Steve Eppley speculated about what I mean by acceptance in Approval. But all I mean by it is voting yes on a candidate's offer to govern in one's behalf, by giving that candidate a vote in Approval.
And, by non-approval, I mean not doing that. As I'd said before, it's the voter's own business why s/he accepts or rejects a candidate. Maybe s/he goes by principle and rejects candidates who aren't good enough, regardless of winnabililty. For instance, maybe s/he rejects those who are dishonest or corrupt, or significantly so. Candidates who don't deserve a vote. I often say that I myself usually vote that way. Or maybe s/he asks "How far down do I have to compromise, in choosing whom to accept?" So maybe it's situational acceptability that s/he goes by. But, as I said, it's none of our business why a voter accepts or rejects a candidate. Just count hir ballot as voted. Steve speculated that maybe some rank methods would do a better job of minimizing the number of people to whom the outcome is unacceptable, according to some unknown definition of "unacceptable", a definition different from my simple operational definition. I doubt it, because the best rank methods, the ones that Steve and I advocate, are majoritarian methods, and they don't try to count what Approval counts. I want to repeat, though, that I'd rather give the voters one of the best rank methods (such as SSD, MAM, MDDA, or MAMPO) than Approval, because those best rank methods make less demands on voters' judgement about what they strategically need to accept, or about what an acceptable candidate is. The progressives who vote Democrat have been down so long that it looks like up to them. So I agree with Steve in preferring to give voters the best rank methods. The differences between the best rank methods, the ones that I listed above, are negligible. It seems to me that Steve doesn't agree with me on that. By the way, SSD was devised by Steve Eppley and me. Markus Schulze had previously posted about CSSD, and its equivalence with BeatpathWinner, but Steve and I were unaware of that. I'd thought that Markkus was describing a new variant of BeatpathWinner, and using "Schwartz set" with a different meaning. I probably didn't completely read those postings. Anyway, SSD is not CSSD, though they're equivalent in public elections. In small committees, they can give different results, and in small committees, (but not in public elections) SSD can violate clone-independence. But SSD has a more natural definition, with more obvious motivation and justification, and is the Condorcet version to propose for public elections. I have no criticism of MAM, but I've encountered non-understanding when offering it to a local Greens organization. SSD makes no mention of cycles, but MAM must either mention them or indirectly refer to them, even if not by name. I personally feel that SSD has a more publicly understandable definition. Mike Ossipoff ---- Election-Methods mailing list - see http://electorama.com/em for list info
