slashdot nailed moderation, no one has attempted something similar. most systems are simple up/down vote or like/report
i am also starting to wonder if the agegroup being hired to implement "social" for websites is now young enough to have missed slashdot in it s prime.
the fact that people are still brainstorming from scratch instead of talking about how to improve slashdots model reeks of reinventing the wheel because they never heard of it.
That's me! Can you explain the Slashdot model and why it worked? Or point to a good write up about it somewhere else?
Slashdot's model was perhaps a little overcomplicated, but my favourite feature was the ability to tag up/down votes with flavours. +1 Informative was different to +1 Funny, and "Factually incorrect" was a different downvote to "Off-topic spam" (whatever they were called).
Other quirks off the top of my head: it capped at +5 and ... -1, I think? The score represented a thing closer to the up/down ratio than "Facebook likes". There was a dedicated -1 Overrated moderation for "I don't disagree that it's interesting, just not +5 interesting".
Also, logged in users got a fixed number of moderation points at random intervals, and you couldn't moderate in a story that you commented in. I'd like to believe this discouraged "throwing away" points on low-effort joke comments, but I'm not sure the facts of Slashdot comments entirely bears that out.
EDIT: And then there was meta-moderation...
especially all the Moderation, metamoderation, and Karma sections!!
A great deal of high-quality commentary was buried, however, often the best and most informative. That's fairliy much par for the course.
Much the early vibe on the site came from the fact that it was simply where intelligent people were commenting online -- especially the early Free Software crowd (well, early in terms of Web 1.0 -- there was the whole 1980s and early 1990s contingent as well).
ESR (before he went fully whackjob mode), Ted T'so, Alan Cox, Bruce Perens, Rasterman, and others.
Much that group seems split amongst HN, LKML, LWN, and Google+ these days, along with some blogs.
I'm about ready to blackhole that domain.
I don't think you're correctly identifying the problem. In my experience, the problem with comments, especially on news sites, is a glut of bad comments, rather than a lack of good comments. This solution doesn't disincentivize bad comments.