Between last week and now, things didn’t change much: only one poll aged out (Fox News) and only one new poll (Suffolk University) was added to the mix. Other than that, just updates from the weekly polls. Here’s the numbers:
Late with this week’s edition - again, busy times on the home front. Hopefully things should start settling down now.
Also, 538 is messing with my head. You may remember that I’ve been giving their A-rated polls a weight of 5; B-rated, 3; and C-rated, 1. Only now they’ve put a number of pollsters on the A/B and B/C borderlines, so I had to revise my spreadsheet to allow for weights of 4 and 2, respectively. Fuckers.
Anyhow, besides that, we’ve got almost an entirely new crop of polls. A-rated Monmouth, Fox News, and Washington Post/ABC; A/B rated NBC/WSJ and IBD/TIPP; B-rated Ipsos, Suffolk, and the weekly YouGov; B/C rated USC/LA Times and the weekly Morning Consult; and C-rated daily HarrisX. (Whew!)
But despite all that, the numbers didn’t change much (but you’ll have to scroll right to see this week’s numbers):
According to Betfair, here in order are the 14 persons most likely to be elected President a year from now. The 2nd number is the chance of winning the Democratic nomination.
40.5% - Donald Trump
14.9% 29% Elizabeth Warren
11.5% 22% Joe Biden
7.6% 16% Pete Buttigieg
6.8% 11% Bernie Sanders
2.8% 3.6% Michael Bloomberg
2.7% 5.8% Hillary Clinton
1.9% 3.6% Andrew Yang
1.8% - Mike Pence
1.2% - Nikki Haley
1.0% 1.7% Amy Klobuchar
0.8% 1.7% Tulsi Gabbard
0.6% 0.9% Kamala Harris
0.5% - Mitt Romney
Warren was 52% to win the nomination on October 15, so has fallen sharply over the past four weeks.
Chances of Democratic control according to Predictwise:
House of Reps 73%
Presidency 53%
Senate 36%
There’s not a lot of point in even putting up this week’s numbers, so little has changed.
Since last week, here’s all that’s happened: the Suffolk poll has aged out, and YouGov and Morning Consult have done their weekly updates. Even HarrisX has taken a break from updating.
That’s it. Not a single qualifying new poll. So the numbers have changed only minimally. And here they are:
Just in case I’ve never said it all in one place, here’s how I do my averaging:
Each Wednesday (usually), I go to 538’s list of recent national polls.
I include those polls whose middle date in the field is within the past 2.5 weeks (17.5 days).
For instance, a poll that was in the field November 1-3 would have middle date November 2, and would sunset in between 11/19 and 11/20. A poll that was in the field November 1-4 would have middle date November 2.5 and would still be good on 11/20, then sunset before 11/21.
Polls with a 538 rating in the A range are given a weight of 5.
Polls with a 538 rating of A/B are given a weight of 4.
Polls with a 538 rating in the B range are given a weight of 3.
Polls with a 538 rating of B/C are given a weight of 2.
Polls with a 538 rating in the C range are given a weight of 1.
Then I average them with those weights.
Unrated polls aren’t included. Only national polls are included, no polls of individual states or groups of states are included.
Can you do one of just the swing states? National polls don’t really give you much of an idea of who is going to win in a close election.
Hillary won California by 4.5 million votes. Her entire popular vote margin was only 3 million.
The only polls that matter are the ones in swing states and I wish our lead was bigger in some of those swing states. I’d like to see a decisive victory but at this point any victory would be nice.
Weighted average of polls (from fivethirtyeight.com) from past seven weeks. First figure is Pennsylvania, 2nd is Wisconsin.
Biden 27% 22%
Warren 16% 25%
Sanders 12% 18%
Buttigieg 5% 6%
But is this really the stat you want? These poll Democratic primary voters. We need to know how Archie Bunker will vote in the general election.
The reason I wouldn’t even if I had the time is that primary polls don’t map well onto general election polls. One can draw some inferences, but they’re arguable.
