Hacker News new | past | comments | ask | show | jobs | submit login

I applaud your ability to maintain an optimistic viewpoint about humans even though all evidence points the opposite direction. What you're suggesting is completely unprecedented on the internet. What always happens is that an open forum is eventually overwhelmed by trolls and morons, and all the smart people give up. There has never been a case where a troll or a moron turned into a useful contributor because of thoughtful response from a good member of a forum. Just review the history of Digg, Slashdot, Reddit and almost any other open forum, all the way back to ancient USENET times...

Wasn't this forum created partially because reddit was overwhelmed by the hordes and all the interesting startup information was lost in the noise?




Actual trolls we'd ban. Merely dumb posts slide harmlessly into oblivion because they don't get any upvotes.

We'll never get the onslaught of 14 year olds they got at digg and reddit because most of them aren't interested in startups. You don't get idiots in your comment threads till you have this on your frontpage

http://power-robot.blogspot.com/2007/08/when-in-rome.html

so as long as we kill offtopic submissions, we should be safe.


Doubly damning is that the post is a fake, at least in the context everyone thinks it's in. It's a screenshot from a porn movie, not a random picture.

(Thinking out loud, just skip the rest of this.)

Posts like that do attract fourteen year olds, and once you do you can never, ever get rid of them. It's self-reinforcing.. Fourteen year olds attract other fourteen year olds, who create submissions like the one above, which attracts more fourteen year olds. One way to fight it is to implement a harsh banning system, the way Facepunch Studios did. Check out their ban list: http://forums.facepunchstudios.com/showbans.php ... I count 93 bans today alone. Most for a week or longer, complete with reasons and viewable to the public.

And yet, a brute-force approach like that only sorta works. What's needed are strong community values combined with a strong punishment system. But not so much as to stifle conversation innovation - just enough to filter the noise.

Karma works, but some people need to be made more equal than others in that case; someone that's been with a community from the beginning is a hundred times as valuable than a newcomer (not an exaggeration), because they help enforce community values. It follows that they should get a karma vote weight proportional to their contributions. This means karma points with a decimal value instead of an integer.

To decide how much vote weight someone gets, you could use a simple factor like 10% of total karma. But much more interesting would be: If you upvote foo's comment, you boost foo's vote weight by 5% of your vote weight. If you downvote their comment in the future, that 5% becomes negative to their vote weight. What that does is ensure that at any given instant, foo only has as much power as he should have. Time with the community only gives foo a small power bonus, since his power is mostly determined by current public perception. And it forces people to really back up what they say. If they say something horrible, they may be rendered powerless tomorrow.

If everyone's karma starts at zero, and everyone's contributing 5% of it, then how does anyone gain karma? Ahh, well.. That's where being with the community from the beginning plays a part. For each day you make a comment that is upvoted, you get one karma point which can never be taken away.

If you're a part of something small, you feel like you're a part of something special. So one way to fight noise is to stay small, or divide your community into sub-communities (almost imopssible).

But I think most important is a goal, the kind Startup News has (you better be submitting stuff related to startups). With a community goal, it's at least an order of magnitude easier to filter out the noise.


A more complicated karma system is a good idea, but it would take a lot of tinkering to get it to work. If someone figures out how to do it right, they'd have the beginnings of a reputational economy, like in Doctorow's Down and Out in the Magic Kingdom.


Perlmonks is the best example I've ever seen of maintaining a great place for smart people. It's been running non-stop for a decade. It's very open (even allowing anonymous posts) but has community rules that mostly reflect common decency. Over the years I've seen countless disrespectful newbies and trolls turned into contributing members of the community. The solution there has been to have long-time members volunteer to enforce the published rules in a consistent and fine-grained way. It requires implementing a user hierarchy, but we already have that here with the "editors". Publishing some rules and systematizing the editor controls would solve most of the problems reddit has faced.


You'll very much enjoy reading http://www.shirky.com/writings/group_enemy.html

But no, it was created primarily as a test for Arc. The fact that it's so beneficial and awesome is a wonderful side effect.


Max Goldberg, the creator of YTMND, has this to say:

"I'm straying off topic, but the point is, the users are what make YTMND. Crappy users make the site crappy, and rather than trying to delete crappy content, it would be far easier to just keep users who create crappy content off YTMND. The hard part is figuring out who to let in and who will add value to the community and the site as a whole. "


Well, sort of true:

http://ycombinator.com/announcingnews.html

Yes, because of reddit. Yes, because of Arc. But mostly because it's a good way to learn about people.


That group enemy link is really insightful. Mod parent up :)




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: