Re: Constitutional amendment: Condorcet/Clone Proof SSD vote tallying

[Sam Hartman <hartmans@debian.org>]

>>>>> "Andrew" == Andrew Pimlott <andrew@pimlott.net> writes:

    Andrew> ?  As far as I can see, all you need is enough D voters
    Andrew> that B voters can cause D beats A.

But if B voters can cause D beats A, how is this not honest?  If I'd
rather see B win or no decision made I rang A below D, otherwise if I
consider either acceptable I rank A above D.

To show a strategy, I think you need to show me a situation where it
is in my advantage to either:

1) Believing that A is acceptable and that I prefer either B or A to
   no decision getting made, rank A below D.


2) Believing that A is unacceptable and prefering no decision to A,
    rank A above D.

What you seem to be saying is that I prefer B over A strongly enough
that I'm willing to risk D winning, but I still find A unacceptable.
I think that seems rather unlikely, and many people who vote that way
are close to finding A unacceptable and thus actually end up voting
honestly.

I expect that if people are voting honestly on a Debian ballot about a
heated issue, you will see them rank the default option fairly near
the top of their ballot.  I.E. I suspect that where people feel
strongly about an issue, yes -> default -> no will be a much more
honest vote than yes -> no -> default.

--Sam