[EM] smith/schwartz/landau

[robert bristow-johnson]

---------------------------- Original Message ----------------------------

Subject: Re: [EM] smith/schwartz/landau

From: "Curt" <accounts at museworld.com>

Date: Sun, March 25, 2018 2:12 am

To: "election-methods at lists.electorama.com" <election-methods at lists.electorama.com>

--------------------------------------------------------------------------



>

>

>> On Mar 24, 2018, at 9:21 PM, robert bristow-johnson <rbj at audioimagination.com> wrote:

>> ---------------------------- Original Message ----------------------------

>>

From: "Curt" <accounts at museworld.com <mailto:accounts at museworld.com>>

>> --------------------------------------------------------------------------

>>

>> > A: “elect a Condorcet Winner if one exists”

>> > B: “elect a winner other than the Condorcet Winner”.

>> > [within A: ]

>> > C: identify a “candidate or candidates that would defeat all other candidates head to head”

>> > D: “elect a single winner that is not a Condorcet Winner if a CW does not exist”.

>> >[…]

>> > Group D is “decisive” where Group C is not. In these cases I would argue decisiveness is overvalued.

>>

>> well, organizations and governments have to move on. they *need* answers and elections are held to provide answers.

>>

>> for a single-seat (usually executive) office, what would you suggest? a runoff?

>>

> Appropriate remedies might be a runoff after another period of consideration, or a power-sharing agreement. But this ties into what it means to have a multi-candidate Smith Set.

>> > What do you believe the Smith Set signifies? Is it meaningless to you other than something from which a winner should be algorithmically selected?

>>

>> it's not meaningless. it just need not be a concept coded in election law.

>>

> But I'm honestly curious what you believe the multi-candidate Smith Set signifies or means.
well, what i think about the Smith Set, whether it be single or multi-candidate set in a single-winner race, is that there is no compelling reason that any candidate in the complement of the Smith
set should be elected.  because the electorate is clear that there are other, more preferred, candidates in *any* scenario.  but it doesn't have to be encoded into law.

>> remember (i am not sure you got this point), Ranked-Pairs and Schulze do **not** select a winner from the Smith set.

>>

> Is this a semantic argument, or are you saying that ranked-pairs and schulze can elect single winners that are not in the Smith Set? Can you produce an example ballot set?
we're having a semantic difference.  RP and Schulze elect a candidate drawing from the entire pool of
candidates.  they do not identify a Smith set and then proceed to select from the Smith set.


so it's just an election procedure say like IRV is or Bucklin is.  same ballots, different way of looking at the ballots. and **if** a CW exists, the RP winner and Schulze winner will be the CW. (and it turns out that RP and Schulze will elect the same candidate in the case of a Smith set of 3,
and i am convinced a bigger Smith set is highly unlikely to occur in a real governmental election.  RP is simpler to explain to legislators and the public than is Schulze, in my opinion.  so it really just matters whether or not we consider the RP winner to be the best indication of the
least disapproved candidate.)

> I’m currently under the impression that is incorrect, as Wikipedia says both are Smith-compliant. I would love to be corrected if Wikipedia is wrong. I would not regard a method as Condorcet-compliant if it is not Smith-compliant.
i am not taking an issue with that.  i
understand why, in RP at least, why the winner is in the Smith set.  but it's not the same as a method that acts in one way to get the CW, and then, if no CW is forthcoming, does a secondary algorithm to pick the winner.  RP and Schulze do what they do on the entire pool of
candidates.
> At any rate, I am not here to lobby for a particular term for the methods in Group D. Only to argue that a distinction between groups C and D should exist, and without it, we do a disservice to Condorcet Methods by allowing them to be describe as “flawed” in ways
that only group D is. I am open to suggestions.
>
i just wanna know what we wanna call a ranked-ballot election method that elects the Condocet Winner when such exists.  what should such a class of methods be called?
--



r b-j                         rbj at audioimagination.com



"Imagination is more important than knowledge."

 
 
 
 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.electorama.com/pipermail/election-methods-electorama.com/attachments/20180326/825544de/attachment.html>