Polling, Will it ever be reliable again.

Does polling matter

  • Total voters
    60
#26
#26
LouderVol said:
I think for some of these polls there is little difference, now, between them and the MSM they have worked with for however long.
Click to expand...
I can’t say you’re wrong and I would be lying if I said I didn’t think there was some confirmation bias going on with the individuals involved in a lot of these polls.

But like I’ve said because of the day job I try to give the rank and file pollsters the benefit of the doubt that they take their job seriously that’s all.

Here’s another example with the current fiasco going on. Election officials and vote counting. Specifically PA. Look at all of the accusations flying of fraud. While I’ve been happy to stoke the flames a bit myself if you press me I’m going to admit that I largely believe these people are behaving professionally and take their task seriously and do their best to conduct themselves accordingly. I’m sure in Philly county most vote tabulators cringe when he/she sees a vote for Trump but they none the less correctly tabulate that vote.

If we largely believe the voting is corrupt then we’re a banana republic. It’s as simple as that 🤷‍♂️

BTW I voted no above. In their current state they are clearly broken.
 
  • Like
Reactions: Orangeslice13
#27
#27
Yes and no... just take what they say and assume the margin of error is more like 10-12pts instead of 4.

It seems like the two campaign’s internal polling was pretty accurate though. Biden camp knew it was tighter than media polls reflected and Trump camp knew his supporters in rural areas were under represented.

FWIW though even with the high turnout on both sides, Biden over performed almost everywhere even while losing ground with certain demographics (which was predicted by the polls).
 
#29
#29
This question is like asking if the news will ever be reliable again. No, it’s Dem agit prop. Nate Silver is a grifter who drives clicks for his website. The individual campaigns clearly don’t trust the polls, why should we?
 
  • Like
Reactions: tbwhhs and spyfish007
#31
#31
golfballs said:
This question is like asking if the news will ever be reliable again. No, it’s Dem agit prop. Nate Silver is a grifter who drives clicks for his website. The individual campaigns clearly don’t trust the polls, why should we?
Click to expand...
Agree Silver is toast and needs to refresh his burger flipping skills. He is a one hit wonder and his 15 minutes have long passed.
 
  • Like
Reactions: tbwhhs and spyfish007
#34
#34
burntorangeVOLffle said:
Yes and no... just take what they say and assume the margin of error is more like 10-12pts instead of 4.

It seems like the two campaign’s internal polling was pretty accurate though. Biden camp knew it was tighter than media polls reflected and Trump camp knew his supporters in rural areas were under represented.

FWIW though even with the high turnout on both sides, Biden over performed almost everywhere even while losing ground with certain demographics (which was predicted by the polls).
Click to expand...
How can you possibly even try to make that statement with the figure put forth in golf balls post?! 😳
70E90194-E0CD-4B6E-AB6E-5DAECD380F52.png
 
#35
#35
NorthDallas40 said:
How can you possibly even try to make that statement with the figure put forth in golf balls post?! 😳
View attachment 320690
Click to expand...

The polls predicted Biden losing ground with certain demos. Mostly Cuban Americans and AA males while gaining ground with White Males w/o college degree. That doesn’t mean either candidate WON those demos. The polls did show a shift though.

On a whole, this is the popular vote swing margin thus far. Still a lot of counting left in a bunch of states so it’s fluid. Right now the trend is Biden over performing the 2016HRC margins even in the counties/states he lost.

C6A880EA-3FEE-4495-B65C-EC2A09A3D530.jpeg

You can get into the weeds here if you want...

2020 Popular Vote Tracker | The Cook Political Report
 
#36
#36
burntorangeVOLffle said:
The polls predicted Biden losing ground with certain demos. Mostly Cuban Americans and AA males while gaining ground with White Males w/o college degree. That doesn’t mean either candidate WON those demos. The polls did show a shift though.

On a whole, this is the popular vote swing margin thus far. Still a lot of counting left in a bunch of states so it’s fluid. Right now the trend is Biden over performing the 2016HRC margins even in the counties/states he lost.

View attachment 320694

You can get into the weeds here if you want...

2020 Popular Vote Tracker | The Cook Political Report
Click to expand...
The graph above gives me all the weed I need these polls as a whole appear to be as or more inaccurate than 2016 what your pushing frankly sounds like a damage control narrative by the polling firms to me.
 
#37
#37
NorthDallas40 said:
The graph above gives me all the weed I need these polls as a whole appear to be as or more inaccurate than 2016 what your pushing frankly sounds like a damage control narrative by the polling firms to me.
Click to expand...

LOL ok man... I acknowledged the polls as a whole are off. I said expect more of a 10-12 MoE. Which is about what we are seeing on the Presidential polls. National pop vote will end up being more like 4-5 points off which is pretty close to the MoE just like in 2016.

The Senate and House races are even more off though. I have no explanation for those.
 
#39
#39
The fundamental problem seems to be national firms trying to run state polls.

Ann Seltzer’s Iowa polls have nailed the past two presidential elections and I think she got the first count of the Dem Primary Caucus pretty close, even though her last poll went unpublished.