But if you’re interested, just go to 538’s primary polls and where that drop-down menu at the top says ‘National,’ find the states you’re interested in on that drop-down, and you’ll get the most recent primary polls for that state. There really aren’t that many polls for any given state, so you can get a sense quite quickly of roughly where things stand.
Do we though? The drop-off between AA turnout in 2012 and 2016 was about 6 percentage points, from about 65% to about 59% IIRC. If it had been more like 60% or 61%, Hillary would have gotten enough additional AA votes in Milwaukee, Detroit, and Philadelphia that right now we’d be bemoaning Hillary’s inability to do anything as President, and worrying about her re-election prospects.
(This despite the fact that her AA support in the primaries was way more than where Biden’s polling with black voters.)
After almost four months leading the pack, and getting as high as 54%, Liz Warren has fallen back to below Biden in the prediction markets for the D nomination. Latest numbers from Predictwise.Com:
23% Joe Biden
21% Elizabeth Warren
18% Pete Buttigieg
12% Bernie Sanders
4% Michael Bloomberg
4% Andrew Yang
2% Amy Klobuchar
Why the big fall for Warren? Her kicking an own-goal by wanting to take away the private insurance many Americans are happy with?
These numbers sum to much less than 100%. The markets seem to give high probability to a dark-horse coming from out of nowhere to save the day!
I admit to not understanding the numbers for the R nomination:
78% Donald Trump
7% Nikki Haley
7% Mike Pence
3% Mitt Romney
A stupid and hateful criminal who behaves as though he really is Putin’s puppet? Sure, let’s re-elect him!
I skipped last week, and should have said something about it. Basically, all the old polls had aged out, and only two new polls (other than our three that are there every week) had replaced them: B-rated Ipsos, and Emerson, recently promoted from a B to an A. Just seemed like too few polls to average, so I let myself be lazy.
In addition to those, we’ll have a new Quinnipiac poll (newly demoted from A to B) in this week’s average, and we’ll see what else shows up tomorrow.
That, and a ton of generally critical press.
I’m a Warren fanboy, but if I had to put down a bet right now, it would be on Biden.
He’s the weakest anointed frontrunner ever, but he never seems to fall out of that 27-28% range very far or for very long. If there’s a candidate who can break the 40-year-old pattern and win the nomination after losing IA and NH, it’s him.
The only way he’s losing the nomination is if he’s impeached and removed, and that’s a long shot. What’s to understand? GOP voters don’t care if he’s Putin’s puppet. America being the world power is soooo 2015, as far as Republicans are concerned. Can’t own the libs like that!
Short version: Warren’s plummeted, Sanders has gone up a tad, and Buttigieg’s firmly into double digits. And Bloomberg’s above this average’s 2% threshold. The numbers:
Unless a new poll is released tomorrow, there will be no new polls from last week, just the updates of the three weekly polls. But since this will be the last week with Harris in the polls, I’ll go ahead and put up the new numbers once YouGov is updated tomorrow, even if they’re only barely changed from last week.
We did have one additional poll: C-rated Change Research, which had Biden, Warren, and Sanders in basically a 3-way tie. Now we know why they’ve got a C rating at 538, and also why I downweight C-rated pollsters. Anyhow, Kamala’s last hurrah:
aSince last week, all of last week’s polls have aged out or (in the case of our Weekly Three) been replaced. We have new polls by A-rated Monmouth and B-rated Quinnipiac and Ipsos, and the weekly polls from B-rated YouGov, B/C rated Morning Consult, and C-rated HarrisX. And here we go:
The changes aren’t earthshaking. Buttigieg’s dropped out of double digits, and Bloomberg’s picked up a couple of points, but let’s see if those changes hold up after another week. Similarly, Biden’s at his lowest point so far since I started keeping this average, but not by that much. And even if that change does hold up after another week, he’s still eight points ahead of Sanders and ten ahead of Warren: can’t beat something with nothing.
The Betfair numbers reported by Predictwise are also moving:
22% Biden
18% Sanders
14% Warren
14% Buttigieg
8% Bloomberg
4% Yang
3% Klobuchar
1% Gabbard
0% Booker
Sanders is at 2nd, up from 4th last time I posted. The infarction helped his campaign? :dubious:
