On Wed, Nov 29, 2000 at 06:39:16PM +1000, Anthony Towns wrote: > Huh? Look, all I'm trying to say is that the straightforward and > obvious reading of the constitution leads to a result that doesn't > make any sense. > > I should expand on my concern about your interpretation of the > constitution. If you're claiming that A.6(2) is ambiguous, and doesn't > mean what, IMO, it clearly and obviously means, and if your claim has > any merit, then I don't see how it's possible to have the constitution > ever be unambiguous. If we use traditional mathsy terms and say > things must be `strictly greater' to ensure no one mistakenly assumes > `greater than or equal to', somehow this translates to increased > ambiguity rather than less. I don't see how anyone could hope to write > a clear and unambiguous constitution under those circumstances.
What's ambiguous is the concept of preference, in a circular tie. If we accept it as axiomatic that a result which doesn't make sense is preferable over one which does, then we get your kind of interpretation -- where the concept of "preference" is purely a matter of pairwise comparison. If we accept it as axiomatic that a result which does make sense is preferable over one which doesn't, then we get my kind of interpretation -- where the concept of "preference" must consider not only pairwise comparison but transitive effects. If you want it expressed in "mathy" terms, the question is: does "strictly prefer" refer to a lattice, or some other [less strict] sort of relation? The constitution doesn't explicitly say. > > But: I'm not asking "are there other methods", I'm asking "what's a > > better criteria?" and "why?" > > And I'm telling you, I don't know. There are plenty of fairness criteria > out there, there are plenty of justifications for resolving circular > ties in various ways. Ok, and I'm telling you that I, personally, at least, don't find that a convincing reason to accept that the single transferrable vote mechanism of the constitution should be eliminated. I'm willing to entertain the concept, if someone wants to put forth some kind of argument about why some other method would be better, but until then, what's the point? > There are plenty of fairness criteria on the fortunecity page, many of > which aren't met by STV alone. I've no idea how many of them are met > when you apply STV to the Smith set: certainly more of them than when > you apply it to the entire vote, but I don't know how many. And then there's statistical effects -- how many of them are met "most of the time". [But, by the way, I don't consider "number of fairness criteria mentioned on the fortunecity page" as some kind of absolute metric of vote goodness.] > There's a discussion of this in the February archives of this list. I'm kind of busy right now. Is there anything in particular that I should look for, if I read those archives? [Like, maybe, some kind of concept of "vote goodness" that the constutions could be measured against?] > > I'll note that the URL you cited doesn't have anything equivalent to > > Single Transferrable Vote. > > http://www.fortunecity.com/meltingpot/harrow/124/methods.html#IRV > > It's the second method listed. Hmm.. at first glance, this isn't the same. It doesn't make any kind of explicit reference to first preference. However, you're probably right. > > It really sounds more as if you want to find faults in the constitution > > than you've thought this through and have a better alternative to propose. > > If you'd care to go back to the first few messages in this thread, you'll > find the constructive suggestions. Oddly, you won't find all that many > in my responses to you, since I've had to spend most of those rebutting > your continual assertions that I don't understand what I'm talking about. Eh? I'm not trying to say that you've never said anything positive -- you have (though, I read the begining of this thread and I didn't see anything constructive in the specific direction of suggesting constitutional ammendments). Thanks, -- Raul
