[EM] Comments on Approval posting

[Markus Schulze]

Dear Russ,

you wrote (26 Jan 2005):
> Also, when you insisted that I post pseudocode for CSSD at the site, I
> learned a lot about your level of sophistication. I told you that the
> Python code that I had already written essentially serves as pseudocode.
> That's partly why Python is a very "popular" language. (Note that I
> defined the word "popular" above for Mike's sake.) But you said that you
> couldn't understand the Python code. Why? Apparently because it doesn't
> have explicit "endif" and "endfor" statements. You must have a very
> simple mind, Mike. Also, the pseudocode you sent me was poorly organized
> and had errors, which you never noticed even after I asked you to check
> it out on the website. But if someone else had noticed the errors, I
> have no doubt that you would have tried to blame them on me. That's the
> kind of person you are, Mike. You've already proven it.

Also appendix 2 of my paper contains a pseudocode for CSSD:
http://groups.yahoo.com/group/election-methods-list/files/schulze1.zip

My pseudocode has two advantages.

First: It has a runtime of O(N^3), where N is the number of candidates.

Second: When there are pairwise defeats each with the same number of votes
for the winner, then it uses the margins of these pairwise defeats to
decide which of these pairwise defeats is the strongest. This is also how
the Debian project handles such situations. See appendix A ("Standard
Resolution Procedure"), section A.6 ("Vote Counting"), paragraph 7.1 of:
http://www.debian.org/devel/constitution

Markus Schulze