Trump Senior Advisor: Polls Have Massive Oversampling Of Democrats

[ReubenSherr]

Polls showing big Biden lead 'have massive oversampling of Democrats,' Trump adviser Cortes claims

Less attention should be paid to national presidential election polls than those in individual states, Trump 2020 campaign adviser Steve Cortes said Monday, as his candidate trails Joe Biden in the latest ABC News/Washington Post survey by a dozen points.

www.foxnews.com

Polls that show former Vice President Joe Biden with a big lead over President Donald Trump "have massive oversampling of Democrats," Trump 2020 campaign adviser Steve Cortes argued Monday.

"I think polls matter," Cortes told host Bill Hemmer. "We pay some attention to them. It's very critical to look at the inputs because the inputs often determine the final output numbers.

"Many of these polls have massive oversampling of Democrats," Cortes added.

The facts here?

Several polls showing Biden up, nationally, by double digits have very interesting samples.

An ABC poll showing Biden up 12 gathered data, in part, by calling landlines and specifically asking for the youngest member of the household. A practice not conducted by most other polling firms. In this election cycle, an NBC poll found that Gen Z voters are 4 times as likely to support Biden as average voters.

ABC News' polling methodology and standards

A summary of ABC News polling standards and methodology.

abcnews.go.com

A CNN poll showing Biden up 16 reached a sample, in which 84% of those respondents had watched the Presidential Debate. A debate which was viewed by 22% of Americans on television.

Fewer Americans Watched Trump and Biden's Debate Than Hil...

Viewership for the first presidential debate dropped nearly 15 percent from 2016 to 2020, but the reason why isn't entirely known.

www.newsweek.com

A recent NBC poll showing Biden up 14 reached a sample in which 9% more Democrats were contacted than Republicans.

Meanwhile, Trump's approval rating is ticking up, slightly.

And his RCP average is tightening in many key battleground states, including Pennsylvania, Ohio, and Arizona.

All this doesn't necessarily point to a Trump victory on November 3. But we could be dealing with a great deal of slick political maneuvering in order to drive fundraising into the Biden campaign, or perhaps, suppress fundraising for the Trump campaign.

We'll see on November 3.

 

[d0gbreath]

Forget the polls and vote.

 

[Decypher]

Tell me more, Trump 2020 campaign adviser Steve Cortes.

 

[Mycroft]

I've been telling people on this forum about this oversampling nonsense for years.

 

[Deuce]

I've been telling people on this forum about this oversampling nonsense for years.

Saying it more often doesn’t make you right. You keep comparing an arbitrary party ID poll to other polls and the math doesn’t support you doing that.

Did it occur to you that maybe you are wrong, instead of every single polling group in the country being wrong?

 

[ReubenSherr]

Saying it more often doesn’t make you right. You keep comparing an arbitrary party ID poll to other polls and the math doesn’t support you doing that.

Did it occur to you that maybe you are wrong, instead of every single polling group in the country being wrong?

To be fair, there is a very obvious split. With just a few exceptions.

Polling firms affiliated with liberal universities, and liberal media outlets, nearly always have Biden with the largest leads.

Polling firms affiliated with conservative outlets, or locally owned outlets, nearly always have Biden with the smallest leads. With the exception of a single Rasmussen poll.

 

[ttwtt78640]

Forget the polls and vote.

Yep, that’s the pollIng which actually matters.

 

[Mycroft]

Saying it more often doesn’t make you right. You keep comparing an arbitrary party ID poll to other polls and the math doesn’t support you doing that.

Did it occur to you that maybe you are wrong, instead of every single polling group in the country being wrong?

Gallup is not "arbitrary". They are the gold standard.

 

[beefheart]

Oh gee, a Trump advisor doesn't like the polls.

Color me surprised.

Another steaming shovel full of propaganda and disinfo from Reuben.

 

[Anagram]

Gallup is not "arbitrary". They are the gold standard.

What about their methodology makes them the gold standard to you? Gallup does not poll elections any more after missing pretty badly from 2008-2012 so we have no real tangible standard to judge the accuracy of their polls. What about their methodology compared to other pollster's methodology, makes you trust their party affiliation numbers implicitly over every other pollster?

 

[Threegoofs]

Polls showing big Biden lead 'have massive oversampling of Democrats,' Trump adviser Cortes claims

Less attention should be paid to national presidential election polls than those in individual states, Trump 2020 campaign adviser Steve Cortes said Monday, as his candidate trails Joe Biden in the latest ABC News/Washington Post survey by a dozen points.

www.foxnews.com

Polls that show former Vice President Joe Biden with a big lead over President Donald Trump "have massive oversampling of Democrats," Trump 2020 campaign adviser Steve Cortes argued Monday.

