On the list just now, I said that people proposing new methods
should either show that their method is better by our standards
or why we should change to their standards. I should add that of
course Schulze's method does meet our standards & criteria.
It could even be argued that those clone subcycle situations are
LO2E situations, so that Schulze does _better_ than SC, by LO2E.
I don't deny that.
As I said on the list, I wouldn't object to a proposal to indicate
to ER that Schulze's method is better than SC, adding Schulze's method
as a recommendation (but not withdrawing the recommendation of SC
or plain Condorcet(EM).
***
Also, I just said that, with Condorcet(EM), with regard to LO2E
or protecting a Condorcet winner, it isn't necessary rank a lower
choice equal to a higher choice. Actually, in a subcycle situation
that isn't entirely true. But those situations aren't as important
because their special, less-common situations. Also, as I've said,
I feel that the smaller a subset of the voters we're talking about,
the more we can expect them to get it together with regard to eachother's
candidates.
But I recognize that Schulze won't have subcycle fratricide in
a clone-set or near clone-set, and that constitutes some degree of
improvement when it comes to LO2E.
Mike
