[EM] Manipulability stats for (some) poll methods

[Filip Ejlak]

wt., 14 maj 2024, 17:59 użytkownik Kristofer Munsterhjelm <
km_elmet at t-online.de> napisał:

>
> IRV vs Benham does look strange, and I would like to investigate it
> further when I have the time. For instance, it's difficult to reconcile
> IRV's problems with not electing Condorcet winners with its low
> strategic susceptibility (since the existence of an unelected CW is a
> strategy opportunity).
>
>
Well, applying an electoral method to an election space (let's ignore
cycles for a moment and assume that it's a space with an honest CW in every
case) divides this space into unmanipulable elections (with resistant CWs)
and manipulable elections (with vulnerable CWs). In the manipulable space,
these manipulability stats don't care about whether we choose the honest CW
or not.
I guess the situation is that Benham and IRV create *more or less* the same
electoral space division, but don't agree about winners in the manipulable
subspace (meaning choosing / not choosing CWs - which doesn't create a
difference in stats).

>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.electorama.com/pipermail/election-methods-electorama.com/attachments/20240515/85ff84b1/attachment.htm>