Monday, October 10, 2016

On Sunday posts, electoral math, voter psychology, marketability, and the care and feeding of black swans.

[This is try number two. The first was lost when my iPhone decided that a 40% charge on the battery was dangerously low and decided to shut down just as I was about to hit send on the email of my dictation. This is yet another issue I wish Apple had prioritized over coming up with new headphone jacks.]

I recently posted a couple of items on undervoting and its implications. These were done in haste and there were a few points I did not have a chance to get to.

Sunday posting

Regular readers know that we generally take the weekends off here at West Coast Stat Views. Recently, though, events have been moving so rapidly that a delay of 48 or even 24 hours can mean the difference between looking prescient  and being behind the curve.

Electoral math

I may not have been clear about this the previous posts. In terms of electoral outcomes, there is virtually no difference between the undervoting I described and voting for a fringe candidate ('fringe' implying that here she has no chance of winning).

Voter psychology

Here is where we start seeing a difference. The meaning of fringe party votes is inevitably muddy. Are you voting for them because you genuinely think they are the best candidates or simply because they do not have Democratic or Republican after their names? Leaving the race blank on the ballot sends a clean message.


In terms of going viral on a national level, one of the problems with fringe-vote as protest is that the voters' options vary from state to state and race to race.  #LeaveItBlank is simply more manageable than...


Care and feeding

I apologize, but the discussion is going to get a bit meta now. I may get some pushback on this, but I believe that most data journalists and quite a few political scientists covering this race have made one of the most fundamental errors in analytic thinking. They have all acknowledged that these are abnormal times but they have locked themselves into normal mode. Think about all of the articles we've seen over the past year that start by admitting that long-held assumptions have been violated and that well-established models are proving unstable, then go on to use those same assumptions and models to make confident predictions.

I'm not making predictions here. I am simply throwing out what I think are plausible but generally not likely scenarios. I have no intention of applying even a rough subjective probability at this point. I've been watching this game for a while and I've seen way too many people fail to find the queen.

What I am saying is that we need to think about these problems using tools and approaches that are appropriate for abnormal times.

No comments:

Post a Comment