CmdrTaco's Journal: Expanding Meta Moderation Context

Context is a tough problem for meta moderation... what we're asking users to do when they M2, is to decide a simple Yes/No on the moderation. Good/Bad? Fair/Unfair? Etc. This gives us a benchmark for accuracy of moderations... since M2 requires 5 "Votes" for each M1, we can usually see quite clearly if a user is moderating badly.

But context is tricky. There are facts that are not available to the meta moderator that might influence their decision. The most obvious fact is what the moderator did with their other 4 mod points. What about when a moderator uses all 5 mod points on a single user? What if they make a habit of it? I've found accounts where they've used 10-15 mod points on a single other user... every time they get 5 mod points, they immediately go moderate up 5 comments posted by their fortunate benefactor.

As often as not, a scan of the logs reveals that Moderator and Moderatee share an IP. In other words, the user simply waits until he gets points, and blows his wad on his other account.

Now meta moderation can't really solve this... and I don't know what a solution would be to this problem. It does happen. Maybe only a couple times a week, but it does. It's not scalable for an admin to manually seek out this sort of abuse.

We provide a 'See Context' link which allows you to get a glimpse of the environment where the post lives. Why not provide a 'See Context' link that shows what the moderator did with their other mod points? Maybe limit that to other comments posted by the same person... A simple one-liner note saying "This moderator has used N points on user Foo" with linkage.

It would add a level of complexity to meta moderation, and thats always something that I avoid doing... but when you stumble accross an M2 where you see a single user has spent double digits of mod points on a single user, you could at least check and see if each and every one of these posts were of sufficient quality to warrant a moderator singling out this user account.

At best, it would encourage users to spread their mod points around. And worst, it might just encourage a troll to create more accounts to use for his nefarious purposes.

Note As with most of my journal entries, these are just ideas I'm banging around. Most never get implemented because they are either daft... or we just don't have time. I'm always open to feedback, but don't freak out just because I'm thinking out loud and you disagree with something.

Automate it...

[singularity]

I think that it should be written into the moderation guidelines that excessive use of moderation points on one individual will be penalized by loss of future moderation privs. (as well as cancellation of the moderations done).

Then completely automate it. There is no reason for MM'ers to have3 to go through that. SlashCode could easily go through by itself and count. Once you hit a magic limit (both per five points and cumulative - no reason someone should use two points each time on his friend of himself

Context is enough

[fm6]

The only thing a metamoderator should care about is the appropriateness a mod. Is it really flamebait, or offtopic? Is the moderator just penalizing a difference of opinion? The fact that a moderator is expending all his points on a single user says nothing about that. I've done this myself, when I thought a user was being compulsively obnoxious, or particularly insightful. But I've never moderated my own posts -- I don't even have a second user name.

As for moderations that are posted from the same IP as

a fair extension of friend/foe

[JDizzy]

If I choose a nick as a friend, then I would not be very friendly if i didn't pump his/her comments. On the flip side I would scrutinize that nickname with a little red or orange dot with extra vengeance. For example, if I see the little red dot(foe, or foe of friend) when moderating the probability that I would simply not moderate very high, and a negative moderation even higher. What I'm saying here is that mod points plus the notion of a friend foe system invites this sort of scandal.

I propose that when

Re:a fair extension of friend/foe

[JDizzy]

As an aid to the meta-moderator, is it possible to put a line in somewhere that says "this nick was moderated N time previously by the same moderator"? Low numbers would be expected, and anything in the double digits is serriously questionable.