This is _not_ normative. See it more as an investigation into the possibility to use a wiki to manage quite formal processes.
- including an audit to see if there is really some shenanigans going on
This approach opens up lot's of new questions. Especially the vote classes are unclear now and that's not good. As I see it there seems to be more then two but less then three classes of votes if/when the WikiVote is implemented. Indirect Votes are encapsylated in the WikiVotes, but the question of where are the Direct Votes and how they could be trusted seems like a real integrity hazard, but actually they've been that way for a while --Juxo 14:24, 13 Oct 2003 (EEST)
- Best way to deal with the trust issue is to have a Consumerium Troll account for each untrusted or suspicious group of edits. For instance if you think every edit that mentions KitKat McFlurry is free advertising for the two companies that own those brand names, then, you can link all such edits together into a common Troll:Sugar-coated spam account, and see if they are promoting other products bad for your teeth. If so they may be being paid by the Federation of Dentists, etc..
- A Role: account like User:Mediator (to deal with disputes) or User:Terminator (to figure out which trolls or users are really being paid to skew ratings one way, and get rid of them) could also exist.
- The only difference between a Role: and a Troll: is that you associate an edit with a role before the fact, and with a troll only after it is made. But this is very important symmetry...
A better name for this is edits, votes and bets. Since this is all about the relationship between those things.
- It's actually about EXPERIMENTING to let SEVERAL ways of relating those thigns compete. Each faction might have its own way of doing it, and the least controversial faction might get to edit the Signal Wiki/Content Wiki. Obviously we already have factions forming:
Juxo position: betting is bad, voting is not much better, editing can be free-form like Wikipedia and managed by the same informal methods - anything that needs formal process goes to Content Wiki. Trolls argue that the Content Wiki will be about consensus so the Opinion Wiki or the troll proposed The Consumerium Exchange must be about controversy and create contrasts that let people make the choices that they count on Consumerium to help them make. So this implies a different kind of formality.
- Why not consider these two factions, and see if we can figure out how they would dance to make the long term Research Wiki work? We seem to be doing fine at this troll-friendly dance so far, and obviously other large public wikis could learn from our approach. Even when counter-trolls came in and tried to satirize the early trolls, they were easily responded to and their comments were absorbed, despite the fact that they were partially vandalism. But when people from a sysop power structure from another large public wiki came in to excuse their abominable sysop vandalism, they were driven off by trolls. So far, this is basically the right way: probably THE SAME PEOPLE were better treated when they came in as trolls making fun than when they came in as sysops claiming authority. This is a lesson to them, and maybe to everybody.
Bleagh. This permission-based model nonsense has to GO. It has no purpose. If this wiki is only for opinions, things that don't get to the Consumerium buying signal, then, who cares? NO one will go to this kind of bother. Recall that nupedia got TWENTY articles in the same time span than Wikipedia got TWO HUNDRED THOUSAND. That's what happens when you make people prove who they are. These are not "requirements", and if they were implemented, the whole thing wouldn't be used. So the clock's ticking on this.
Besides, until the whole process behind the signal is clear, there's no point trying to figure out what its interfaces are. So why not work on the Consumerium Process instead? One service cycle for this whole thing?