Awesome news! While we will never get to perfection (Arrow’s theorem), there must be some measure on which ranked choice voting can be quantitatively compared to single-choice vote.

For instance, if you took the sets of all candidate pools and voters with all permutations of preferences 0-1 for each candidate and prior beliefs that each candidate will be voted for by others, by how much average preference does ranked-choice voting improve the situation?

Is it ever worse for aggregate preference to use ranked choice voting?