3 ms·
IRV is only awful in irrelevant corner cases that only political science wonks care about. When the electorate's preferences is a statistical tie, it doesn't m
by gowld 8y ago
IRV is only awful in irrelevant corner cases that only political science wonks care about.
When the electorate's preferences is a statistical tie, it doesn't matter who wins, as long as one party isn't forcing all the tiebreaks to go their way.
- function_seven 8y ago> When the electorate's preferences is a statistical tie, it doesn't matter who wins I wish this was more front and center in these types of discussions. Imagine a race between Alice and Bob. Alice gets 9,500 votes, Bob 9,521 votes. But—due to some controversial rules—10 of Alice's and 40 of Bob's votes are invalidated. Alice wins. This is not a scandal or a problem. The original vote totals were a tie. The new vote totals are also a tie. Essentially a coin flip determined the winner. For purposes of determining who won, 49%===51%. If a candidate is truly wanted by the electorate, then they need to do better than 50% +1 anyway.
- wnoise 8y agoIt's awful precisely when it would start delivering a different result than FPTP, because that's when strategic voting comes into play.
- harshreality 8y agoI don't think that's true at all. It provides the illusion of compatibility with 3rd party voting, while saddling it with strategic voting pathologies not so different from winner-take-all. Here's a summary: https://rangevoting.org/IrvPathologySurvey.html https://rangevoting.org/IrvPathologySurvey.html
- aaron_m04 8y agoIt seems to me that the weirdness comes from this if/else block in how IRV works. if (highest voted candidate has more than 50%) { That candidate is winner } else { use 2nd picks of voters for least voted candidate IRV() // recursive } A really desirable algorithm wouldn't be a piecewise function like this.