[EM] Let's clear up some confusion

Wed Oct 3 10:37:13 PDT 2012

On Wed, Oct 3, 2012 at 12:39 PM, Juho Laatu <juho4880 at yahoo.co.uk> wrote:

> Do you assume that the voters will know that the method will treat tied first and tied last in a different way than tied middle?

No. I assume that voters will be assured (correctly) that they never
have any reason to vote anyone over their favorite. They'll be assured
that they never have any need or incentive to not rank their favorite
in 1st place.

And they'll also be assured that, when making out there ballot, when
they've voted for some candidates there, they have no need to add any
other candidates to their ballot in order to fully help the candidates
for whom they've already voted.

The first paragraph above is about FBC. The 2nd paragraph above is
about Later-No-Help.

An additional result of Later-No-Help compliance is that if if there
are unacceptable candidates who could win, so that the only important
thing is ensuring that an acceptable, instead of an unacceptable will
win, then they have no reason to rank any unacceptable candidates.

None of those guarantees can be made for unimproved Condorcet ("Strong
Condorcet"), but they can all be made for Symmetrical ICT.

Additionally, as I said, Symmetrical ICT doesn't have the chicken
dilemma, but Strong Condorcet has the chicken dilemma.

Juho continued:

If they know, then you >could say that the interpretation and sincere
wishes of the voters are different for the middle preferences.

[endquote]

You'd have to re-word that. I don't know what it means. Yes, things
can be said about equal ranking at top, and at bottom, that can't be
said about equal ranking at middle. That's why Symmetrical ICT
interprets equal ranking differently only at top and bottom, where
certain clear statements can reliably be made, regarding the voter's
preference, intent, and wishes--statements that call for Symmetrical
ICT's different interpretation of equal ranking at top and at bottom.

Juho continues:

> (In that case, probably you should include that difference also in the definition of what the ballots mean.)

Wrong. My definition of Symmetrical ICT fully specifies the method and
its count rule.

I stated, above, the guarantees of which the voter can be assured,
when the method is Symmetrical ICT, but not when the method is
unimproved Condorcet ("Strong Condorcet").

Michael Ossipoff

(The text copied below doesn't contain any new answers from me, or any
new comments from Juho or anyone. It's included only for possible use
for reference, to clarify what Juho was referring to in his
above-quoted comments and questions.)

