August 10 Poll Report

Good afternoon everybody.   I have been on vacation visiting my sons on the West Coast, and, while I tried to update the polls every day, there were quite a few days in there when I wasn’t able to get it done.  I am back at home now, so I should be able to do regular updates from now on.

It has been four days since I have done an update, so let me begin by reporting on the four polls which have wrun since I have ritten.  In the poll of August 6^th, Elizabeth Warren drew 56% against an expected 50%, with the gains coming entirely at the expense of Kamala Harris:

In the standings posted below, Warren will be up 33 points and Harris down 36 points because of this poll.   The candidates listed in that poll control 37% of the vote, which is more than the other three polls combined, and that poll was 88% consistent with previous poll results. 

In the poll of August 7^th, Michael Bennet staggered a weak polling group, taking 44% of the vote:

Bennet was up seven points from expectations, while all three of the others were down.  Those four candidates control only 6½% of the vote, and predictions for the poll were 85% accurate.   Michael Bennet will be up 11 points in the Support Scores (below) due to that poll, and also the removal of an old poll. 

Mike Gravel was scheduled to be included in the poll of August 8^th, but he pulled out of the race a few days before that, so we wound up polling only three candidates, which I don’t like to do although there is no clear reason not to.   (If you go from four candidates to three the norm goes from 25% to 33%, but that has nothing to do with the math, since those numbers don’t play into the computations.  What goes into the computations is how one candidate fares vs. the other candidates in the group.  A win of 50% to 25% is exactly the same, in the computations, as a win of 40% to 20%.)  Anyway, Gravel hit the road, which left us with three candidates, one of whom was much stronger than the other two:

The system is set up to generate matches between candidates of relatively equal strength, but with a random input.  Occasionally the random numbers work out such that you get a strong candidate matched against non-competitive candidates.   Those three candidates represent 12% of the Support, and predictions for the poll were 97% accurate. 

And in yesterday’s poll, the poll of August 9^th, we had Donald Trump polling his usual 25% and splitting the rest of the vote.  But there is a surprise there, which can not be explained with reference to Trump.   Gillibrand won the poll with 29%.  Accepting that most of my twitter respondents would vote against Trump in a polling group with O.J. Simpson, Seabiscuit and Frosty the Snowman, we still would have expected Gillibrand to finish behind her two Democratic rivals, not ahead of them.   Don’t know what happened there; anyway, there are the poll results:

Those candidates control 13% of the support, and predictions for that poll were only 58% accurate, which is very, very low. 

The changes in the scores since our last update are actually very modest, despite four polls being added to the system and four removed.  Since our last update:

Elizabeth Warren is up 33 points,

Kirsten Gillibrand is up 22 points,

Tulsi Gabbard is up 17 points, since the removal of an old poll outweighs her disappointing performance in yesterday’s poll,

Jay Inslee is up 13 points,

Michael Bennet is up 11 points,

Pete Buttigieg is down 10 points,

Kamala Harris is down 36 points,

Beto O’Rourke is down another 34 points due to the removal of an old poll from the data considered relevant, and

Donald Trump is down 40 points. 

These are the updated Support Scores:

The Green List, which at one point had 12 names on it, is now down to four names.  Andrew Yang has temporarily dropped off the Green List, but will probably be back on it tomorrow; he is up 32% in the last 30 days.  The requirement is that a candidate has to be up 33% in the last 30 days to qualify for the Green List, but he is up 32% and is performing extremely well in the poll which is currently running.

I have two changes to the process to announce.  One is that I will start now removing TWO days of old polling each day, until the number of polls included in the system is down to 40, rather than 50.   This has to do with a few candidates dropping out of the race or being eliminated from the polling.  I need at least 15 "positioning estimates" or position points for each candidate to feel comfortable. . . well, I really need 10 points to feel pretty comfortable, but it is always better to have more data than you absolutely need.  At one point I was polling 30 or 32 candidates or something; at one point I was polling Wayne Messam and Eric Swalwell and Jeff Flake and Stacey Abrams and John Kasich.   You get 12 positioning points from each poll; with 40 polls, that’s 480 positioning point estimates to work with.  With 30 candidates, that’s just 16 per candidate—not enough to feel comfortable.  But as more candidates drop out of the race, I am more comfortable relying on fewer polls. 

The other change is that I am not going to try to write DAILY updates; I’ll do them at least twice a week and probably every two days.  I don’t mind the work, but it’s kind of the "watched pot never boils" syndrome.  When you watch the results too closely, nothing seems to be happening.  I think the reports will be more interesting if I don’t do them as often. 

In the reader comments to the last poll there was discussion about Climate Change/Global Warming, and, while this is not a Climate Change discussion page, I thought I would comment briefly about that in an effort to move the discussion.   There are what I call Spin-Your-Wheels discussions, and there are discussions which make progress. 

In the larger community, Climate Change has obviously devolved into a Spin-Your-Whells discussion in which people keep making the same arguments on both sides over and over, without many people moving their positions, sort of like a chess game in which both players keep moving their knights and bishops back and forth so that the game merely returns to where it was.  In our little community, though, could we perhaps start moving the discussion along?

I don’t see, personally, that there is any reasonable doubt that the world population is facing a tremendous problem which will have serious consequences.  I don’t think there is any doubt that the sooner we can find consensus on what needs to be done, the better off our grandchildren will be. 

Part of the problem, though, is that while the Climate Change deniers are (I think) clearly wrong, not everything which is said by the Climate Action Advocates is true.   SOME of it is true, much of it is true, but a great many things are said about Climate Change which are just complete nonsense, and many other things are said which may be true or may not.  From my standpoint, it is almost impossible to tell whether new claims are credible or are not—and I am sincerely trying. 

When people say things about Climate Change which are not true or are absurdly difficult to substantiate, this gives ammunition to those who want to deny that Climate Change is happening at all.  From my standpoint, what would appear to be helpful to moving the discussion forward would be if people who are genuinely knowledgeable about the subject would organize and bat down those who make widely publicized irresponsible or undocumented claims about Changes to the Environment.   That would be my thought.

Thank you for reading.