election-methods@mailman.electorama.com

Technical discussion of election methods

View all threads

Fwd: Fwd: agenda landau winner

SS
Susan Simmons
Sat, Jul 31, 2021 1:46 AM

Sent from my MetroPCS 4G LTE Android Device

-------- Mensaje original --------
De: Kristofer Munsterhjelm km_elmet@t-online.de
Fecha: 29/7/21 6:31 a. m. (GMT-08:00)
A: Susan Simmons suzerainsimmons@outlook.com
Asunto: Re: [EM] Fwd: agenda landau winner

On 7/29/21 3:32 AM, Susan Simmons wrote:

¿Am I correct in understanding that in this agenda contexto Min Pairwise
Support would satisfy strong mono-raise, along with Range in general,
Above MidRange Approval, and Chiastic Approval?.. (my favorite agenda
setters)

Preliminary result: every monotone rating system where the score (some
element of a total ordering) of a candidate is a function of only that
candidate's ratings, and the social ordering is in order of candidate
score, passes strong mono-raise as long as an uprating of the winner
isn't accompanied by a downrating of someone else.

This includes, I think, Range, MJ, and Approval. I haven't checked
Chiastic Approval.

This class of rating methods is also the one for which Approval strategy
is an optimum under perfect information. They also all pass a
restrictive form of IIA (IIA when the scale is independent of the
candidates' presence). Maybe we should find a name for the whole class.

I haven't checked min pairwise support either; I'll look into it.

-km

Sent from my MetroPCS 4G LTE Android Device -------- Mensaje original -------- De: Kristofer Munsterhjelm <km_elmet@t-online.de> Fecha: 29/7/21 6:31 a. m. (GMT-08:00) A: Susan Simmons <suzerainsimmons@outlook.com> Asunto: Re: [EM] Fwd: agenda landau winner On 7/29/21 3:32 AM, Susan Simmons wrote: > ¿Am I correct in understanding that in this agenda contexto Min Pairwise > Support would satisfy strong mono-raise, along with Range in general, > Above MidRange Approval, and Chiastic Approval?.. (my favorite agenda > setters) Preliminary result: every monotone rating system where the score (some element of a total ordering) of a candidate is a function of only that candidate's ratings, and the social ordering is in order of candidate score, passes strong mono-raise as long as an uprating of the winner isn't accompanied by a downrating of someone else. This includes, I think, Range, MJ, and Approval. I haven't checked Chiastic Approval. This class of rating methods is also the one for which Approval strategy is an optimum under perfect information. They also all pass a restrictive form of IIA (IIA when the scale is independent of the candidates' presence). Maybe we should find a name for the whole class. I haven't checked min pairwise support either; I'll look into it. -km