National Polls.It Comes Out To About 50?,60? Votes Per State.So How Can They Be Accurate?

LeeOnLido

Gold Member
Jul 1, 2018
5,703
1,675
190
This is one reason why we cant rely on National Polls. Look at how many are polled. 2000? 3000? so divide that among 50 states. What does it come out to? between 40 to 60 votes a state? and thats what we are supposed to go by? Even if they polled 5000 people, its still miniscule to how many were polled in each state. But of course, we have to keep in mind that they are polling a few thousand out of about 130 to 140 million potential voters.
:1peleas:
 
Did no one learn a lesson from 2016? The accuracy of these polls are fucked. I'd trust a psychic fortune teller more than these polls.
 
It's accurate because they tell you it's accurate(to within 3%+or -)
i wonder if they poll some of those 5000 people who live under the sinking 53rd state of Guam
Most of those people don't have underwater phones...
i wonder how many alligators in florida get these calls on thier cell phones, for all we know they tell the pollster that they havent decided yet and they have to go because its lunch time
 
It's accurate because they tell you it's accurate(to within 3%+or -)

That error bar DEPENDS on them doing about a dozen things RIGHT.. Like GUESSING the mix of Repubs and Dems.. Which they now suck at because increasingly, voters are FLEEING both parties as loyalists. And they are NOT generally weighting Independents. Also depends on the accuracy of them vetting and weighting by age, registration, and voting frequency..

So -- in short -- it's junk with that SMALL a population...

Because it's WORSE then just 50 or 60 people per state. Because 1/2 of their poll population is in California, New York/Jersey and Florida if they do it right. Montana probably gets one or two people if any...
 
This is one reason why we cant rely on National Polls. Look at how many are polled. 2000? 3000? so divide that among 50 states. What does it come out to? between 40 to 60 votes a state? and thats what we are supposed to go by? Even if they polled 5000 people, its still miniscule to how many were polled in each state. But of course, we have to keep in mind that they are polling a few thousand out of about 130 to 140 million potential voters.
:1peleas:

It’s accurate. Ask Rachel Maddow. He is the one who pretty much guaranteed Hildo would be the Orange Clown based on those very polls.
 
This is one reason why we cant rely on National Polls. Look at how many are polled. 2000? 3000? so divide that among 50 states. What does it come out to? between 40 to 60 votes a state? and thats what we are supposed to go by? Even if they polled 5000 people, its still miniscule to how many were polled in each state. But of course, we have to keep in mind that they are polling a few thousand out of about 130 to 140 million potential voters.
:1peleas:

It’s accurate. Ask Rachel Maddow. He is the one who pretty much guaranteed Hildo would be the Orange Clown based on those very polls.
so whats the latest CNN Poll? Biden 547 electoral votes,,,,,Trump 3?
 
This is one reason why we cant rely on National Polls. Look at how many are polled. 2000? 3000? so divide that among 50 states. What does it come out to? between 40 to 60 votes a state? and thats what we are supposed to go by? Even if they polled 5000 people, its still miniscule to how many were polled in each state. But of course, we have to keep in mind that they are polling a few thousand out of about 130 to 140 million potential voters.
:1peleas:
its called statistics and yes, national polling is very accurate because the source size is so large that most variables can be averaged out. The problem with polling is NOT small sample sizes or accuracy, such polling works well on 300 million Americans that are extremely diverse in virtually all aspects of life but can be reasonably averaged. This does not seem to be the case on smaller scales in individual states. The relationship of population size and polling sample size is not linear by any means. It would be almost impossible to accurately assess the voting result of 100 people unless you spoke to damn near all of them - the variables are just to massive to average out. The same is simply not true for a sample size of 300 million. It was those local polls that were grossly inaccurate last cycle.

The problem for most of the current polls cited is they are national and the president is not elected nationally. People seem to have a real problem getting that through their head.

To pretend that the current polls do not reveal a problem with Trump's coming election though is asinine. The numbers do not look good because they are simply not good. That is obvious. Of course that does not mean Trump will not win, he has not even started campaigning in any serious fashion, but he does have a tough row to hoe.
 
This is one reason why we cant rely on National Polls. Look at how many are polled. 2000? 3000? so divide that among 50 states. What does it come out to? between 40 to 60 votes a state? and thats what we are supposed to go by? Even if they polled 5000 people, its still miniscule to how many were polled in each state. But of course, we have to keep in mind that they are polling a few thousand out of about 130 to 140 million potential voters.
:1peleas:
The make up the sample being poll is more important than the number being poll. You can poll a million people but if they are not represented of the population, you can end up with results worst than a well designed poll of only a thousand. The secret to a good poll is having a good sample.

A presidential poll that seek to predict the winner of the electoral college can't not be single poll but rather 50 state polls each with a good representative sample of that state. This makes presidential election polling both very expensive and difficult to design.
 
This is one reason why we cant rely on National Polls. Look at how many are polled. 2000? 3000? so divide that among 50 states. What does it come out to? between 40 to 60 votes a state? and thats what we are supposed to go by? Even if they polled 5000 people, its still miniscule to how many were polled in each state. But of course, we have to keep in mind that they are polling a few thousand out of about 130 to 140 million potential voters.
:1peleas:
If we tell you that the average of national polls had Hillary up on the popular vote by 3% and she won by 2%, you would pretend that she lost the popular vote because illegals. No matter what, you will not accept being stupid.
 
i wonder what Trumps approval would be if we didnt poll anyone in Cally/NY
 
This is one reason why we cant rely on National Polls. Look at how many are polled. 2000? 3000? so divide that among 50 states. What does it come out to? between 40 to 60 votes a state? and thats what we are supposed to go by? Even if they polled 5000 people, its still miniscule to how many were polled in each state. But of course, we have to keep in mind that they are polling a few thousand out of about 130 to 140 million potential voters.
:1peleas:
They are somewhat accurate because of the mathrmatics of scientific polling. And the more polls agree, the clearer the signal.
 

Forum List

Back
Top