On Thu, Jun 12, 2003 at 05:26:57PM -0400, Nathanael Nerode wrote: > Incidentally, this is by *far* the most controversial aspect of the > amendment. This mechanism causes the "Condorcet winner"/"ideal > democratic winner" to lose under certain contrived circumstances. (They > are rather contrived, admittedly, and strong arguments have been made > that these circumstances will not really happen.)
When there's more than one condorcet winner, and voter participation is extremely low, this mechanism can cause the selection of a different condorcet winner than plain sequential dropping. Is that what you're talking about? Or are you talking about the fact that the entire vote is tossed out if the ideal democratic winner doesn't have enough votes to meet quorum? > Also, the mechanism used here is called "quorum" even though it bears only a > passing resemblance to ordinary quorums, which has been quite confusing. Perhaps, instead, we should have a quorum which, in essence, shuts down debian-vote during weeks or months where traffic is too low? That's about as close as we could get to conventional quorum. [Personally, I think this would be even more confusing and, in fact, would have some bad properties if it were closely investigated.] Or, perhaps, we should have a formal "pre-voting" period where people declare their willingness to participate in the vote. That would actually work reasonably closely to the original intent of quorum, but would add even more red tape to the voting process. Also, there are some strange results here from people registering on quorum and not voting, and vice versa. Anyways, of the quorum alternatives to what the amendment proposes, I think this is probably the best mechanism. [Interestingly enough, the people "confused" by the amendment's usage of the word quorum didn't seem to agree with me on this issue -- or, at least, this didn't turn into a formal amendment.] For some reason, some people think that quorum should be assessed after the vote and should be used to toss the vote if not enough votes were received. That has bad properties which can discourage some voters when participation is already low. Interestingly enough, most of the people who have been "confused" by the concept of quorum have been advocating this approach. In essence, the issues we're trying to address, with quorum, are different from the "Everybody meets in the big hall and votes, and if not enough people show up we can't vote" concept of classic quorum. If people are confused by the fact that we're voting on a mailing list (rather than in person), that's sad but it's not a reason to use an inferior mechanism. -- Raul
