[EM] The AMS has switched from Ranked Pairs to Instant-Runoff Voting

[Kevin Venzke]

Something I meant to comment on:

Le jeudi 19 octobre 2023 à 15:00:31 UTC−5, Kristofer Munsterhjelm <km_elmet at t-online.de> a écrit :
> On 2023-10-19 21:42, Markus Schulze wrote:
> > The following explanation is now given:
> >
> >  > We decided to switch because Condorcet can give you no
> >  > result: nobody would win. Instant Runoff is preferred
> >  > by most who use a ranked system. It's much easier to
> >  > read the results with Instant Runoff.
> >
> > https://www.ams.ubc.ca/wp-content/uploads/2023/04/Council-Minutes-February-1-2023-Redacted.pdf
> 
> This seems like a weird explanation. It's true that Condorcet isn't
> always decisive, but Ranked Pairs is. And is a Sankey style transfer
> diagram that much more easy to read than a Ranked Pairs pairwise
> preference order?
> 
> But it is what it is, I guess.

What I see is that the principle of Condorcet may have legitimacy, but a method that
ensures Condorcet doesn't necessarily obtain its own legitimacy just because of it.

Yes, RP would give them a decisive result, but they wanted to sign up for Condorcet,
not RP.

As far as I know they didn't even encounter a cycle in one of their elections.

I see this as a blow to the stance that it really doesn't matter how a method
proposes to resolve cycles. Whether one's theory is that Condorcet is the only
desideratum, or whether it's that cycles would not show up in practice, this is
probably wrong from the standpoint of getting people to buy into it. People need to
perceive the merits of the approach for all cases.

Kevin
votingmethods.net