4 ms·
I think you're doing a good job in turning off flags for some devisive posts. I was just surprised at how easy it is to kill the post from just a couple (?) fla
by j_maffe 2y ago
I think you're doing a good job in turning off flags for some devisive posts. I was just surprised at how easy it is to kill the post from just a couple (?) flags especially after others have vouched for it. There was this weird behaviour going on where the post got killed then revived because some user vouched for it and then another flagged it.
It's also worthy of note that this system is ripe for abuse, because the visibility of the post is asymmetric. It gets killed as soon as it gets flagged, so no new users see it and hence hard for someone to vouch. There's also the obvious problem of ganging on a post to flag/vouch for it, which again seems like it can be easily abused by small groups.
- jazzyjackson 2y agoIs it really abuse for a small number of people to decide right off the bat that a discussion is going to be, erm, unproductive?
- j_maffe 2y agoIt is if the post consistently gets vouched for whenever others stumble across it. Of course if it only got flags then it likely represents a widespread opinion.
- pvg 2y agoSome users flag, others vouch, that's a disagreement rather than abuse in some obvious way.
- j_maffe 2y agoYes, but again see my previous comment about the asymmetry of visibility.
- pvg 2y agoThere are lots of such asymmetries (although yours isn't very accurate since posts don't get hidden the moment they are flagged) but however the system is tuned, it's still not evidence of abuse. People disagree on what stories belong where and the system may work differently than the way you prefer or assume but neither of these are evidence of abuse. That's not saying there isn't abuse it's just not clear how you reach that.
- j_maffe 2y agoYeah, you're right, it's not evidence of abuse. I just find the designed system is vulnerable to it. Whether that's the case would require more data.