Quote:
Originally Posted by Firebot
My issue with polls, is that the only ones who respond to them are ones which would hit a mindset, and they try to portray a narrative. They are biased in nature of how they are conducted. Since most polls are generally media controlled and most media is left wing, the results tend to be heavily slanted.
A reminder of how 2016 was looking around this time.
https://projects.fivethirtyeight.com...tion-forecast/
Oct 17 2016 had Clinton winning at 88.1%
People are really betting way too much on the polls and I feel like Trump could win and be a shock to everyone...again...to anyone who doesn't realize that large demographic of the population isn't on reddit or on forums and online posters don't necessarily take the time to vote.
Trump wasn't supposed to win in 2016, yet he did. |
That poll projects Clinton to have 48.5% of the vote. She received 48.2% of the vote.
The poll had Trump at 44.9% of the vote. He ended up with 46.1%.
It was within 0.3% for Clinton and 1.2% for Trump.
The poll was bang on accurate.
Looking at the state level. Pennsylvania for example, 538 had Clinton polled at 48.9%, Trump at 45.2%. Clinton actually received 47.5% of the vote (1.4% difference) and Trump 48.2% (a variance of 3%).
I don't see a margin of error listed, but a quick search shows CNN's latest poll has a margin of error of 4%.
It wasn't the polls that were wrong, it was the analysis of the numbers.