Do you have any suggestions for things that might fix the problem? I have all kinds of ideas I'm half considering, but I'm curious to hear what you think. We're on mostly uncharted territory here.
Maybe I'll ask HN generally...
Do you have any suggestions for things that might fix the problem? I have all kinds of ideas I'm half considering, but I'm curious to hear what you think. We're on mostly uncharted territory here.
Maybe I'll ask HN generally...
Upvoting a comment that's mean-spirited, unhelpful, or otherwise of negative value should carry a visible cost. But who makes that decision? You can't rely on the community to do so, because that puts you in the same position. Perhaps you need 10% of your community to be "trusted elders," and they have the right to declare comments to be of negative value. Then any upvotes such a comment accrued, count as negative karma for the upvoter. There might be other, simpler and more visible ways to achieve the same net result.
I know good hackers are interested in many things besides hacking, but "intellectual curiosity" is being interpreted very broadly. Any definition will lead to people playing the lawyer game on the words, so that's a hard one to try to narrow down.
Perhaps bands of karma, expressed as a percentage of the population (or something) could be used to grant and deny more access to features. Making someone's position is the ranking visible will help that be more transparent. The idea would be that higher ranks get to decide if the decisions by lower ranks are right or wrong, and hence those of lower rank can advance if they behave in a manner you choose to reward.
Or something.
These are quick "toss in the air and see what happens" ideas. They won't work, but perhaps suitably clever people on HN can find mutations of them that will. A few ideas were being tossed around here:
http://news.ycombinator.com/item?id=2387873
... although that was more specifically to try to solve the "submission race" problem that jacquesm had identified. But some of those ideas, mated with some of the above, might produce more fit children.
"Perhaps bands of karma, expressed as a percentage of the population (or something) could be used to grant and deny more access to features. Making someone's position is the ranking visible will help that be more transparent."
I think any attempt to automate this will be susceptible to gaming, groupthink, and so forth, and it makes karma too much like a game.
Mutating them into something that does work is more important, interesting and useful. There's more discussion here:
http://news.ycombinator.com/item?id=2403696
Enjoy!
So, rather than trying to weed out poor up-votes, leave it to the community to manage the down-voting. Thus, a quality counter argument will be safe from down-voting and, hopefully, out-vote the poor quality post.
But if there's one thing I've learned from a lifetime in tech, it's that attempts at adjusting behaviour by indirect means always leads to unexpected behaviour.
a) It's a stupid comment b) It's a mean comment c) It's a troll d) I disagree with it (despite user education, it seems that people downvote for this reason all the time) e) I dislike the commenter (ad hominem)
Perhaps explicit options to flag a comment as a) stupid, b) mean or c) a troll would give you enough data to help solve the comment quality issue.