4 ms·
Ah, thanks. Title changed. edit Hm, seemed to preempt myself on that. I guess the post is fixed now.
by hoprocker 11y ago
Ah, thanks. Title changed.
edit Hm, seemed to preempt myself on that. I guess the post is fixed now.
- dang 11y agoSomeone emailed about it and we fixed it. That's the preferred way to ask us a question. We don't see all the posts that try to ask us something, but we do see all the emails. For completeness, here's what I told the user who emailed: Users flagged it. User flags make a post drop in rank and, when enough users do it, eventually kill the post. When there's an ongoing discussion, we typically unkill the post so that comments can continue. That's what I've done here.
- nkurz 11y agoWhen you say you "fixed it", does this include removing the rank penalty? Or does the post remain lowered in rank? I'd presumed that you tended to manually review flagged posts, and then make a judgement as to whether the flags are appropriate. But if even "inappropriate" use of flagging can keep posts off the front page, it's still a pretty powerful tool. I'd feel more comfortable knowing that permanent removal from the front page requires an explicit editorial decision, at least upon review.
- dang 11y agoIn most cases, we unkill the thread so it stays open for comments, but do not turn off the rank penalty. Occasionally, though, we do turn it off. Some easy cases are when a good article has a bad title and people were flagging the title (solution: change the title and turn off flags), or when there is evidence of a flagging ring (solution: turn off flags and ignore future flags from the offenders). The harder cases are posts like this one, where the community is divided on a political question, some want it on HN and others don't. In that case we usually stop short of overriding the rank penalty. The voting and flagging systems have been in a stable equilibrium for years, and we've learned that it's a mistake to fight that very much or very often. But we do do it sometimes, I'd guess maybe 10% of the time (but haven't looked at the data). For example, sometimes an article is so substantive and otherwise so obviously a good fit for HN that it doesn't seem fair to the community to let the flags have their usual say. Sometimes a thread turns out to be unexpectedly high-quality and not a flamewar. And sometimes a story is of so much interest to the community that we couldn't stop it even if we wanted to (a recent example being the Strange Loop/Yarvin thing), in which case we turn flagging off for hopefully the best submission on the topic and treat the others as dupes. In the end, there's no escaping editorial judgment, though not every HN user shares your positive view of it. We try not to over-exercise it, but never doing so would be suboptimal for the community. There's a tragedy of the commons paradox in there somewhere.
- hoprocker 11y agoThanks for the insight into how you folks try keep things sane on here.