3 ms·
Yes, you're correct, they were mostly racist or far-right subreddits. But that didn't ever make much sense to me, because even if the person was evil elsewhere,
by easton 5y ago
Yes, you're correct, they were mostly racist or far-right subreddits. But that didn't ever make much sense to me, because even if the person was evil elsewhere, a comment that didn't have that sort of content shouldn't be moderated IMO. Brigading should be detectable regardless of what subreddits a user is a part of, given it usually coincides with bizarre usage patterns (downvoting everything in new as soon as it appears or something).
I get moderation is hard, but this makes it too easy. What if someone was evil but then saw the error of their ways? Should they be forced to make a new account? Or what if they were trying to start debates (as unlikely as that is) in said evil subreddit?
- woodruffw 5y ago> I get moderation is hard, but this makes it too easy. What if someone was evil but then saw the error of their ways? Should they be forced to make a new account? Or what if they were trying to start debates (as unlikely as that is) in said evil subreddit? This is a great point, and it's a problem that's created by these extensions. I'd argue that it's a lesser problem than the original one, and so it's a pill I'm willing to swallow (and subsequently tackle). In general, I think that anonymized reputation and vouching systems are the way forwards for online communities. Vouching solves the "blacklisting" problem: if we incentivize high-reputation people to vouch for good-faith accounts, then people who fall into the categories you've mentioned can be correctly exempted from the filter. But that's awfully hard to do.