If the numbers quoted by Daniel Linden can be believed, close to 2,000 times per day, Residents pull down the Help menu, select “Report Abuse…” and fill out the form to report abuse of one sort or another. But, once they hit the button to send those reports on their way, some Linden has to look at all those ARs…a Sisyphean task, to say the least. LL has concluded that “something must be done” about this, as, obviously, putting more people on reviewing ARs will work for awhile, but won’t work forever. Daniel’s posting on the Grand Unified Linden Blog says:
The main thrust of this project is to move our process away from the current one report/one resolution model and towards a system that with will quickly and accurately identity and manage those individuals and behavior that make Second Life feel unwelcoming or unsafe. The revised system will focus also on moving problems towards more useful paths for resolution – specifically by enabling and encouraging the development of inter-Resident mediation and dispute resolution options for those issues Linden Lab isn’t equipped to resolve. A further emphasis will be placed on self-resolution — by improving existing tools like mute and parcel-based access restrictions.
But what does that mean, exactly? Tateru, writing for Second Life Insider, seems to be waiting for the other shoe to drop. Prokofy, on the other hand, writing for the Second Life Herald, sees the specter of Resident government rearing its ugly head. Apparently, that’s been proposed at least once…and most Residents preferred the “benevolent dictatorship” of the Lindens. But if that threatens to devolve into anarchy because the Lindens are overwhelmed…what then?
Perhaps the answer lies, not in a full-fledged “Resident government,” but in a smaller-scale endeavor, a “peer court” system, if you will. Here are some sketchy notes on how such a thing might work; parties for whom this piques their interest are invited to comment on it and fill in any holes I might have left. (And I will not be banning anyone from this blog. I won’t even delete comments unless they’re obviously garbage, and I pledge to make a note of it if I am forced to do so. Akismet should catch all the obvious spam anyhow.)
Appointment to the “peer court” would be by petition, with each Resident wishing to become a peer judge required to submit the signatures (validated in some fashion) of N other Residents, where N is some relatively-large number, large enough to make gaming the system impractical, yet small enough that people won’t give up in frustration. I’m thinking N=100 for starters. Additional restrictions might be imposed on signatures, too, such as “no more than X% of signatures can share a group membership with the petitioner.”
Peer judges would be randomly assigned Abuse Reports to review, where they would make recommendations to LL as to what the consequences should be. (There should be a set of “best practices” for peer judges to follow, which would have to be carefully drawn up beforehand.) They would be permitted to question the victim, the accused, and any witnesses, but would be required to render a recommendation within some small time limit (72 hours?). Note that this would just be a recommendation, which LL would have to either confirm or not. The recommendation would include a report on anything the judge found 0ut in the course of investigation that would justify the intended consequence. At this stage, a lot of the obvious junk can be weeded out and the clear-cut cases handled quickly; extra time would have to be taken to review the ARs where facts are in dispute.
Since LL would review recommendations before implementing them, it would act as both the “court of last resort” in appeal and a check on the performance of the peer judges; peer judges who diverged too often from the “best practices” would be required to justify themselves to LL, on pain of removal from their peer judgeships and other possible disciplinary action. If they can justify themselves…well, perhaps the “best practices” need to be revised; this possibility should be accounted for. There would also need to be a “grievance procedure” for Residents who felt they hadn’t received a fair shake at the hands of the peer judges.
Now, I will acknowledge that there are probably more holes in the above scheme than in a shotgunned Swiss cheese; in particular, I’m assuming that all participants are reasonable people. It’s probably closer to the truth that people would try to game this system six ways from Sunday. Griefers would try to get appointed as peer judges so they can let all their friends off scot-free; the random assignment of ARs to peer judges would make this a dicey proposition at best, but it must be considered. (One might be required to pass a “background check” with LL before becoming a peer judge…but would that encourage griefer groups to try a “carnival booth attack” to find someone that would slip past that requirement?) Peer judges might act capriciously in dispensing justice, letting their friends off and throwing the book at their enemies; would LL’s review of their actions against the “best practices” be enough to act as a check on this? And real troublemakers would be the type that would cry “foul!” on the judges even when there was no impropriety; would this result in more work for LL, having to deal with grievances filed by every two-bit cagegunner who doesn’t like the thought of being held responsible for his actions? There’s also a big issue to consider: what’s in it for the peer judges? Why would someone become one? Would they be compensated for this work? (Probably not…LL has been moving away from compensation schemes like that. Example: the end of Instructor subsidies.) And I won’t even get into the issue that LL itself may be biased towards or against certain people…though some who read this will no doubt be thinking of that.
Perhaps LL would implement such a scheme, if it does so, as a pilot program, sending some ARs through this system (at random) while handling the others in the old way. This would be necessary to keep the initial set of peer judges from being swamped, too, until their numbers can be ramped up (presuming the scheme is found to work).
Anyway, that’s just one possible way LL could choose to handle this. Is there another way that would work better? If I’m completely off-base (not out of the question), tell me why. If this scheme could be improved (highly likely), tell me how. Just remember: If LL intends to devolve some of these functions on Residents, it would behoove us to carry on a discussion beforehand as to how that might be accomplished, rather than wait until LL imposes a solution on us. (Which they may very well do anyway…but I’d rather see some discussion of this before that happens than none at all.)