Welcome to DU! The truly grassroots left-of-center political community where regular people, not algorithms, drive the discussions and set the standards. Join the community: Create a free account Support DU (and get rid of ads!): Become a Star Member Latest Breaking News General Discussion The DU Lounge All Forums Issue Forums Culture Forums Alliance Forums Region Forums Support Forums Help & Search

Yavin4

(35,433 posts)
Wed Nov 4, 2020, 11:05 AM Nov 2020

If you think that polling is bunk, then why are you wearing a mask?

Polling (really sampling) is a cornerstone of the scientific method. The same underlying mathematics that says wearing a mask lowers contagion from covid-19 is the same underlying mathematics that used to sample voter preferences for an upcoming election.

And just like there's a margin of error in the projections of health benefits from wearing a mask, there too is a margin of error in election polling.

EVERY polling outfit got it wrong. Every one of them. And not only did they get it wrong, but their margin of errors always favored Trump. EVERY time.

Now, you could rightly say that the methodolgies of the pollsters are deeply flawed, but we didn't see this in 2018 nor for any other previous election except for 2016. Something is not right with this vote.

Be careful when you dismiss polling because that same argument can be used against you when people question wearing a mask or social distancing. It's all the same math.

13 replies = new reply since forum marked as read
Highlight: NoneDon't highlight anything 5 newestHighlight 5 most recent replies
If you think that polling is bunk, then why are you wearing a mask? (Original Post) Yavin4 Nov 2020 OP
Go back and look at the 2018 polls. Ace Rothstein Nov 2020 #1
Where are the results in our favor? n/t Yavin4 Nov 2020 #2
NV and AZ Senate Ace Rothstein Nov 2020 #5
The MOE swung in Trump's favor in both states. Yavin4 Nov 2020 #8
Polling like all statistics are only as good as your inputs. I felt like FL was going to be hlthe2b Nov 2020 #3
True. That's why a consensus is the key. Yavin4 Nov 2020 #4
And I read too that in 2016 non-college educated people were Laura PourMeADrink Nov 2020 #11
As you say, Wellstone ruled Nov 2020 #9
My apologies to SurveyMonkey. Buckeye_Democrat Nov 2020 #6
While I don't totally dismiss Turin_C3PO Nov 2020 #7
Polling models can be off, though... SidDithers Nov 2020 #10
I read a couple years ago that Trump did have a hidden Laura PourMeADrink Nov 2020 #12
Apples and Oranges StarryNite Nov 2020 #13

Ace Rothstein

(3,160 posts)
1. Go back and look at the 2018 polls.
Wed Nov 4, 2020, 11:09 AM
Nov 2020

A lot of them were actually quite a bit off. Mostly in the favor of Republicans but a few in our favor as well. FL was way off.

Yavin4

(35,433 posts)
8. The MOE swung in Trump's favor in both states.
Wed Nov 4, 2020, 11:37 AM
Nov 2020

Not fully, but so far, he's gotten the lion's share of it.

hlthe2b

(102,225 posts)
3. Polling like all statistics are only as good as your inputs. I felt like FL was going to be
Wed Nov 4, 2020, 11:12 AM
Nov 2020

incredibly difficult to accurately poll because of the volatility and extreme variances in subpopulations.

That hardly means ALL polls are unreliable. Like all science, it depends on the methods, the inputs, and the analysis--but most of all the consistency of validating studies, using similar, but not always identical methods.

Yavin4

(35,433 posts)
4. True. That's why a consensus is the key.
Wed Nov 4, 2020, 11:14 AM
Nov 2020

Not all pollster use the same methodology. Some are landline only. Some are mobile. Some are online polls which are anonymous. All of them had it wrong. All.

 

Laura PourMeADrink

(42,770 posts)
11. And I read too that in 2016 non-college educated people were
Wed Nov 4, 2020, 11:50 AM
Nov 2020

Underrepresented and that not all pollsters had adjusted for this this go around.

 

Wellstone ruled

(34,661 posts)
9. As you say,
Wed Nov 4, 2020, 11:41 AM
Nov 2020

statistics and Algorithms used in polling are totally dependent on the inputs. Here in Nevada we are seeing in real time how data collected missed a critical aspect. The missing key seems to be the Covid impact and how it is and did effect the Dem Party vote totals.

Some how People think the Gaming Industry can just throw the doors open and flip a switch and we are back to pre Covid economic status. And the Votes seem to track that false Narrative. Sad to say,people oft times vote against their best interests.

Turin_C3PO

(13,964 posts)
7. While I don't totally dismiss
Wed Nov 4, 2020, 11:18 AM
Nov 2020

the possibility of vote flipping, I think it’s just hard in this volatile atmosphere to get a true reading from the polls. That doesn’t change the fact that election insecurity has to be addressed.

SidDithers

(44,228 posts)
10. Polling models can be off, though...
Wed Nov 4, 2020, 11:43 AM
Nov 2020

At their heart, polling firms are still making an estimate of mix of D/R/I voters in a voting population, in order to create their modeling.

My concern is that polling models aren't taking into account voter disenfranchisement. If they survey 500 Democrats for a likely voter poll, who fully intend to vote, but because of suppression efforts, only 475 of those Democrats are actually able to cast their ballot, then the poll may be overstating the Democratic response by 2%.



Sid

 

Laura PourMeADrink

(42,770 posts)
12. I read a couple years ago that Trump did have a hidden
Wed Nov 4, 2020, 12:15 PM
Nov 2020

Enthusiasm/GOHV ability that gave him a hidden boost. That he could win a place if he had a 45% (ironic #) approval. That was the magic number. Why I've closely followed his approval #s for four years. Last preelection avg on 538 was 44.6. in the end, this may indeed pan out. He came up slightly shy of 45 and he will end up slightly shy of winning too.

Many of his followers think he is a messiah. And he knows they will do what he asks. Preelection he was asking them to find 9 people and get them to go vote. Like if God came down and appeared to you and asked you to do something, you'd do it.

Latest Discussions»General Discussion»If you think that polling...