I had a look at FairVote’s STAR write-up, and the rather contradictory logic (concern about center-squeeze while claiming IRV condorcet failure was rare, concern about center-squeeze while insisting on later-no-harm) annoyed me, so I wrote a rebuttal.
I know Equal Vote already published this piece a while ago, part of the idea here was that since FairVote’s arguments sound reasonable if you are unfamiliar with the subject, the counterarguments need to make it clear why the original arguments don’t hold water to someone unfamiliar with the subject. I think the Equal Vote piece relies on a fair amount of assertion of concepts that aren’t obvious to people who haven’t thought much about any of this (such as the claim that tactical voting in STAR generally backfires). Hopefully some of the explanations here are useful as ways to frame the issue.
That said, I won’t claim that this a good argument to show people on its own. Some of the snark directed at FairVote might seem uncalled for. It also spends a lot of time criticizing IRV rather than defending STAR. The reason for this is that a lot of the original piece touts various features of IRV that STAR lacks, and I was trying to explain why they aren’t desirable. The other thing is that it harps more IRV criteria failures than STAR passes - most notably participation failure. I still think it’s legitimate, as this kind of pathology seems both less likely and smaller magnitude (i.e. the ‘best winner’ and the ‘actual winner’ will be closer to each other) in STAR than IRV, but I didn’t mention that in the rebuttal.