What I'm curious about is what sort of data is available on these negative comments. If we were to identify (and agree) on a large enough sample population (n>30, probably) can we start seeing any trends in terms of downvotes, account age, posting frequency, etc? Do any of those correlate with a sample population of comments we identify as positive?
Other half-formed ideas I have that may be interesting:
- mandatory cooling off period (one to two weeks) before account activation
- analyzing downvotes by user and, once a threshold has been reached, putting them in a "time out."
- being able to block certain users from appearing in stories once you've logged in