Friday, April 29, 2067: Polling Roundup and JULYSSA Forecasts Vol. 2

In the Democratic primaries, not much has changed in the last week.  Abdullah is still the front-runner, with a nearly 3-in-5 chance of victory:

Here are the events-based simulation results.  Note that I'm showing a different report than I showed you last time.  This report includes the odds of each candidate winning the nomination; these are simply the number of times each candidate won the nomination in 100,000 simulated primaries.

Note that the events-based simulation gives Carpenter a much higher chance of winning the nomination than the basic, polls-based model.  Meanwhile, Hancock has surged, winning 512 delegates in this week's best fit simulation, as opposed to 10 last week.  Fairchild, Hillman, Underwood, and Knotts do much better, delegate-wise, than last week as well.

What is happening here?  To be honest, I'm not entirely sure.  None of those five candidates have had an especially good week; JULYSSA could be picking up random noise.

Meanwhile--there's been a poll including Harper Henry!!!  In fact, only one (reputable) poll has included Henry in the week since she launched her candidacy.  Oddly, two polls included possible candidates, but not Vickers' actual challenger!

To say Henry has long odds is a bit of an understatement!  Indeed, in the best fit model for the convention, she receives no delegates at all.  As mentioned in the post about her campaign launch, she is very similar to Vickers politically except in an area--polyamory--where her views don't match the party as a whole.  JULYSSA knows this, and thus doesn't expect her to catch on.

The output for the general election simulations, I've realized, is a bit hard to read, so I need to clean it up before I post it.  Stay tuned for those results!

Tuesday, April 19, 2067: And Then There Were Two

It's now official: Congresswoman Harper Henry is running against Vice President Vickers for the Republican nomination.

It seemed very unlikely that Vickers would face a primary challenge; the 35th Amendment was largely supported by Republicans because they hoped he would run for president someday.  But it's been a long time since 2058, and Vickers is now so far right, especially on robot and AI issues, that the party's liberal and moderate wings have been openly clamoring for someone to challenge him.

Harper Henry would be an odd choice for them; her ADW-NOMINATE score is actually further to the right than Vickers was when he was last in Congress.  Indeed, they agree on almost everything: supporting the Gomez-Lausen Act and denying robots civil rights, expanded global warming mitigation efforts, and foreign intervention and an expanded role of the US in the Democratic Alliance.  There is, however, one issue where they differ significantly: polygamy.

Vickers' support for a constitutional amendment to ban group marriage is a hallmark of his campaign.  Henry, on the other hand, is polyamorous herself: she has lived with her two boyfriends and their children for almost twenty years.

It's hard to know what damage Henry can do to Vickers based on this position.  As I have mentioned in my paper on JULYSSA, the model includes Polling USA's monthly American Values Survey, and in last month's survey, Republican voters were opposed to polygamy 61% to 29%, with 10% undecided.

Then again, I have also mentioned that JULYSSA does not use a Rational Actors Model, because we (Alice and I) have shown that voters are not rational actors, and their preferences can be, frankly, weird.  In other words, voters may choose Henry if they feel Vickers has an electability problem, even if she doesn't match their values on marriage and is otherwise politically similar to him.

Right now, JULYSSA has Henry's chance of winning the primary at 12.5%, but this estimate must be regarded with extreme caution: she has not, as of yet, been included in any polls, and JULYSSA assigns it's greatest model weights to polling results.  More time will be needed to see what Henry's chances really are.