Hacker News Reader

An interesting idea. But, unfortunately, it's very open to abuse. There is no guarantee the flagged comment has been flagged correctly.

Really, you're stuck with the same problem: just as there could be mass upvotes of "poor" comments and mass downvotes of "good" comments, there could be mass flagging of "good" comments as "spam", "dumb", or "mean".

View on HN

It shouldn't be too hard to limit abuse. Set appropriate thresholds, give less weight to new accounts with low karma, ignore flags from users who frequently flag inappropriately (this last one could be automated by comparing upvotes from a list of "known good voters" against the flags. For example, if you flag stuff that pg upvotes, you're probably flagging inappropriately).

Reply on news.ycombinator.com