[EM] Request for sincerity formulae and 2 sincerity numbers; SDSC

[Craig Carey]

Mike Ossipoff has been writing about reducing/minimizing the
"need for insincerity".

I request that the "need for insincerity" numbers be stated.
The moment they are defined, there could be a simple minimizing of
the need for insincerity quantities over all the finite number of
different sets of possible winners. Imaginably some highly
unsatisfactory preferential voting method would be found but I
guess that Mike Ossipoff can't estimate what the need for
insincerity is.

Can Mr Mike Ossipoff tell me whether the sincerity, or 'falsity',
of "voters" (in a Russ Paielli webpage sense) is perfectly
independent of the preferential voting method, ?.

The idea of sincerity does seem to be present in 3 of the criteria,
in a form where no one is able to change that idea into one of
"a need for insincerity".

http://home.pacbell.net/paielli/voting/criteria.html

SFC:
    "Members of that majority should be able to sincerely rank all
    the candidates..."
[That is about actual (in)sincerity, not any 'need for insincerity'.
"Should" means "it is true that"]


GSFC:
    "If there's no falsification of preferences, and if a majority
    of all ...
    The members of that majority should be able to sincerely rank
    all the candidates, and B still shouldn't ..."

[The word "falsification" really does seem to be all about actual
sincerity, rather than any "need" for its presence or diminution.
The words "should be able" mean "it is true that the majority is
able to ...". This is interesting since it suggests that sincerity
is a Boolean quantity. I did ask if sincerity was discrete rather
than continuous. We are not considering a "need for insincerity".
The page doesn't define either. Can you tell explicitly that
sincerity is Boolean. It could have more than 2 states.]


WDSC:
    "... without any member of that majority voting a less-liked
    candidate..."
[Here is an instance of embedding an entire undefined voting method
inside of a definition of a rule.  The majority might have spread
their opinions over 2,000 papers that have an average preference
list length of 8. Voting methods say whether or not candidates
win and lose, and many know methods do not return information on
which of 2 candidates that both either win or lose, is the more
preferred. "Liked" perhaps doesn't mean 'preferred' when there are
more than 1 winner.]




At 16:53 00.09.13 +1200 Wednesday, Craig Carey wrote:
>At 03:31 00.09.13 +0000 Wednesday, MIKE OSSIPOFF wrote:
>
>>In a reply to Markus I spoke of minimizing need for insincerity.
>>When I say that I want criteria that measure how well a method
>>minimizes need for insincerity, I mean that I'd like to
>>minimize the degree of insincerity needed, and it would be good
>>for that protection to extend to as many voters in as many
>>situations as possible.
>>
>>FBC doesn't in any way limit who is protected. SARC only

...

>Have you got any existing research into sincerity that you can copy
>into the mailing list?.
>
>This message's questions are retracted to the extent reasonable
>if Mike knows that the Russ Paielli sincerity rules are due to
>collapse or be withdrawn. This is a list on mathematics of
>preferential voting, not ideas that had to be withdrawn because only
>a bit undefined.

...


A comment. Mike asked for comments on how to fix his rules up.
This is not something that any subscriber need answer. The rules
are flawed. I suggest that Mike should instead check his rules
and discard the bad ones. What is this in "SDSC: Strong Defensive
Strategy Criterion"

      "If a majority of all the voters prefer A to B, then they
       should have a way of voting that will ensure that B won't
       win, ..."

No one can regard that as a good rule, when they consider what
is likely to happen in an election that has 10,000,000 candidates,
and that needs to elect 9,999,999 winners. In such an election, both
A and B will likely win. [The idea of a voter has been replaced with
the idea of a 'paper'].
SDSC is now a dead rule. Mike is asking for help to improve his
rules.  I wrote this to Russ Paielli and Mike Ossipoff in 27 August
2000:


>Ossipoff was thinking 1 winner (I suppose) but the definition does
>not say one winner. He got told and now you have the intentional lack of
>clarity on
>http://home.pacbell.net/paielli/voting/criteria.html

(parts of the site are at http://www.electionmethods.org/ )

To get SDSC improved, so that it does not reject 100% of multiwinner
methods, the number of winners needed to be constrained.
Why design a rule that rejects approx 100% of known preferential
voting methods that elects more than 1 winner?. It rejects the
Approval Vote. But that was not enough to lead to it being fixed.

Today it is like a wrecked French-British Concorde, destroyed
hopelessly as if SDSC was never once tested ever by anybody anywhere,
yet in a week it will coursing through the fumes tainted air again,
and online at the small website of Mr Russ Paielli.

Mike, can you comment on this please?: how many winners are there.
Why doesn't Russ Paielli fix problems when told about them?.

|
|    SDSC: Strong Defensive Strategy Criterion
|
|    [Same as WDSC except that "...over a more liked
|    candidate" is replaced by "...equal to or over a
|    more-liked candidate"]
|
|    If a majority of all the voters prefer A to B,
|    then they should have a way of voting that will
|    ensure that B won't win, without any member of
|    that majority voting a less-liked candidate equal
|    to or over a more-liked one.
|