[EM] 1-Person-1-Vote has been abandoned.

[Stephane Rouillon]

I do not hate Middle, I just dislike under and over representation
when it comes from the system...

As for a complete mathematical and thorough definition of
reciprocal fairness, try this.

Suppose two sets, S1 the set of voters and S2 the set of candidates.
Suppose an electoral method that produces scores for each candidate.
If you can split S1 in |S2| subsets each of a cardinality equal to the score
obtained
by the corresponding candidate, you can link these two sets using a bijective
mapping. Each voter contributes to one and only one candidate.
If an electoral method produces scores that verify this property,
it respects reciprocal fairness.

Is this well defined enough for you?
Steph.

MIKE OSSIPOFF a écrit :

> Steph wrote:
>
> Mike, you have a way of twisting reality...
>
> Oh? Did I or did I not ask you & James if you could justify 1p1v
> in terms of something more fundamental? Was that or was it not quite
> a while ago? Did you do so during that time?
>
> Steph continued:
>
> I do not remember who was supporting me,
> maybe Mr Gilmour.
> If I remember well, an honest summary
> was more like:
>
> It seems two different understanding of 1p1v exist.
> Some people think that an equal opportunity
> for each voter is a valid definition of 1p1v.
>
> I reply:
>
> 1 person 1 equally-counted ballot is what it's always meant.
> It happens that those ballots are Plurality ballots, currently,
> in a number of countries including this one.
>
> But anyone can define something how they like, and of course that's
> valid too.
>
> Steph continued:
>
> Approval for example respects this definition.
> It provides different voting power between ballots
> but equal opportunity.
>
> I do agree about finding approval "fair" between
> voters. But as I said before, there is more to
> 1p1v than only fairness from the voters point of view.
>
> Fairness has to be respected from the voted point of view too.
> In that sense, plurality could be seen as non-respecting 1p1v
> when multiple clones come to share the same votes.
> In the same way, I have already explained why
> I consider Approval not passing the same test.
> Approval solves the splitting-vote issue but it cannot
> map a bijection between electors and candidates.
> So, even if approval is fair to voters, it is unfair to
> some candidates (namely the extreme ones, extremeties
> being defined according to the small number of alternative
> preferences, not according to ideology).
>
> Maybe the "1person-1vote" name is not what you would like
> because it does not correspond to the criteria I just explained.
> "Reciprocial fairness" might best fit. I do not care about the name,
> except if a lot of people see 1p1v as I do. If I remember well,
> some people did see things as I do.
>
> I reply:
>
> Name it what you want, but state it more precisely than you have.
> Tell exactly what 1p1v or Reciprocal Fairness requires, and why
> it should be accepted as a fundamental standard. Or justify it in
> terms of a fundamental standard. That was what I asked you do do a
> long time ago, and what you still haven't done.
>
> Steph continued:
>
> PS: I had mail problems, this is why I lost some of
> our arguments. If I remember well you did not answer neither:
>
> I reply:
>
> The mail problem that you had might have prevented your CW
> example from posting. I answered your argument about the desirability
> of electing extremists in violation of majority rule. I'll answer
> it again, maybe today, maybe on a subsequent day. The mail problem
> that you had might have kept you from receiving my reply, but it's
> in the archives. Either I'll post the message number, or will
> repeat my reply.
>
> Steph continued:
>
> -the example where approval lost a Condorcet winner.
>
> Would you tell me the message number of that posting in which
> you posted that example, or else re-post it? You know, no one claims
> that Approval meets the Condorcet Criterion. Approval fails CC as I
> define CC. As some others define CC, they simply say "Don't apply CC
> to Approval or Plurality [because if you did, they'd pass CC]."
> But, in any case, it's well established and universally accepted that
> Approval isn't a Condorcet Criterion method.
>
> Steph continued:
>
> -my proof attempt that margin and relative margin pass
> an extended Ideal Democratic Winner protection from truncation criteria.
>
> I reply:
>
> I replied to that. You'd posted a significantly weakened version
> of SFC, and said that relative margins passes it. wv of course passes
> the stronger original SFC & GSFC.
>
> Protection from truncation isn't such a useful approach to a criterion,
> since, for instance, Plurality & IRV don't even have offensive truncation
> strategy. But Plurality & IRV fail SFC & GSFC, as do
> the versions of Condorcet that measured defeat strength by margins
> or relative margins.
>
> Steph continued:
>
> -the philosophical point about an extremist governement as an attempt
> would benefit more to society than immuability caused by an always
> consensual method as approval.
>
> I reply:
>
> Excuse me? Approval causes immutability? If you want to vote someone
> out, then don't vote for him, and vote for someone better.
> As I've often said here, Approval quickly homes in on the voter median,
> and stays there. Maybe that's what you mean by immutability, but it's
> a misuse of the term, because the voters could change it if they wanted
> to. But the voter median candidate, when compared separately to any
> other candidate, is preferred to him by more people than vice-versa.
>
> Steph continued:
>
> Of course this is a matter of proportion,
> but the electoral method should not maintain forever the statu quo according
> to me.
>
> I reply:
>
> Do you mean that the electoral method shouldn't maintain a status quo
> chosen by Steph, or that, according to you, the electoral system
> shouldn't maintain the status quo? Even when leaving a particular
> status quo involves a big violation of majority rule?
>
> Steph continued:
>
> Sometime any extremity would be better than the consensus.
>
> I reply:
>
> Better according to whom? Not better according to the majority whose
> expressed wishes you'd violate in otder to go to the extreme.
>
> Some methods deny this by construction.
>
> I reply:
>
> Good.
>
> Even if you hate the middle, you should still prefer methods that
> honor majority rule. That's because if the voting system doesn't
> protect the voter median candidate, then voters will do whatever it
> takes to do so themselves. For instance, in Plurality, IRV, margins
> and relative margins, they'll often do so by the defensive strategy
> of burying their favorite.
>
> Steph continued:
>
> If you prefer to discuss with Donald and Craig, it is your prerogative.
>
> I reply:
>
> Should that be taken as a call for assistance from Don & Craig?
> Anyway, I've never evaded discussing these issues with you.
>
> MIKE OSSIPOFF a écrit :
>
> >Some time ago, two people on this list were advocating the use
> >of 1-person-1-vote as a criterion for judging methods.
> >
> >I pointed out to them that 1p1v is a rules criterion, rather than
> >a behavioral criterion. It simply says what someone believes a
> >method's rules should be like.
> >
> >I pointed out that EM members surely wouldn't accept a rules criterion
> >as a fundamental standard.
> >
> >I asked if the proponents of 1p1v consider it a fundamental standard,
> >a derived standard, or just a criterion, either of which need
> >justification in terms of a fundamental standard.
> >
> >I then asked those 2 people if they could justify 1p1v in terms
> >of a fundamental standard.
> >
> >Their failure to do so, after all this time, is their way of telling
> >us that they can't justify 1p1v in terms of a fundamental standard,
> >and that 1p1v apparently has no justification.
> >
> >Mike Ossipoff
> >
>
> _________________________________________________________________
> Help STOP SPAM: Try the new MSN 8 and get 2 months FREE*
> http://join.msn.com/?page=features/junkmail
>
> ----
> For more information about this list (subscribe, unsubscribe, FAQ, etc),
> please see http://www.eskimo.com/~robla/em

----
For more information about this list (subscribe, unsubscribe, FAQ, etc), 
please see http://www.eskimo.com/~robla/em