The reverse is happening precisely because there are so many new users who are voting. I’d say that the way LW started out could be used as an estimate of what that would look like. It was very rare for a comment to reach as many as 5 upvotes, and if you see an old comment that has more than that, most likely it had help from someone more recently upvoting it.
Obviously, it would not be entirely the same, and I would place more weight on the up and downvotes being more accurate if this were put into place now, but it would make it much more difficult to get to that point.
The reverse is happening precisely because there are so many new users who are voting. I’d say that the way LW started out could be used as an estimate of what that would look like. It was very rare for a comment to reach as many as 5 upvotes, and if you see an old comment that has more than that, most likely it had help from someone more recently upvoting it.
I agree LW in say 2010 seems an ok proxy for what it would be like. With one key difference, posting Main articles is much more karma rewarding than it was back then. Articles did get over 10 or 20 karma even back then
We should remember that we don’t really care how many of the lurkers become posters. Growing the number of users is not a goal in itself, though I think for some communities it becomes a lost purpose. What we actually care about is having as much high quality content that has as many readers as possible.
I would argue the median high quality comment is already made by a 1000+ user. In any case the limit is something we can easily change based on experience and isn’t something that should be set without at least first seeing a graph of karma distribution among users.
What we actually care about is having as much high quality content that has as many readers as possible.
I would argue the median high quality comment is already made by a 1000+ user. In any case the limit is something we can easily change based on experience and isn’t something that should be set without at least first seeing a graph of karma distribution among users.
There’s probably a corollary to Löb’s theorem that says a community of rationalists can’t add new members to the community and guarantee that it remains a rational community indefinitely. Karma from ratings is probably an especially poor way to indicate a judgement of rationality because it’s also used to signal interest in humor (to the point that slashdot doesn’t even grant karma for Funny moderations), eloquence, storytelling, and other non-rational things. Any karma-increasing behavior will be reinforced and gain even more karma, and the most efficient ways of obtaining karma will prosper contrary to the goal of creating high quality content. Does every user with more than 1000 karma understand that concept sufficiently to never allow a user who does not understand it to reach 1000 karma?
To be honest I didn’t fully grasp the concept until just now. I was ready to start talking economics with karma as the currency until I realized that economics can not solve the problem.
Okay, in regards to the misinterpretation:
The reverse is happening precisely because there are so many new users who are voting. I’d say that the way LW started out could be used as an estimate of what that would look like. It was very rare for a comment to reach as many as 5 upvotes, and if you see an old comment that has more than that, most likely it had help from someone more recently upvoting it.
Obviously, it would not be entirely the same, and I would place more weight on the up and downvotes being more accurate if this were put into place now, but it would make it much more difficult to get to that point.
I agree LW in say 2010 seems an ok proxy for what it would be like. With one key difference, posting Main articles is much more karma rewarding than it was back then. Articles did get over 10 or 20 karma even back then
We should remember that we don’t really care how many of the lurkers become posters. Growing the number of users is not a goal in itself, though I think for some communities it becomes a lost purpose. What we actually care about is having as much high quality content that has as many readers as possible.
I would argue the median high quality comment is already made by a 1000+ user. In any case the limit is something we can easily change based on experience and isn’t something that should be set without at least first seeing a graph of karma distribution among users.
There’s probably a corollary to Löb’s theorem that says a community of rationalists can’t add new members to the community and guarantee that it remains a rational community indefinitely. Karma from ratings is probably an especially poor way to indicate a judgement of rationality because it’s also used to signal interest in humor (to the point that slashdot doesn’t even grant karma for Funny moderations), eloquence, storytelling, and other non-rational things. Any karma-increasing behavior will be reinforced and gain even more karma, and the most efficient ways of obtaining karma will prosper contrary to the goal of creating high quality content. Does every user with more than 1000 karma understand that concept sufficiently to never allow a user who does not understand it to reach 1000 karma?
To be honest I didn’t fully grasp the concept until just now. I was ready to start talking economics with karma as the currency until I realized that economics can not solve the problem.