really mattered were the swing states. You have to pick one of 2 options:
1. The swing state polls got it right.
2. The swing state polls got it wrong.

I don't see how you can pick #1. Which means you're left with #2.

I went through this earlier: 10 of the 11 swing States you're talking about were indicated as "toss up." How can you be wrong about who wins when your conclusion is that it is a "toss up?"

You don't have to pick one of your two options because, hopefully, you understand that with survey sampling there is a realm in which you have to say you can't say what's going to happen one way or the other.

10 of the 11 swing states you're talking about are states such that no RealClearPolitics "pick" based on polling was made. You can't be wrong when you don't make a pick. And the 11th was a situation in which they just said the State "leans" in one way. That means they were saying it was very "iffy." It wasn't even a State they were willing to say would "likely" go that way.

Why is it so important to you people to believe that the polls "got it wrong" when they didn't?

[JohnStOnge]

Maybe not in terms of the national polls, but in terms of the 11 (identified by Politico & others) swing states, they did. And the only polls that really mattered were the swing states. You have to pick one of 2 options:
1. The swing state polls got it right.
2. The swing state polls got it wrong.

I don't see how you can pick #1. Which means you're left with #2.

I went through this earlier: 10 of the 11 swing States you're talking about were indicated as "toss up." How can you be wrong about who wins when your conclusion is that it is a "toss up?"

You don't have to pick one of your two options because, hopefully, you understand that with survey sampling there is a realm in which you have to say you can't say what's going to happen one way or the other.

10 of the 11 swing states you're talking about are states such that no RealClearPolitics "pick" based on polling was made. You can't be wrong when you don't make a pick. And the 11th was a situation in which they just said the State "leans" in one way. That means they were saying it was very "iffy." It wasn't even a State they were willing to say would "likely" go that way.

Why is it so important to you people to believe that the polls "got it wrong" when they didn't?

[YoUDeeMan]

The polls got it wrong.

[JohnStOnge]

The polls got it wrong.

Why is it so important to you to say that when it's so obvious that it isn't true?

I think it must be a thing associated with Trump support. Saying things that aren't true seems to be a big part of the Trump phenomenon.

[cx500d]

Finally got around to looking directly at what polls were saying immediately before the election and, as I expected, the perception that the polls "got it wrong" is false. Any reasonable interpretation of the body of polling would have led to the conclusion that the election was too close to call.

See the RealClearPolitics poll summary from just before the election at http://www.realclearpolitics.com/epolls ... e_map.html . You can see by clicking on "Virginia" in the table that the last poll it includes was taken November 1 - 6. The RealClearPolitics people projected Clinton getting at least 203 electoral votes and Trump getting at least 164 electoral votes. But 171 electoral votes were up for grabs. Toss up.

The site had 18 jurisdictions indicated as being "solid" for one candidate. The candidate indicated as favored won all 18 of those jurisdictions.

It had 15 jurisdictions indicated as "likely" for one candidate. The candidate indicated as favored won all 15 of those jurisdictions.

It had 6 jurisdictions indicating as "leans" to one candidate. The candidate indicated as the one the State was leaning to won 5 of those jurisdictions.

When you see that margin of error thing with a poll, that's a 95 percent confidence interval. You expect the truth to be outside of the 95 percent confidence interval about 1 in 20 times over a large number of polls. The polls in aggregate were only "wrong" with respect to 1 jurisdiction out of 38. And, really, the 95 percent confidence thing probably only applies to the "solid" category. "Leans" is not a confident assertion.

Meanwhile, you can go to the RealClearPolitics average of polls for the popular vote just before the election at http://www.realclearpolitics.com/epolls ... -5952.html. It has Clinton winning the popular vote by 3.3 percentage points and she won it by 2.1 percentage points. The difference is well within what one should expect it to be.

Basically, if you were trying to anticipate what would happen based on polling results available on the night of November 7 you would say Clinton would win the popular vote but the electoral vote was too close to call.

And it WAS very close. Trump won Wisconsin, Pennsylvania, and Michigan by 0.8, 0.3, and 1.2 percentage points. It's not reasonable to expect polls to be able to pick the winner in States that close.

The la times poll consistently showed trump up.

http://graphics.latimes.com/usc-preside ... dashboard/



Sent from my iPhone using Tapatalk

[AZGrizFan]

Yeah JSO, nobody in the media thought she'd win. Enjoy.

phpBB [video]

That is just fucking delicious to watch their collective heads explode.

[YoUDeeMan]

The polls got it wrong.

Why is it so important to you to say that when it's so obvious that it isn't true?

I think it must be a thing associated with Trump support. Saying things that aren't true seems to be a big part of the Trump phenomenon.

Most of the polls said that Hillary was ahead and was the favorite to win.

They were wrong.

The exit polls were also wrong.

How hard is that to understand.

[Ibanez]

John, it's been over 2 months. Trump has been inaugurated. You need to let it go. The polls were showing a Clinton victory. They were wrong. Does it matter if they were wrong by .3% or 15%? No, they were wrong. At this point, what difference does it make?

Good grief Ibanez. The polls were not showing a Clinton victory. They were showing too close to call. How do you interpret a situation in which the polls were showing neither candidate having locked up enough electoral votes to win as "showing a Clinton victory?"

And BTW I was posting on this forum to that effect on the night before the election. I was posting that the polls did not provide sufficient evidence to say Clinton was going to win.

You need to separate what some pundits who don't understand survey sampling were saying from what the polls were saying. The polls were NOT saying that Clinton was going to win. They were saying there was not sufficient evidence to make a call either way.

You're wrong.

[houndawg]

The polls got it wrong.

Why is it so important to you to say that when it's so obvious that it isn't true?

