GA-Landmark: Trump +4

[Hammy]

Landmark was the most accurate in 2016/18. Georgia flipping was always a pipe dream.

[Hammy]

Landmark was the most accurate in 2016/18. Georgia flipping was always a pipe dream.

You flip from doomer to non-doomer and back again so fast and so often it's giving me whiplash.

Not dooming--I've never expected Biden to win GA/NC/TX and there are statistical reasons backing that, and polling discussion is sport for me.

[Hammy]

Landmark was the most accurate in 2016/18. Georgia flipping was always a pipe dream.

You flip from doomer to non-doomer and back again so fast and so often it's giving me whiplash.

Not dooming--I've never expected Biden to win GA/NC/TX.

It's dooming to say it's a "pipe dream" for Biden to win two states where he has a lead in the averages. Or even one where he is only very slightly down.

Well yes, when you have more bad quality polls than good quality ones, it's going to throw off the polling average. And as I've pointed out in other threads, NC for example seem to over-poll Dems in the same way Arizona does with Republicans.

[Hammy]

The senate numbers in this poll are really off too. No other poll has had Perdue at 50% and Collins in second. Also Lieberman surging back from the brink with over 8%.

Too R-heavy sampling perhaps?

[Hammy]

And this is Landmark's third poll this month, which is suspect considering that they don't generally poll that much. What are they up to?

From the polling history on Atlas it seems they poll more often during presidential years--in 2016 they had three between Sep 22 and Oct 20.

[Hammy]

Landmark doesn’t weigh for education.

https://twitter.com/justingraywsb/status/1319747508797607938?s=21

If Biden is winning back WWC/non-college voters, would that cause a reverse 2016 type of error that ends up being to Biden's benefit?