ABC/WP poll's inclusion of people not registered to vote makes Biden look weaker than he is

[highplainsdem]

The ABC and Washington Post poll has been including people who aren't registered to vote -- their stories note that there's still plenty of time to register -- and while they do report the results for registered voters, they're giving most of the attention to the poll numbers that also include those people who aren't registered, whether for the primary polls or head-to-head match-ups against Trump.

I didn't pay enough attention to this in their earlier polls, but did today, when I posted an OP about the different results:

https://www.democraticunderground.com/1287269685

In today's poll, there was only a 2% different, with Biden leading by 10 points instead of 8 points among registered voters.

In their July poll, Biden did 5 POINTS better among registered voters, but ABC put that info a paragraph -- and a graph -- below the numbers from all the adults surveyed, including those not registered to vote.

https://abcnews.go.com/Politics/harris-scores-debate-performance-electability-biden-front/story?id=64092090


ABC News hasn't yet released the results of the GE polling I believe they did at the same time as today's poll. In July, they released those poll findings a few days after their story about the Democratic primary results

https://abcnews.go.com/Politics/trump-reaches-career-high-approval-faces-range-reelection/story?id=64117018

I think the main takeaway from that story, for most readers, would have been the first numbers reported, in the first paragraph of the Trial Heats section, which indicated that at least 5 of our candidates could beat Trump:

Even while it’s up, Trump’s historically low approval rating makes him vulnerable in the 2020 elections – but hardly a pushover. Among all adults (there’s plenty of time to register to vote), Joe Biden leads Trump by 14 points. But that narrows among the other four Democrats tested against Trump in this poll – an 8-point lead for Kamala Harris, a slight 7 points for Elizabeth Warren, 6 for Bernie Sanders and 4 for Pete Buttigieg. The latter two don’t reach statistical significance.

You have to keep reading to find out that the story's quite different when only registered voters are polled:

Among registered voters, moreover, Biden still leads, by 10 points, but the other races all tighten to virtual or actual dead heats – Trump a non-significant -2 points against Harris, -1 against Sanders and exactly tied with Warren and Buttigieg.

You have two strikingly different stories there. Five Democratic candidates would beat Trump by at least 4 points. Or: Biden beats Trump by 10 points, while his four closest rivals are in a statistical or exact tie with Trump.

I don't know why they're highlighting the findings for all adults versus the findings for registered voters. As brooklynite said in my topic about the registered-voter results for today's poll, polling adults who aren't registered voters is meaningless -- or no more meaningful than a Facebook poll.

But the end result of their doing this makes Biden look weaker compared to his Democratic rivals, and makes those rivals seem stronger against Trump than the registered-voter findings suggest they'd be.

[BlueMississippi]

1. Media wants a horse race to drive up ratings nt

[highplainsdem]

7. Apparently.

[calguy]

2. Until we have actual primary results

All polls at this point are pure speculation. Im glad my guy Joe continues to lead the virtual pack, but still don't put much into it until poll results get backed up by vote results. Still early folks.

[5starlib]

3. Silly to do a poll for all adults when only 50% of the country votes

If we were in Europe, it would be fine to do a poll of all adults as they routinely have 95% of the population voting. But here, a poll among all adults is virtually useless.

[Skya Rhen]

4. Some of these pollsters are so manipulative. SMH...

[Thekaspervote]

5. Per usual highplainsdem... great break down ... sleuthing out the truth ...thx!

[Scurrilous]

6. K&R