>>915
In "The Alternative Vote Explained" (https://www.youtube.com/watch?v=3Y3jE3B8HsE), CGP Grey claims the Alternative Vote (aka Instant Runoff) doesn't have the spoiler effect. Strictly speaking, this isn't quite true. Whenever no Condorcet winner exists, you have cyclic majorities, and when you have that, you can always construct several alternative scenarios where all but two candidates drop out, with no candidate being preferred by a majority in all scenarios they appear in. So, all methods that elect the majority winner in an election with only two candidates with 100 % probability suffer from the spoiler issue, though to different extent, and I agree that Instant Runoff is better than Plurality in that regard.
The scenario where Runoff works better than Plurality is with a two-candidate election turning into an election with three candidates, the third-party candidate being somewhat similar to one of the two already in the race, getting much fewer votes than the two established candidates, but, by drawing support away from one of the candidates, enough to change the winner to the one least liked by the supporters of the two similar candidates. To be clear, we are assuming here that candidate A has more votes than either candidate B or C, but less than both of them combined, and B and C as well as their supporters consider A the worst. A wins in Plurality, but is prevented from winning by Instant Runoff (or a real second voting round). And A is also prevented from winning by the methods Top Flop and Co-op presented in the pdf, and both these methods are easier to count.