"I think polls matter," Cortes told host Bill Hemmer. "We pay some attention to them. It's very critical to look at the inputs because the inputs often determine the final output numbers.

"Many of these polls have massive oversampling of Democrats," Cortes added.

The facts here?

Several polls showing Biden up, nationally, by double digits have very interesting samples.

An ABC poll showing Biden up 12 gathered data, in part, by calling landlines and specifically asking for the youngest member of the household. A practice not conducted by most other polling firms. In this election cycle, an NBC poll found that Gen Z voters are 4 times as likely to support Biden as average voters.

ABC News' polling methodology and standards

A summary of ABC News polling standards and methodology.

abcnews.go.com

A CNN poll showing Biden up 16 reached a sample, in which 84% of those respondents had watched the Presidential Debate. A debate which was viewed by 22% of Americans on television.

Fewer Americans Watched Trump and Biden's Debate Than Hil...

Viewership for the first presidential debate dropped nearly 15 percent from 2016 to 2020, but the reason why isn't entirely known.

www.newsweek.com

A recent NBC poll showing Biden up 14 reached a sample in which 9% more Democrats were contacted than Republicans.

Meanwhile, Trump's approval rating is ticking up, slightly.

And his RCP average is tightening in many key battleground states, including Pennsylvania, Ohio, and Arizona.

All this doesn't necessarily point to a Trump victory on November 3. But we could be dealing with a great deal of slick political maneuvering in order to drive fundraising into the Biden campaign, or perhaps, suppress fundraising for the Trump campaign.

We'll see on November 3.

Right.

That’s what top advisors of campaigns have to say when their candidate is getting creamed.

Gotta squeeze out that last few weeks of consultant payments and maybe a bonus.

 

[Decypher]

What about their methodology makes them the gold standard to you? Gallup does not poll elections any more after missing pretty badly from 2008-2012 so we have no real tangible standard to judge the accuracy of their polls. What about their methodology compared to other pollster's methodology, makes you trust their party affiliation numbers implicitly over every other pollster?

He likes the results. There’s a couple of self-styled polling experts on this site (and in this thread) who have all the time in the world for one set of polling data but dismiss any other set which doesn’t provide the results they like. It’s almost as if it’s the result they favour rather than the process.

 

[Mycroft]

What about their methodology makes them the gold standard to you? Gallup does not poll elections any more after missing pretty badly from 2008-2012 so we have no real tangible standard to judge the accuracy of their polls. What about their methodology compared to other pollster's methodology, makes you trust their party affiliation numbers implicitly over every other pollster?

They've been doing it for more than 16 years. They get it right. And party affiliation polling is much different than other polling.

But hey...you go ahead and find someone else who does what they do...see if they do it better...and go with them, if you like.

 

[Deuce]

Gallup is not "arbitrary". They are the gold standard.

Doesn’t matter. You’re trying to arbitrarily change reported party affiliation in a poll. You think polls should lie about what respondents said.

 

[Mycroft]

Doesn’t matter. You’re trying to arbitrarily change reported party affiliation in a poll. You think polls should lie about what respondents said.

I'm not changing anything.

It's the pollsters who say, Screw the data on party affiliation. We'll make up our own party affiliation numbers. Who cares if they bear no resemblance to reality. The useful idiots will believe us."

I'm rejecting those pollsters.

 

[ReubenSherr]

I tried to calculate national party affiliation a couple months ago, using voter registration in most states, and 2016 primary registration by party in states which didn't mandate voter ID by party, and there was a 3-4% difference between nationally registered Democrats and Republicans.

Obviously, this was just me doing the calculations.

But NBC literally had 9% more Democrats than Republicans surveyed, there is no way that's accurate. Or even close.

Backing this up is the fact that swing state polling isn't even coming close to mirroring what these double digit national polls are claiming is true about the race.

Selective new polls have Trump down by 2 in Nevada, down by 2 in Pennsylvania, up by 4 in Florida and Arizona.

 

[Anagram]

They've been doing it for more than 16 years. They get it right. And party affiliation polling is much different than other polling.

