So as a moderator on Ask Ubuntu yesterday, I found a sockpuppet account and nuked it from orbit. I then suspended the puppeteer account through the standard moderator messages interface.
Usually when this happens people either admit to their crime or try to spin an epic tale of how their account was hacked or aliens did it... Or you know, their "friend" came round and used their wifi, slipped in the shower and accidentally gave them ~200 reputation over a couple of hours.
None of that happened this time.
Rather than lie, he demanded to know what rule he'd broken. I started rubbing my hands together, ready to get biblical... But where is using multiple accounts to artificially inflate your reputation against the rules? I really thought it was going to be easy but after ten minutes bumbling around the best I was able to come up with was section 4 in the Terms and Conditions (restrictions) that states:
Under no circumstances will Subscriber use the Network or the Service to [...] (c) create a false identity or to impersonate another person
It's a bit of a feeble interpretation. I couldn't really find anything in the documentation that would, for example, disallow me starting an Oli2 account (clearly meaning me) and voting for all my main account's posts.
I appreciate that the T&Cs is full of "SE can do whatever the hell they like" and that this user was just trying to troll a mod but I can't see how it would hurt to just lay out the ground rules that people are expected not to falsely or unfairly boost their or others' reputation. Could that be added?
You might think this sort of thing should be common sense but even when it's just friends in a voting circle, things can spiral out of control pretty fast. 15 months or so ago we expunged 1850 reputation between just three users who thought they weren't at fault. That doesn't happen often but when it does...
Another aspect to this is moderators aren't being told when the automatic checks and balances nullify suspicious votes. The only reason I found this sockpuppeteer is because he (and his puppet) were raining downvotes on another user. When I inspected their voting patterns, that was when I found that one account had been voting for the other, but some of the votes were already dealt with.
The user was able to vote again at a less suspicious rate and beat the system.
I really appreciate the auto-tools (they save us a ton of work) but dealing with the problem without dealing with the user isn't a great idea. If at all possible, when somebody has had a block of votes nuked, could the tools flag them for moderator/staff investigation?