alt Hacker NewsAs always, thanks for the transparency.
It seems that people, even "established HN users" will flag literally anything. Do you feel that there is any remaining article quality signal that can be obtained from the current flagging mechanism?
As always, thanks for the transparency.
It seems that people, even "established HN users" will flag literally anything. Do you feel that there is any remaining article quality signal that can be obtained from the current flagging mechanism?
If the above list gives the mistaken impression that flagging is basically random, that's an artifact of the way I cherry-picked the list. The flagging system has problems, for sure, but it's a vital part of how HN's system functions.
If you squint and look closely, though, I think you can detect this in the above list. The weirdest "wtf?" cases of flagging are ones where the threads had a lot of comments and were on the frontpage. That means upvotes won the tug-of-war with flags, as they should have in most of those cases.
Conversely, it you look at the submissions in the list which had 0 comments or very few, it looks to me like most were either spam, low-quality articles, or dupes.
Remember, also, that some flags are just mistakes - the link is easy to fat-finger or misclick, and the UI doesn't provide feedback about that. That's likely to change soon as part of work that tomhow and I are planning.