Creme de la Creme Polling Averages! Only the best polls for the DU readers!

[bluewater]

Crème de la Crème Polling Averages! Random Live caller, A rated polls within the last 30 days only!

Every wonder what the polling averages would look like if they only used random live caller A rated polls?

Wonder no more... I present my homespun Crème de la Crème Polling Averages!

No online polls.

No robo-caller polls.

No unrated or C rated or even B rated polls!

All polls have to close within the last 30 days to be included, so here are the results:

Quinnipiac University: A- rated = Warren 28%, Biden 21% : +/- 4 margin of error (Oct 17-21)

CNN/SSRS: A- rated = Warren 19%, Biden 34% : +/- 5 margin of error (Oct 17-20)

Fox News/Anderson: A rated = Warren 22%, Biden 32% : +/- 5 margin of error (Oct 6-8)

IBD/TIPP: A- rated = Warren 27% Biden 26% : +/- 5 margin of error (Sept 26-Oct 3)

Monmouth University: A+ rated = Warren 28%, Biden 25% : +/- 5 margin of error (Sept 23-29)

Crème de la Crème Polling Averages : Warren 24.8% to Biden 27.6%

Lol THAT was amazing fun!

There are two big take-aways from this little exercise:

1. Warren and Biden are in a statistical TIE in A-rated truly random Live Caller polls, Biden with a +2.8% lead while the moe's are in the +/-4% to +/-5% range.

2. These A-rated live caller polls do not occur very frequently and there are so few of them. The polling landscape is overwhelmingly dominated these days by online polls, which I don't think most people realize.

It's amazing how this A-rated live caller poll average is so similar to the results on The Economist's Primary Polls tracker:

https://projects.economist.com/democratic-primaries-2020/

They use some online polls while excluding others, but their model also has the race in a virtual tie with Biden leading +1%.

All in all this was a fun and thought provoking little exercise.

[HeartlandProgressive]

1. Thanks for posting this!

I was surprised there are so few live caller only polls. Were there any B-rated live caller polls?

Looking at 538's list, I can't find any. It seems if a poll is live caller only 538 rates it at least A-.

[bluewater]

3. ywyw

It's amazing how this A-rated live caller poll average is so similar to the results on The Economist's Primary Polls tracker:

https://projects.economist.com/democratic-primaries-2020/

They use some online polls while excluding others, but their model also has the race in a virtual tie.

All in all this was a fun and thought provoking little exercise.

[boomer_wv]

2. Cherry picking.

Why are you leaving out the CNN poll?

That one had Biden up 15 last week.

[bluewater]

4. Opps.Thanks for pointing out I jumbled 2 polls. The OP is corrected now and is pretty much the same.

Thanks.

There is no Fox News/SSRS poll as I originally posted... there is the Fox News/Anderson poll and the CNN/SSRS poll that you pointed out was missing. I went back and corrected the OP for this jumbling of those two polls.

Sorry for that mistake, and thank you for pointing it out. It change the average result a few points, but not outside the margin of error.

Sooo, we See Biden with a +2.8% lead in the A-rated live caller only polls in the last month. And all those polls had marins of error of either +-4 or +/-5 percent.

[CTyankee]

11. I don't see the size of any of these polls. Since the moe goes up the smaller the poll size and

vice versa with a large poll size, this makes a difference in the way we look at poll veracity.

[bluewater]

12. The margin of error for each poll is included. Did you overlook those perhaps?

[CTyankee]

13. mi scusi! It's right there, aagh

[bluewater]

14. np. My eyesight is so bad I misread things all the time myself.

based on the moe's, all the polls appear to be roughly the same size and smaller than the online polls that survey larger preselected panels.

I believe that's because it's more expensive and time consuming to actually do live caller polling.

But as they old saying goes... you get what you pay for.

[CTyankee]

16. Well, I was at fault for not reading more closely.

Anyway, interesting polls.

But ya know, it's only the end of October and I'm exhausted already. First I was for Kamala because we needed a woman and a woman of color to boot. But the debates and lots of online reading has changed me to EW (I think Booker would be a great VP...a real plus for community organizing and who would probably accept because he would be looking to his own run in 8 years).

I still like to hear from lots of the talking heads, particularly the ones from the NYT, Wash. Post, and some members of Congress. I like to hear their perspectives...

[bluewater]

17. Thanks for the discussion!

[Thekaspervote]

5. Good question isn't it!!

[HeartlandProgressive]

6. Asked and Answered though.

lol Do keep up.

[DownriverDem]

8. It's still clear

that Biden is the Dems best choice to beat trump.

[bluewater]

9. Has Biden ever won a single Presidential primary in a single state?

I don't think he has.

I believe the only national primaries he ever won were on Obama's coat tails running for VP.



Joe will probably win some this year though, since the 2020 Primary race is in a virtual tie between Warren and Biden.