Why do you think they get it right? The last three elections they participated in were the 2008 presidential election, the 2010 congressional ballot, and the 2012 presidential election. They were either the most inaccurate, or tied for the most inaccurate, pollster in all three among the 15 pollsters in 2008, 7 pollsters in 2010, and 10 pollsters in 2012 that made it to RCP. Those are the only things that Gallup polled that produce tangible numbers. There is no objective standard to check Gallup's numbers against for party affiliation or presidential job approval or any of the many other things they measure. The only ones you can objectively measure, they were bad at the last time produced them.

If you look at the polling in previous years, you would expect to see evidence of pollsters using what you would claim as "bogus numbers" for party affiliation getting inaccurate results. Yet in 2016, all the national polls that were supposedly "wildly oversampling Democrats" ended up pretty close to dead on. There wasn't a single miss from the last week outside a standard deviation. They averaged about a 1% miss.

In 2012, Gallup was again the worst performing pollster overall. Dean Chambers made the same claim that you were making about the polls oversampling Democrats. Source. The final result was that every single pollster ended up underestimating Obama's margin (or having Romney actually up, like Gallup).

So if this is a constant theme among pollsters, why do we not get pollsters consistently overestimating Democratic support? Sometimes there's a slight bias in the Democrats favor like 2016, but just as often it's in the Republicans favor like 2012. This despite those same polls "oversampling" Democrats compared to Gallup.

So I'm honestly asking, why do you think Gallup is the gold standard? I cannot find any data showing them to be right more than any other pollster, nor can I find a consistent track record of pollsters who sample more Democrats than Gallup does being wrong in favor of the Democrats. The data just doesn't support it.

 

[Airyaman]

Unskew the Polls!

 

[Anagram]

I'm not changing anything.

It's the pollsters who say, Screw the data on party affiliation. We'll make up our own party affiliation numbers. Who cares if they bear no resemblance to reality. The useful idiots will believe us."

I'm rejecting those pollsters.

That's not what they're doing. Most pollsters don't weight by party affiliation at all. They weight by race, education, age, etc., and then record the party affiliation of the voters after weighting by the other factors. It is not the other way around.

 

[Aunt Antifa]

Tell me more, Trump 2020 campaign adviser Steve Cortes.

Just imagine if they spent as much time actually trying to win the election vs whining about it.

 

[Anagram]

I tried to calculate national party affiliation a couple months ago, using voter registration in most states, and 2016 primary registration by party in states which didn't mandate voter ID by party, and there was a 3-4% difference between nationally registered Democrats and Republicans.

Obviously, this was just me doing the calculations.

But NBC literally had 9% more Democrats than Republicans surveyed, there is no way that's accurate. Or even close.

Backing this up is the fact that swing state polling isn't even coming close to mirroring what these double digit national polls are claiming is true about the race.

Selective new polls have Trump down by 2 in Nevada, down by 2 in Pennsylvania, up by 4 in Florida and Arizona.

By selective polls you mean Trafalgar, Trafalgar, Trafalgar, and the Nevada Poll commissioned by Sheldon Adelson that polled Nevada without calling cell phones, weighting Hispanics at 13%, and using the pollster used by Adam Laxalt in 2018 that had him winning correct?

I would again point out that state polling done in the last week has Biden improving 2016 democratic numbers by an average of 7%, which would be consistent with about a 9% point lead nationally. But of course, if you only look at the polls in which Trump does better than average, then Trump will do better than average.

 

[Helix]

i hope that the election is an anti-Tweety wipeout.

 

[beefheart]

Womp womp....

 

[beefheart]

By selective polls you mean Trafalgar, Trafalgar, Trafalgar, and the Nevada Poll commissioned by Sheldon Adelson that polled Nevada without calling cell phones, weighting Hispanics at 13%, and using the pollster used by Adam Laxalt in 2018 that had him winning correct?

I would again point out that state polling done in the last week has Biden improving 2016 democratic numbers by an average of 7%, which would be consistent with about a 9% point lead nationally. But of course, if you only look at the polls in which Trump does better than average, then Trump will do better than average.

Reuben is deliberately being dishonest about the polls.
Which figures, he is only here to sow disinformation.

 

[ReubenSherr]

By selective polls you mean Trafalgar, Trafalgar, Trafalgar, and the Nevada Poll commissioned by Sheldon Adelson that polled Nevada without calling cell phones, weighting Hispanics at 13%, and using the pollster used by Adam Laxalt in 2018 that had him winning correct?

Fox Orlando as well.

Keep in mind how accurately Trafalgar called Pennsylvania in 2016.