Atheists, agnostics, long-haired, weirdos, short-haired, weirdos, vandal,

Gula.

I love the government, love the government, love the government, hug the government, love, the government.

Welcome to the politics, guys, placed for bipartisan, rational, and simple debate on American politics and policy.

I'm Michael Barronowski, Professor of Political Science at Northern Kentucky University.

Today, as the last pre-election episode of our Election 2020 series, we'll be discussing how accurate election polls are, what happened in 2016, if we can trust the polls this year, what those polls are telling us and our predictions for

the presidential race as well as control of the Senate and House of Representatives.

What have major polling organizations done to try to adjust their methods based on what they've

learned from 2016? Who wants to start us off? Faith. Yeah, kind of what happened in 2016 is not so

much that the polls were completely wrong, but a lot of the polls actually fell within their

appropriate margin of air. But the fact that Trump was actually able to pick up a couple swing

states that these polls didn't quite have in his favors, but kind of really threw it off.

And so, like, a lot of things that polling organizations have attempted to do this time around

is consider different factors that have been going on, especially considering looking at

education and seeing kind of where respondents' education is. They've also done things in terms of kind of shifting

polling, going away from traditional polling types like random digit dialing to using texting,

which I know some polls have said that has allowed them to get younger respondents and even men

to increase. So I think anything else that they've been doing has really kind of been beneficial.

Also kind of taking away from the idea that some pollsters have said that, like as Trump

refers to them as the silent majority, didn't throw them off.

But a couple of posters saying that it did in this time around, they are taking more into

account to make sure that those respondents actually give a direct

...