GA (Landmark/WSBTV): Trump +4

[Buzz]

https://www.realclearpolitics.com/docs/2020/FINAL_WSBTV_Landmark_Poll_Georgia_PresSenate_Nov_1_2020.pdf

50-46

[Bootes Void]

this is a fairly friendly trump pollster.

[LimoLiberal]

They're either going to be wrong or right.

[Buzz]

They're either going to be wrong or right.

Bold take!

[Alben Barkley]

They're either going to be wrong or right.

Insightful!

[redjohn]

They're either going to be wrong or right.

this is going to be my quote of the night

[Lent Marslink]

Why do they keep getting overwhelmingly white demos? It's 65% white that higher than in 2014.

[LimoLiberal]

If there was a real late break to Trump, is this what it would look like? That's the question that keeps spinning through my head.

[Alben Barkley]

Other than this poll and I guess the AZ tie from Marist, mostly good polls for Biden.

Landmark has also been kind of all over the place but skewing towards Trump relative to other GA polls. So I'm not too worried about it.

If there was a real late break to Trump, is this what it would look like? That's the question that keeps spinning through my head.

Considering how many people have already voted, I'm not sure it even matters anymore. Indeed whatever "tightening" we may be seeing in some polls (not all or even most) could just come down to fewer people who have already voted are responding (mostly Democrats).

[Hammy]

If there was a real late break to Trump, is this what it would look like? That's the question that keeps spinning through my head.

Or, as others have already pointed out, many people who have already voted are no longer answering polls

[DrScholl]

They do one day samples which is very bad for polling.

[VAR]

Lief is clearly trolling.

[libertpaulian]

What was their final number for Kemp/Abrams?

[emotional hardcore]

What was their final number for Kemp/Abrams?

Kemp +2 which is basically dead on. However it was released more than a month before election day.

[Buzz]

Their final 2016 number was Trump +3

[Penn_Quaker_Girl]

They do one day samples which is very bad for polling.

One day samples?

[wbrocks67]

Yes, it totally makes sense that Trump will nearly replicate hits 2016 margin here!

[BudgieForce]

They do one day samples which is very bad for polling.

One day samples?

They do 1 poll over a single day.

[President Johnson]

It's just not there yet

[RBH]

Landmark's post-Labor day Georgia polls

9/30: Biden by 2
10/7: Trump by 2
10/21: Trump by 4
10/28: Trump by 1
11/1: Trump by 4

If they did a poll today, they'd probably get Trump by 1/2. Rollercoaster pollster.

[President Sestak]

They have Biden at 28.3% of WHITES. That's a win for sure if he actually gets that number; idk why they have such a white sample lol.

[SirWoodbury]

Lean R.

[kwabbit]

Landmark's post-Labor day Georgia polls

9/30: Biden by 2
10/7: Trump by 2
10/21: Trump by 4
10/28: Trump by 1
11/1: Trump by 4

If they did a poll today, they'd probably get Trump by 1/2. Rollercoaster pollster.

That kind of movement is to be expected with random sampling. In fact, it would be suspect if they kept on finding the same number (like Trafalgar getting Trump +2 in Michigan 7 times in a row). If the 'true' result for Landmark was Trump +1, then this would be completely normal and I wouldn't really say these polls say anything different from each other.

[forsythvoter]

GA is a really hard to poll state, because there's a huge divergence between rural (10-15% D), suburban (30-40% D), urban (60%-70% D) white voters vs. black voters (90-95% D). If you don't get the mix right, the results can easily be R+5 or D+5.

At this point, though, I doubt there's that much movement in the last 3 days. It's likely 80% or so the total vote is already in, so even a sharp last minute move won't move the needle much.

[RBH]

Landmark's post-Labor day Georgia polls

9/30: Biden by 2
10/7: Trump by 2
10/21: Trump by 4
10/28: Trump by 1
11/1: Trump by 4

If they did a poll today, they'd probably get Trump by 1/2. Rollercoaster pollster.

That kind of movement is to be expected with random sampling. In fact, it would be suspect if they kept on finding the same number (like Trafalgar getting Trump +2 in Michigan 7 times in a row). If the 'true' result for Landmark was Trump +1, then this would be completely normal and I wouldn't really say these polls say anything different from each other.

having the pollster for one of the big Atlanta TV stations apparently doing one day samples is great for eye-catching polls. As for accurate data, well.. I guess if the numbers bounce around, one of them is due to be right.