I don't know how the current flagging system works, but dealing with it in a slightly more intelligent way could reduce the number of false positives. Some basi

I don't know how the current flagging system works, but dealing with it in a slightly more intelligent way could reduce the number of false positives. Some basic filtering based on obviously spammy links, domains, language could be used. For flagged posts which aren't caught by such a spam filter, let the total number of flags decide. Right now it seems that it only takes a couple of flags to push a story off the front page.