dave@cogsci.indiana.edu (David Chalmers) (11/20/89)
In article <2440@stl.stc.co.uk> "David Wright" <dww@stl.stc.co.uk> writes: >So the call for votes would list the following options: > >A comp.society.rabbit >B rec.pets.bunnies >C rec.pets.rabbits >D rec.rabbits >E sci.coneys >F talk.rabbit >N no This system as proposed won't work, precisely because of the awkward integration of the "NO" vote with the group name options. To see this, take a simpler example. A rec.aquaria B sci.aquaria N no Assume I don't care whether the group exists or not, but I feel very strongly against the name "sci.aquaria". How do I vote? There's no way I can express this group preference without effectively voting eith "YES" or "NO" to the group itself. Either I can vote (1) rec.aquaria (2) NO (3) sci.aquaria or (1) NO (2) rec.aquaria (3) sci.aquaria. But neither of these is right. The first will count as an effective YES vote for the group, demonstrating an enthusiasm for the group which is not present. Ditto for the second and "NO". There are other problems with this scheme (such as incorporating the 100-vote margin), but I won't go into them. Instead I'll repeat the *only* simple scheme which has a hope of working. (1) YES/NO on charter. (2) STV on name. This avoids the awkwardness of having to express feelings on the charter and on the name in the same breath. Only if the charter vote receives 100 more YES's than NO's (and maybe a 2/3 share) will the name votes need to be examined. Then, everybody's preference counts. On the fish vote, for instance, I could abstain from the charter vote while voting [1] rec.aquaria on the name. My preference will be taken into account only if needed. Making the name vote STV rather than MAUVE allows a much more satisfactory range of preferences to be indicated. I know there have been a thousand proposals, but it seems to me that this is the only scheme that doesn't have terrible problems. -- Dave Chalmers (dave@cogsci.indiana.edu) Concepts and Cognition, Indiana University. "To live outside the law you must be honest..."
bee@cs.purdue.EDU (Zaphod Beeblebrox) (11/20/89)
Said dave@cogsci.indiana.edu (David Chalmers): (in article <29992@iuvax.cs.indiana.edu>) |Assume I don't care whether the group exists or not, but I feel very |strongly against the name "sci.aquaria". How do I vote? There's no |way I can express this group preference without effectively voting |eith "YES" or "NO" to the group itself. Either I can vote | |(1) rec.aquaria (2) NO (3) sci.aquaria |or |(1) NO (2) rec.aquaria (3) sci.aquaria. | |But neither of these is right. The first will count as an effective YES |vote for the group, demonstrating an enthusiasm for the group which is not |present. Ditto for the second and "NO". This can be solved by modifying the STV scheme to allow for ABSTAIN as an option. Thus you would vote as follows: [ groups whose names you vote YES for in STV order ] ABSTAIN [ groups you don't care about ] NO GROUP [ groups you don't want to see created ] Thus for the above you would vote ABSTAIN rec.aquaria NO GROUP sci.aquaria Order is only important for the groups listed before ABSTAIN. If the current leader is one in your ABSTAIN list, your vote is neither YES nor NO. If the current leader is one in your NO GROUP list, then your vote is counted as NO. Note that your vote can change back from ABSTAIN to NO and back depending on what the current leader is. This makes for a system for which the instructions to the voter read as follows: "First list the groups you favor in order that you favor them. Then type ABSTAIN. Then list the groups for which you have no preference (possibly none). Then type NO GROUP. Then list the groups you oppose (also possibly none)." I believe that covers all the reasonable possibilities without being overly complicated. Private note to a few obnoxious flaming jerks (you know who you are): notice that I've changed my .signature. The reason I don't go by my first name is that I would then be confused with Mr. Templeton. B.E.E. -- Z. Beeblebrox | "Ich bin ein Berliner!" -- President Kennedy, 1961 (alias B. Elmore) | "Tear down this wall!" -- President Reagan, 1987 bee@cs.purdue.edu | ..!purdue!bee | Berlin Wall RIP 1961-1989
jbeard@ntvax.uucp (Jeff Beardsley) (11/21/89)
I think it should more properly be: 1: Yes/No/Abstain on the charter 2: STV on name This allows people who dont care whether a group exists or not, to vote against a particular name, without lending support to another name, or to the charter in particular. -- ------- Jeff Beardsley at UNT ----------- <jbeard@dept.csci.unt.edu> -------- altrec.fishy.flamescience: Alternate Rec Group for the scientific study of fishy flamage.
dave@cogsci.indiana.edu (David Chalmers) (11/22/89)
In article <1989Nov21.144817.14808@ntvax.uucp> jbeard@ntvax.UUCP (Jeff Beardsley) writes: >I think it should more properly be: > >1: Yes/No/Abstain on the charter >2: STV on name > >This allows people who dont care whether a group exists or not, to vote >against a particular name, without lending support to another name, or >to the charter in particular. Absolutely. This is what I intended, I'm sorry if I didn't make it explicit enough. Furthermore, there should also be an option to abstain from the name vote if all one cares about is the existence of the group. -- Dave Chalmers (dave@cogsci.indiana.edu) Concepts and Cognition, Indiana University. "To live outside the law you must be honest..."
dalamb@qucis.queensu.CA (David Lamb) (11/23/89)
In article <2440@stl.stc.co.uk> "David Wright" <dww@stl.stc.co.uk> writes: [ long description of STV] In article <29992@iuvax.cs.indiana.edu> dave@cogsci.indiana.edu (David Chalmers) writes: > >This system as proposed won't work, ... It's been a long time since I've looked at Political Economy but... If I remember correctly, in the 60's an economist named Arrow proved a theorem that, with a fairly simple and intuitive definition of "fair", there cannot be any "fair" voting schemes. Thus there will ALWAYS be voting anomalies. You have to decide what kind of anomalies are most important. An Australian friend (who used STV at home) said that STV results in a winner where more people were in favour of the winner than were opposed to the winner. That seems like a good characteristic for newsgroup creation to me. If you want to vote against a name, you write it last. If you want to vote in favour, you write it first. If you don't care, you write it in the middle. If you have choices you rate about equal, you fit them into a category (early, middle, late) and write them in a random order (preferably using dice). David Alex Lamb ARPA Internet: David.Lamb@cs.cmu.edu Department of Computing dalamb@qucis.queensu.ca and Information Science uucp: ...!utzoo!utcsri!qucis!dalamb Queen's University phone: (613) 545-6067 Kingston, Ontario, Canada K7L 3N6