[EM] (no subject)

[James Gilmour]

Dave Ketchum > Sent: Sunday, November 09, 2008 6:02 PM
> James seems to be stretching his interpretation a bit far.
> 
> Agreed that, while the voter can choose to rank all 
> candidates, the voter is permitted to omit those least desired.
> 
> In Condorcet every ballot is counted.  For each the counter considers EVERY 
> pair of candidates, such as A and B.  If the voter has indicated preferring 
> A>B, that is recorded toward A winning; likewise for B>A.
> 
> As to IRV, while using the same ballot, it only looks at enough to satisfy 
> it purpose - which DOES NOT INCLUDE knowing whether the voters like A 
> better than B.

I am not stretching my interpretation too far.

In elections to be counted by IRV or Condorcet rules voters will not mark preferences for candidates among whom they have no
preferences.  Thus in a four-candidate election, a ballot paper marked "A, B" indicates that this voter prefers "A" over "B" and
prefers both "A" and "B" over both "C" and "D", and it tells the Returning Officer that this voter has no preference between "C" and
"D".  In contrast, a ballot paper marked "A, B, C, D" has given the Returning Officer information about all possible preference
comparisons.

It is clear from the affidavits that ONE of the objections to IRV is that the ballot paper marked "A, B" will be treated differently
from the ballot paper marked "A, B, C, D", and hence the voting system will treat the two respective voters differently (and to such
an extent as to be "unconstitutional").

In a Condorcet count these two ballot papers (and hence the respective voters) would also be treated differently, because the voter
who marked the "A, B" ballot paper could not contribute a vote to the "C, D" pair-wise contest that is an essential part of
determining which candidate should be elected.

My question was simply that if the effect of THIS difference in an IRV count is sufficient to make IRV counting "unconstitutional",
why would the effect of THIS difference in a Condorcet count not be sufficient to make Condorcet counting also "unconstitutional"?
I could easily see how, on THIS ground, IRV counting and Condorcet counting could both be considered "constitutional" or could both
be considered "unconstitutional", but I have some difficulty is seeing how, on THIS ground, one could be considered "constitutional"
and the other "unconstitutional".

James Gilmour



> On Sun, 9 Nov 2008 16:20:10 -0000 James Gilmour wrote:
> > Kathy Dopp  > Sent: Sunday, November 09, 2008 2:17 AM
> > 
> >>Your statement oversimplifies and ignores details/differences
> >>between IRV and Condorcet.  IRV proponents may pretend not to 
> >>know that Condorcet methods do not exhibit most of the flaws 
> >>of IRV counting methods. For example, Condorcet, to my 
> >>knowledge treats all voters ballots equally, considers all 
> >>choices on all ballots, 
> > 
> > 
> > If I have understood the various submissions correctly, the 
> principal 
> > objection to IRV on THIS ground, is that the ballot papers 
> of voters 
> > who express different numbers of preferences are thereby treated 
> > differently, and in such a way and to such an extent that these 
> > differences should render the IRV voting system "unconstitutional".
> > 
> > It is correct that Condorcet counting considers all the preferences 
> > marked on the ballot papers, in a sequence of pair-wise contests.  
> > However, Condorcet counting has no option but to treat differently the 
> > ballot papers of voters who have expressed different numbers of 
> > preferences, because such voters will be excluded from some of the 
> > pair-wise counts.
> > 
> > If this difference in the treatment of ballot papers with different 
> > numbers of preferences would be a "fatal" flaw in IRV, would it not 
> > also be a "fatal" flaw in Condorcet counting, and indeed in any other 
> > voting system where voters may express different numbers of 
> > preferences?
> > 
> > James Gilmour
No virus found in this outgoing message.
Checked by AVG - http://www.avg.com 
Version: 8.0.175 / Virus Database: 270.9.0/1777 - Release Date: 09/11/2008 09:53