[uoaei]

> doesn't allow you to express your preferences very well, since there is very limited differentiation between candidates

Again, so what? The function being optimized here is not "fealty to true preferences for individuals", but rather "level of satisfaction with outcome for the entire group of electors". And actually, my preference is to not be asked to rank candidates but just issue binary assessments. Your claim that people want the former is wild, unsupported hypothesis, and the way you phrased it seeks to imply they'd rather not vote than vote approval voting. I think you are strongly (albeit transparently) forcing your own preferences through others' mouths in an attempt to start this more or less malformed debate.

> approval goes head to head with IRV or STAR (the two other movements that seem to have some momentum) I would generally expect the voters to go with STAR or IRV, for better or worse.

IRV is literally the same thing as RCV and so displays the same idiosyncracies and fundamental flaws so I'm not quite sure you know what you're talking about here.

Further, the kind of person attracted to STAR is the same kind of person who can devise linear algebra algorithms for deciding what's for lunch, and go in to change every setting on their new smartphone before the wallpaper. They are micromanipulating technocrats, and while this community massively overly represents that type, it doesn't represent the public per se. I assure you, what you parrot from online discussions by people who very much don't have their finger on the pulse, is then predictably very far from representing views considered "normal" in the literal sense.