>
>
> On 3.10.2012, at 14.56, Michael Ossipoff wrote:
>
>> On Wed, Oct 3, 2012 at 7:24 AM, Juho Laatu <juho4880 at yahoo.co.uk> wrote:
>>
>>> Yes, it seems that the interpretation of the ballots and sincere wishes of the voters are the same in both traditional ranked ballots and your improved >approach.
>>
>> First of all, it isn't _my_ improved approach. It's Kevin Venzke's
>> improved approach.
>>
>> (My innovation was to do it at bottom as well as at top. In fact, I'd
>> proposed it at bottom long ago. At that time I called it "power
>> truncation").
>>
>> Yes, the wishes of the voters don't change, just because one count
>> rule respects their wishes and another doesn't. You're right about
>> that.
>>
>> But no, the interpretation of ballots is not the same in Improved
>> Condorcet and unimproved Condorcet. Improved Condorcet respects the
>> preferences, intent and wishes of equal top and equal bottom ranking
>> voters. Unimproved Condorcet doesn't respect their intent and wishes.
>>
>> Let me again state the definition of Symmetrical ICT, to show how it
>> differs from the versions of unimproved Condorcet:
>>
>> Symmetrical ICT:
>>
>> (X>Y) means the number of people ranking X over Y.
>> (Y>X) means the number of peoiple ranking Y over X.
>> (X=Y)T means the number of people ranking X and Y at top.
>> (X=Y)B means the number of people ranking X and Y at bottom.
>>
>> X beats Y iff (X>Y) + (X=Y)B > (Y<X) + (X=Y)T.
>>
>> [end of Symmetrical ICT definition]
>>
>> So no, Improved Condorcet and unimproved Condorcet do not interpret
>> ballots in the same way.
>>
>>> And the interpretation is the same for all ranks, except that the first and last ranks do not have any candidates above or below.
>>
>> Yes, and that's why Symmetrical ICT treats equal top and equal bottom
>> ranking differently, in keeping with (as I said) the preferences,
>> intent and wishes of the equal top and equal bottom ranking voters.
>>
>> Some here don't like to hear this: The emperor (unimproved Condorcet)
>> doesn't have any clothes.
>>
>> Mike Ossipoff
>>
>>
>>
>>
>>
>>>
>>> Juho
>>>
>>>
>>> On 3.10.2012, at 13.53, Michael Ossipoff wrote:
>>>
>>>> On Wed, Oct 3, 2012 at 3:25 AM, Juho Laatu <juho4880 at yahoo.co.uk> wrote:
>>>>> You explanation sounds like a pretty regular ranked ballot approach. If I rank U and V second, I want them to lose to the firsts and win the rest.
>>>>
>>>> Quite so.
>>>>
>>>> And (regarding your 2nd-ranked candidates), it's because you want
>>>> someone else (your 1st ranked) to win more than you want your 2nd
>>>> ranked to win, and because you also want your 2nd ranked to win more
>>>> than you want your 3rd ranked to win--That's what makes the top and
>>>> bottom rank positions different from all of the other rank positions.
>>>>
>>>> Your top-ranked candidates: You'd prefer that they win instead of anyone else.
>>>>
>>>> Your bottom-ranked candidates. You'd prefer that anyone but them wins.
>>>>
>>>> Neither of those things can be said for any other rank position, other
>>>> than top or bottom rank position. For the reason that you stated in
>>>> your above-quoted text.
>>>>
>>>> That's why, in keeping with what the voter would prefer and wishes
>>>> with hir equal top and equal bottom rankings, Symmetrical ICT
>>>> interprets equal top and bottom ranking as it does. That's why no
>>>> other rank positions are treated in that way--because the voter intent
>>>> and preference that I refer to at top and bottom rank position doesn't
>>>> apply at any other rank position.
>>>>
>>>> Because, when ranking X and Y in 1st place, you'd prefer that the
>>>> winner be from {X,Y}, then you don't want either to pairwise-beat the
>>>> other, which could change the winner from someone in {X,Y} to someone
>>>> else, like your last choice. So Symmetrical ICT lets you have your
>>>> ballot counted as automatically voting between X and Y in such a way
>>>> as to keep either from beating the other.
>>>>
>>>> It's your vote. It's your ballot, and it's your pairwise vote between
>>>> X and Y. It should be counted in your best interest, in keeping with
>>>> what you prefer and intend, when ranking X and Y equal top, or when
>>>> ranking W and Z equal bottom.
>>>>
>>>> Mike Ossipoff
>>>>
>>>>
>>>>
>>>>>
>>>>> Juho
>>>>>
>>>>>
>>>>> On 3.10.2012, at 6.06, Michael Ossipoff wrote:
>>>>>
>>>>>> Juho:
>>>>>>
>>>>>> In improved Condorcet, the voter who equal top ranks X and Y, or who
>>>>>> equal bottom ranks W and Z, doesn't have any more power to vote one
>>>>>> over the other, or to not do so, than any otther voter has to vote one
>>>>>> candidate over the other or no do so.
>>>>>>
>>>>>> Nor does a vote for X over Y, or for Y over X, counted for the ballot
>>>>>> of a voter top ranking X and Y, have any more power or effect as a
>>>>>> pairwise vote cast by any voter between any two candidates.
>>>>>> Likewise for the equal bottom ranking voter who ranks W and Z at
>>>>>> bottom. ("at bottom" means not voted over anyone).
>>>>>>
>>>>>> So then, what makes Improved Condorcet different from unimproved
>>>>>> Condorcet?  How is it more favorable to the equal top or equal bottom
>>>>>> ranking voter, without giving undue power to that voter?:
>>>>>>
>>>>>> With respect to X and Y, hir ballot is counted in hir beat interest,
>>>>>> in keeping with hir preferences, intent and wishes.
>>>>>>
>>>>>> As for what that means, I'll say it again:
>>>>>>
>>>>>> If you rank X and Y both in 1st place, that means that you'd rather
>>>>>> elect one of them (either one of them) than anyone whom you don't rank
>>>>>> in 1st place.
>>>>>>
>>>>>> If you rank W and Z at bottom, that means that you'd rather elect
>>>>>> anyone whom you rank above bottom, instead of W or Z.
>>>>>>
>>>>>> Mike Ossipoff
>>>>>> ----
>>>>>> Election-Methods mailing list - see http://electorama.com/em for list info
>>>>>
>>>>> ----
>>>>> Election-Methods mailing list - see http://electorama.com/em for list info
>>>> ----
>>>> Election-Methods mailing list - see http://electorama.com/em for list info
>>>
>>> ----
>>> Election-Methods mailing list - see http://electorama.com/em for list info
>> ----
>> Election-Methods mailing list - see http://electorama.com/em for list info
>
> ----
> Election-Methods mailing list - see http://electorama.com/em for list info