= 27d567bc2d48d9f40e9ccc20ea370b15)
steven0461 comments on Issues, Bugs, and Requested Features - Less Wrong
You are viewing a comment permalink. View the original post to see all comments and the full post content.
= 783df68a0f980790206b9ea87794c5b6)
Loading…= b715d02e85f7b3b4ae65ca3d3e450868)
Subscribe to RSS Feed
Powered by Reddit
= f037147d6e6c911a85753b9abdedda8d)
You are viewing a comment permalink. View the original post to see all comments and the full post content.
Comments (628)
If karma is the sum of individual post scores, does that reward quantity too much relative to quality?
Not counting the free first point for every comment toward karma would be an improvement, I think.
I disagree. If someone is posting lots of comments, they're either relevant and useful or they're not. If they're relevant and useful comments, then they should be rewarded automatically. If they're not, then someone will notice them 'gaming the system' and downvote the irrelevant comments.
If they're relevant and useful comments, then they should be rewarded by getting upmodded.
Every comment/post you make is an opportunity for the community to subtract karma from you if they feel you are wasting their time.
But there is a clear bias towards not voting if the comment is neutral, as opposed to when it's really good or obviously bad. The cost of wasting anyone's time is subtle and not obvious in any single comment, hurting only in volume. We shouldn't create any incentive for additional waste of breath to fill the comment stream.
I agree. I would like some sort of "karma" score that tells me the average quality of work as well as a score indicating the total contribution.
An "average quality" type score would be more relevant in determining the expected quality of a comment before spending the time to read it.
I'd've thought so too a priori, but Hacker News tried something like this, and people started worrying that e.g. commenting on old posts would drag their average down.
I guess a really good karma system would need a better theoretical foundation, and tweaking outside of that is a very inefficient activity. Possibly one of these is already described in the academic literature. Maybe one of the qualified members of our community could contribute the effort towards finding one or research it from the start.
What if we had a score for the average of a person's best 20%? There'd still be a bit of penalty for commenting on an old thread, or posting something unpopular, but it'd be smaller.
Creating a complex system of rewards is a standard management problem, leading to unhealthy amount of attention turned towards gaming the system. In this tradeoff, a simple inaccurate system may be better than a supposedly more accurate, but complex and theoretically unsound one.