It has two major advantages: respondents can express how much more or less they like different candidates, and respondents can rate two candidates equally (or can rate them differently).
In a municipal election, STAR has implementation downsides (lack of existing use, new ballots, voting software recertification, etc.), but those don’t really apply to your situation: you’re implementing it anew, everyone has used a 5-star rating system, and nobody has a preconceived notion of how your interface should work. Might as well optimize for the more representative result for the least respondent effort. (A simpler system called approval voting delivers almost all of the utility that STAR does for less effort, but since you’re implementing it yourself and aren’t subject to existing constraints, you may as well use STAR. They’re both much easier to tabulate than ranking.)
Hi HN! I love the concept of ranked-choice voting, so I thought it would be useful for situations that often spawn a survey. This site is pretty darn simple--no accounts, totally free for the time being.
I hope others find this useful :) And I'm dog-fooding a bit to figure out what features are worth adding [0]. Looking forward to getting feedback!
It has two major advantages: respondents can express how much more or less they like different candidates, and respondents can rate two candidates equally (or can rate them differently).
This enables it to do a better job determining the respondents’ preferences: http://electionscience.github.io/vse-sim/VSE/
In a municipal election, STAR has implementation downsides (lack of existing use, new ballots, voting software recertification, etc.), but those don’t really apply to your situation: you’re implementing it anew, everyone has used a 5-star rating system, and nobody has a preconceived notion of how your interface should work. Might as well optimize for the more representative result for the least respondent effort. (A simpler system called approval voting delivers almost all of the utility that STAR does for less effort, but since you’re implementing it yourself and aren’t subject to existing constraints, you may as well use STAR. They’re both much easier to tabulate than ranking.)