But who will win the nomination?

Time will tell.

[bluewater]

7. if you average the RCP Polls tracker and the Economist's tracker you get...

pretty much the results of my little homespun crème de la crème polling average!

RCP Polls: Biden 27.2%, Warren 21.8%
https://www.realclearpolitics.com/epolls/2020/president/us/2020_democratic_presidential_nomination-6730.html

The Economist: Biden 25%, Warren 24%
https://projects.economist.com/democratic-primaries-2020/

Average of RCP Polls and The Economist = Biden 26.1%, Warren 22.9%

Crème de la Crème = Biden 27.6%, Warren 24.8%

lol so my little experiment gives results very similar to the two major polling aggregate trackers!

Go bluewater! lol

[RudyColludie]

10. Yeah but the upper bound for Joe's confidence interval is still higher!

Oh yeah, baby! There might be a high degree of overlap but it's still Biden Time!



Vote Joe or Trump won't Go!

[bluewater]

15. As I always say...

the upper bound of the confidence intervals are what the upper bound of the confidence intervals are

[BlueMississippi]

18. When you average polls, the MOE is not averaged

The MOE is recalculated for the combined sample size and so it goes lower.

(That is the point of averaging - to reduce MOE!)

[UncleNoel]

19. "Creme de la Creme" October Poll Averages: Biden 29.75, Warren 22.75, Sanders 16.25 (Biden +7)

I wanna play, too!

Let's eliminate September polls. It is the end of the month now.

Biden 29.75, Warren 22.75, Sanders 16.25 (Biden +7)

Quinnipiac A- (LV) 10/17-21: Biden 21(-6), Warren 28(-2), Sanders 15(+2) [Warren +7]
CNN/SSRS A- (RV) 10/17-20: Biden 34(+10), Warren 19(+1), Sanders 16(+1) [Biden +15]
SUSA A (LV) 10/15-16: Biden 32(-1), Warren 22(+3), Sanders 17(0) [Biden +10]
FOX News A (LV) 10/6-8: Biden 32(+3), Warren 22(+6), Sanders 17(-1) [Biden +10]

Or, post October debate:


Biden 29, Warren 23, Sanders 16 (Biden +6)

Quinnipiac A- (LV) 10/17-21: Biden 21(-6), Warren 28(-2), Sanders 15(+2) [Warren +7]
CNN/SSRS A- (RV) 10/17-20: Biden 34(+10), Warren 19(+1), Sanders 16(+1) [Biden +15]
SUSA A (LV) 10/15-16: Biden 32(-1), Warren 22(+3), Sanders 17(0) [Biden +10]

Either way it is Biden by 6 or 7. Not quite a tie!

[bluewater]

20. Nope, have to keep it 30 days. That's the rule lol It's my tracker

no cherry picking dates! More polls will be coming out before the month ends.

And by the way, Creme de la Creme Poll Averages of A Rated live caller only Polls is TRADEMARKED. lol

Go back to your own homespun polling average with those beastly "B-rated" polls.

man... talk about intellectual property theft lol

lol Joking. Mostly. But no changing the rules, it's my poll tracker, 30 days is the expiration date

I am glad you like my incredibly clever idea that has so much more panache than your idea.

Again, Joking. Mostly

Seriously, glad it peaked your interest, but no changing the rules or stealing the Creme de la Creme brand name.

Thinking out loud:
My Crème de la Creme Tracker has one unique issue to decide how to handle. These A-rated live caller polls come out so infrequently, I need to establish my rule for when the tracker gets updated. I think it would be silly, for example to post results for only 2 or 3 polls within a sliding 30 day window.

For now, I think the Crème de la Crème Poll should be updated when there are a minimum of 5 polls in the current 30 day window. If one drops off, the Crème de la Crème Polling Average will not be updated until at least one additional poll can be included.

And as I said, my brand-named personal homespun tracker, my rules. lol

Have a nice evening.

[LanternWaste]

23. Seems the most effective way to get the data we want to better validate the conclusion we want

"my brand-named personal homespun tracker, my rules. lol"

[bluewater]

24. lol my tracker my rules, and no stealing the brand name!

It's "methodology" is A-rated live caller polls only that have been released in the last 30 days. Minimum of 5 polls needed to up date the tracker.

oooo that's soooo manipulative!

Remember to look for The OFFICIAL ORIGINAL Crème de la Crème A-rated live caller only Poll tracker!

Accept NO substitutes!

An Interview with bluewater, founder and administrator of the Crème de la Crème A-rated Live Caller Poll tracker:

Question: Is it true you started the Crème de la Crème poll tracker as a gentle jibe at other attempts of homespun poll averaging that people are doing here on DU?

bluewater: uhhhhhh...

[Celerity]

21. good post nt

[brooklynite]

22. The thing about The Economict polling model...

...is that it shows Biden static and Warren rising...