That model of having someone in the state who knows what is going on on the ground and understands the demographics just seems superior to some firm in New York or wherever looking at data and making guesstimated judgments about how many of each demo will turn out.
 
  • Like
Reactions: Sandman 423 and LouderVol
#40
#40
burntorangeVOLffle said:
LOL ok man... I acknowledged the polls as a whole are off. I said expect more of a 10-12 MoE. Which is about what we are seeing on the Presidential polls. National pop vote will end up being more like 4-5 points off which is pretty close to the MoE just like in 2016.

The Senate and House races are even more off though. I have no explanation for those.
Click to expand...
A 10-12 MoE is a waste and indicates they should be thrown in the trash can then. How’s that.

And the last sentence in the first paragraph is indeed a damage control sentence. A national poll on popular vote is a meaningless statement. Individual polls for each state to predict the winner and then allocate the EC votes is the only way to predict a winner. The method your trying to show support for doesn’t have anything to do with electing the president and never has.

It’s pretty clear why the Senate and House polls are so inaccurate. The polls as compared to the results are ****.
 
#41
#41
RockyTop85 said:
The fundamental problem seems to be national firms trying to run state polls.

Ann Seltzer’s Iowa polls have nailed the past two presidential elections and I think she got the first count of the Dem Primary Caucus pretty close, even though her last poll went unpublished.

That model of having someone in the state who knows what is going on on the ground and understands the demographics just seems superior to some firm in New York or wherever looking at data and making guesstimated judgments about how many of each demo will turn out.
Click to expand...
Sounds like a reasonable approach to me. Have pollsters who have an actual clue what the electorate is thinking. Because that clearly isn’t the case right now.
 
#43
#43
NorthDallas40 said:
A 10-12 MoE is a waste and indicates they should be thrown in the trash can then. How’s that.

And the last sentence in the first paragraph is indeed a damage control sentence. A national poll on popular vote is a meaningless statement. Individual polls for each state to predict the winner and then allocate the EC votes is the only way to predict a winner. The method your trying to show support for doesn’t have anything to do with electing the president and never has.

It’s pretty clear why the Senate and House polls are so inaccurate. The polls as compared to the results are ****.
Click to expand...

Fine... throw national poll out.

If you are looking for the polls to tell who the winner would most likely be based on EC, the vast majority said Biden. It appears at the moment to be the case.

Be mad at the margins I guess.
 
#45
#45
burntorangeVOLffle said:
Fine... throw national poll out.

If you are looking for the polls to tell who the winner would most likely be based on EC, the vast majority said Biden. It appears at the moment to be the case.

Be mad at the margins I guess.
Click to expand...
Welcome to the point of the discussion! And thanks for providing a shining example of the confirmation bias narrative!
 
  • Like
Reactions: Orangeslice13
#47
#47
NorthDallas40 said:
A 10-12 MoE is a waste and indicates they should be thrown in the trash can then. How’s that.

And the last sentence in the first paragraph is indeed a damage control sentence. A national poll on popular vote is a meaningless statement. Individual polls for each state to predict the winner and then allocate the EC votes is the only way to predict a winner. The method your trying to show support for doesn’t have anything to do with electing the president and never has.

It’s pretty clear why the Senate and House polls are so inaccurate. The polls as compared to the results are ****.
Click to expand...
A 5 MoE would mean they only got 17 right. Only 1 wrong for biden, and then 32 wrong for Trump. That's staggering.

A 12 MoE spread would cover about 30 of the states. And that's a HUGE lot of variability, which like you said is useless.

Now that I think of that a 5 MoE would be +2.5 to -2.5, right? If so they did even worse than what I had.
 
#48
#48
LouderVol said:
A 5 MoE would mean they only got 17 right. Only 1 wrong for biden, and then 32 wrong for Trump. That's staggering.

A 12 MoE spread would cover about 30 of the states. And that's a HUGE lot of variability, which like you said is useless.

Now that I think of that a 5 MoE would be +2.5 to -2.5, right? If so they did even worse than what I had.
Click to expand...
Yep. You just summarized how damn worthless these polls have been the last two election cycles.

People need to ignore them and ask two questions.

Do I want to vote or not?
Who do I want to vote for?

The rest is irrelevant
 
  • Like
Reactions: tbwhhs, Sandman 423 and Orangeslice13
#49
#49
NorthDallas40 said:
Welcome to the point of the discussion! And thanks for providing a shining example of the confirmation bias narrative!
Click to expand...

What exactly are you upset about? In your previous post you wanted a basic EC winner/loser predicted by the polls to be accurate. Well you got it this year. Give or take 1-2 states the ultimate result is the same.

If you are looking for an exact science of margin you never had that anyway.

And remember, votes are still being counted in other states than PA, NV, AZ, and GA right now. All across the country states are still counting including the “toss ups” TX and FL. While it appears Trump’s lead is too big to make up the difference in those places, the margins COULD still get a little closer to the polling MoE. Will they fall in it? Or just outside of it? Remains to be seen.
 
New posts
You must log in or register to reply here.

VN Store



Back
Top