Meta Stack Exchange is where users like you discuss bugs, features, and support issues that affect the software powering all 158 Stack Exchange communities.

What is meta?
Here's how it works:
  1. Any Stack Exchange user can ask a question
  2. The community provides support, votes on ideas, and reports bugs
  3. Your voice helps shape the way Stack Exchange operates

Determining whether a specific user is a sock puppet is not an exact science. While many cases are clear-cut, the line can get a lot fuzzier when for example coworkers and friends are involved. So while I'm confident that we moderators get it right most of the time, there is still a chance of error.

The three main options for dealing with vote fraud are

  • Invalidate the votes between the users and suspend them (invalidation can only be performed by SE employees)
  • Delete the sock puppet
  • Merge the sock puppet into the main account

In ambiguous cases, moderators will usually choose the first option. It deals effectively with the vote fraud and is by far the least dangerous option.

The problematic part is merging, which used to be the default way of dealing with sock puppets. Merging is an incredibly dangerous tool, and the reason for my feature request is that I think it should never be used without explicit consent of the account owners. A hostile merge like one performed for sock puppets also merges login credentials, potentially giving a user access to a different account if such a merge was performed in error. A merge is also not easily reversible, though some rudimentary undo functionality exists now.

I'd like to make merges consensual in all cases, moderators should only merge users if the user themselves request it. This means that we need a different tool for dealing with obvious sock puppets.

So what I imagine is a "This is a sock puppet of user x" button in the mod menu. This button would invalidate votes (pending SE approval, as mods can't do that), probably transfer ownership of all content to the sock puppeteer and delete the account (a reversible deletion might be nice, but probably too complicated to implement). It would explicitly not transfer OpenID logins or any other profile information. The idea is that the process is at least somewhat reversible and safe, that it can't accidentally expose a users account to another user.

Another reason I'm proposing this is that the exact way that sock puppets are dealt with has changed over time, and I suspect that many moderators are not aware of the current SE policy on that subject. Most moderators (except for SO and maybe trilogy mods) don't have to deal with sock puppets often. Making the sock puppet removal an explicit option removes the need to educate all users about the process and all the potential dangers and reduces the chance of mistakes.

share|improve this question
I like this. +several million if I could – ChrisF Nov 13 '12 at 20:29
@ChrisF Make a few million sock puppets now while you still can, then! – JNK Nov 13 '12 at 20:49
At the least, I'd like it to queue invalidation and automatic mod messages with a custom 'please don't vote for posts just because your friend wrote it, looks like fraud, you don't want that ..' message to those involved, along with annotation of their accounts. All an employee has to do is agree and it's done. – Tim Post Nov 14 '12 at 11:49

Here are a few improvements I'd like to see around Moderator tooling for Sock puppetry and targetted upvoting:

  • Allow us to escalate handling a sock puppet or targetted upvoter to the SE community team without having to go through moderator chat. Perhaps through a queue. The issue with moderator chat is that it is hard to follow up and make sure your message to the SE team didn't fall through the cracks. I've had some instances where I left a message in chat about sock puppets and never received a response.

  • Allow users to flag as 'suspected sock puppet upvoting' -- this should only be enabled for users with a certain internal flag weight (or a certain ratio of good to bad flags, since you're not doing that whole flag weight thing anymore); this will keep the everyday users from spamming about it, but would allow trusted users to flag for that. Every user should be able to flag their questions or answers as having been serially down voted, though.

  • Once the SE team 'approves' or determines there is sock puppet upvoting, stick it back into the moderator queue for us to send a message to that user -- the message could even be filled out for us using one of the two messages we have for sock puppetry and targetted up voting

share|improve this answer
Just require 15k reputation for such flags. That'd be a nice filler for the large 10k..20k gap. – ThiefMaster Oct 23 '13 at 11:55
@Thiefmaster Like it. +1 – George Stocker Oct 23 '13 at 11:56
"I agree that chat is a lousy tool for these most of the time. It's pretty good for discussions, advice, and very time-sensitive things, but more and more of these requests end up being things that should just go into a queue.". I thought of ChrisF's feature request there when I read this, and both seem to be trying to solve the same problem. It isn't just sock puppets, we're seeing piles of flags each day for content dissociation requests, removal of sensitive information, and other things only devs can handle. – Brad Larson Oct 23 '13 at 14:52

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .