23 ms·
It's not possible for a candidate with zero first preference votes to win [1]. If they have zero first preference votes, they will be eliminated in the first ro
by harpastum 10y ago
It's not possible for a candidate with zero first preference votes to win [1]. If they have zero first preference votes, they will be eliminated in the first round.
Ranked choice doesn't always get the correct result, but it's a lot closer than a simple vote.
"If no candidate receives a majority, the candidate with the fewest first choices is eliminated and voters who liked that candidate the best have their ballots instantly counted for their second choice."
- http://www.rcvmaine.com/how_does_ranked_choice_voting_work http://www.rcvmaine.com/how_does_ranked_choice_voting_work
- baddox 10y agoHow are you defining "correct result?" One formal definition could be the Condorcet winner (the candidate that would win a one-on-one election against each other candidate), but some elections don't even have a Condorcet winner. Also note that neither plurality voting nor instant runoff voting are guaranteed to select the Condorcet winner if one exists. https://en.m.wikipedia.org/wiki/Condorcet_criterion https://en.m.wikipedia.org/wiki/Condorcet_criterion
- skrap 10y agoI'm of the uninformed opinion that any national election will probably have a Condorcet winner. IF that's true, then we could literally pick any Condorcet method, and just go with that. Condorcet IRV seems a reasonable choice, in that it's relatively easy to explain to people (compare to beatpath, for example). I think all the discussion about the merits of different methods is just so much bikeshedding. It might matter, but it's far more important to pick _something_ than to stick with plurality! In that spirit, I am totally in support of Maine's choice to go with IRV, even with its warts.