I think it must be a thing associated with Trump support. Saying things that aren't true seems to be a big part of the Trump phenomenon.

You remind me of those Japanese soldiers that kept turning up on Pacific islands in the 80s. You lost John.

The good news is that you're only an assassination away from your heart's desire of an evangelical wackjob in the White House.

[ASUG8]

If the media guys in NYC and LA walked outside their studios and asked 1,000 normal "people on the street" who they'd pick and try to extrapolate that two highly liberal voting bases would correlate with results from those of us in the flyover states then they would absolutely get it wrong. This election more than any other in my memory demonstrated that the coasts don't control the results.

[ASUG8]

That is just fucking delicious to watch their collective heads explode.

If anyone on the View was on fire I wouldn't waste my urine to piss the fire out.

[JohnStOnge]

The la times poll consistently showed trump up.

http://graphics.latimes.com/usc-preside ... dashboard/



Sent from my iPhone using Tapatalk

The problem with that is that's a popular vote poll. And Trump lost the popular vote.

[YoUDeeMan]

The la times poll consistently showed trump up.

http://graphics.latimes.com/usc-preside ... dashboard/



Sent from my iPhone using Tapatalk

The problem with that is that's a popular vote poll. And Trump lost the popular vote.

Awesome...that one was wrong along with all the others.

[JohnStOnge]

Since I did this today I'll post it. I looked at where the popular vote count was around mid day. It was at Hillary Clinton 48.2%, Donald Trump 46.1% . I did a simulation where I selected 10 independent random samples of 1000 each from a population where 48.2% voted one way and 46.1% voted the other way.

I got poll results ranging from Trump having the edge by 3 percentage points (48.3% to 45.3%) to Clinton winning by 10.1 percentage points (52.0% to 41.9%). And that is an idealized situation. That's a situation in which pollsters would be able to identify everyone who would actually vote then have every single one of the ones they selected for the sample cooperative respond.

Most people just do not understand what the situation is with survey sampling. They don't know how to interpret sampling error involved even with a perfect survey.

The fact that the polls as interpreted by the RealClearPolitics people were 38 of 39 in States where they actually indicated a favorite is absolutely incredible. People ought to be lauding them for what a good job they did instead of expressing their ignorance by saying "they got it wrong."

[YoUDeeMan]

Since I did this today I'll post it.

Most people just do not understand what the situation is with survey sampling. They don't know how to interpret sampling error

Apparently, you do not know the meaning of the word, "error".

Here, let me help you:

er·ror
ˈerər
noun
a mistake.
"spelling errors"
synonyms: mistake, inaccuracy, miscalculation, blunder, oversight;
the state or condition of being wrong in conduct or judgment.

synonyms: wrongly, by mistake, mistakenly, incorrectly;


Yup...the polls said Hillary was ahead.

The polls made a mistake.

The polls were inaccurate.

The polls made a miscalculation.

The polls made a blunder.

Th polls were in a state or condition of being wrong in judgment.

The pools were wrong.


Thanks, JSO, for working long and hard to confirm what we already knew.

[cx500d]

The la times poll consistently showed trump up.

http://graphics.latimes.com/usc-preside ... dashboard/



Sent from my iPhone using Tapatalk

The problem with that is that's a popular vote poll. And Trump lost the popular vote.

I think they properly discounted the overweighted ca effect in the election total vote and polled using a weighting approximating the electoral vote distributions.


Sent from my iPhone using Tapatalk

[JohnStOnge]

Most of the polls said that Hillary was ahead and was the favorite to win.

They were wrong.

The exit polls were also wrong.

How hard is that to understand.

I am not the one having a hard time understanding what the polls supported saying.

The polls as interpreted by RealClearPolitics did not indicate that either candidate was ahead because there were 171 votes associated with "toss up" jurisdictions. The polls were not able to provide sufficient evidence to even say those jurisdictions were "leaning" one way or another. In a situation like that you don't KNOW who is ahead in the electoral college race. It is, again, too close to call.

Meanwhile, the national popular vote polls did generally show Clinton ahead. I don't know what the confidence level associated with that was. Don't know whether or not there was sufficient evidence to make the call. But she did win the popular vote.

[JohnStOnge]

I am wondering if it's gotten to the point where you guys REALLY don't understand the nature of survey sampling or if you're just jerking chains. But I'll try again.

I found something rare in one story on a national popular vote poll. It's a reference to confidence. It's the last ABC News Washington Post tracking poll (https://www.washingtonpost.com/news/the ... 71de967b5d).

The poll estimated Clinton to be 4 percentage points ahead. But the author of the article wrote this:

Clinton’s edge in the Post-ABC poll does not reach statistical significance

People see a poll like that ABC News Washington Post tracking poll, with Clinton at an estimated 47 percent and Trump at an estimated 43 percent, and they think that means the poll says Clinton is ahead. But it does not. That's why I grit my teeth every time I see pundits during election season hanging on every poll percentage point and saying stuff like "candidate X's lead grew since this poll was taken last month" when the change was something like 48% to 44% vs. 49% t 43%.

Random sampling error such as that illustrated by the simulation I described a few posts ago involves a pure theoretical situation. It is like if the pollster could 100% accurately identify everyone who is going to vote, be 100% sure of being able to contact each one of those people if they are selected for the sample, and be 100% sure that they will answer the questions when they are contacted. And, yes, 100% sure they are going to answer the questions honestly. Obviously that doesn't happen.

For the body of polling to enable the RealClearPolitics people to have the "favorite" win in 38 of the 39 jurisdictions they felt they could designate...with the only "miss" being a jurisdiction in the least confident "lean" category...is remarkable. And to say "the polls got it wrong" under such circumstances is just ridiculous.




Yet, as interpreted by the RealClearPolitics people.

[CAA Flagship]

It was the economy