Nice! I commented in support of it. The current Caucus disaster is insane. Anyone know why caucuses exist like this? I don’t understand the arbitrary requirement of 15% or runoff.
Caucuses are actually a lot like Asset Voting except that the voters, not the candidates, get to negotiate. I agree that 15% is arbitrary, but at least that lets voters whose first pick has truly no chance have some influence.
The 15% is mandated in the DNC Rules, so that’s the part that states have no choice over. It makes sense to me, to mitigate vote splitting, though obviously 14 or 16% would work just as well.
There are so many complicated insane election rules which are all band-aids for the spoiler effect. Top 2 being the most obvious.
The posted proposal doesn’t seem that much like STAR to me.
A better proposal would be Sequential Monroe Voting with STAR for the last delegate.
I would also eliminate only those candidates with < 15% non-zero rating instead of 5 stars.
When implemented for Caucuses, it would be likely that one or more candidates would exceed more than one delegate quota of top rating votes. In that case, they would receive Floor(Total[candidate,top-rating] / quota) delegates, then ballots giving top rating to the winner would be reweighted to remove that number of delegates times the quota.
If a single delegate remains for the last step, STAR could be used instead of simple Range as in SMV.
I have issues with STAR as a single-winner method because it is not clone independent, but in this case, I think it does the right thing in a PR-style context. That being, choosing the most favored candidate within the winning faction.
The point of this post is a primary option that individual states could adopt without the DNC changing the rules. Proportional representation isn’t allowed by the DNC rules, but if we weren’t worried about that, then we could just go with single winner STAR Voting for the entire election. Ie. the STAR National Popular Vote winner would win, no need for delegates, superdelegates or even a convention.