The ratio of meta comments to object comments on this post is about twenty to one right now. That’s a pretty good ratio but I think if we really tried we could push it up to 40:1.
It’s an impractical idea for a thread. People may occasionally bookmark good articles, but (generalizing from my own example) they rarely bookmark good comments. So I’m not surprised that this thread was something of a failure.
Don’t post more than one comment in this thread. If you find more other-person comments you want to add, edit yours to include them.
I don’t understand the rationale for this. It seems contrary to the usual practice in threads of this type, where items are separated in order to allow voting on them individually.
I don’t understand the rationale for this. It seems contrary to the usual practice in threads of this type, where items are separated in order to allow voting on them individually.
I support your reasoning. There is no good reason to group by recommenter.
Mind you there is only one non-meta comment in the thread. It seems to have flopped for some reason.
New terms: hcomment for comments in this thread; o-comment for comments suggested by this thread.
What I figure is that if you like the o-comments, you can vote on them instead of the hcomments. I don’t know if this is a good idea, since it multiplies the effect of selection bias, but I figure comments don’t need two sets of karma scores. Talk about information cascades.
On the contrary, it seems to me that information cascades would be better avoided by separating out voting on the comments “where they occur” from voting on them as “nominees” in this thread.
I had assumed the purpose of this thread was to identify these comments so that they could be collected elsewhere (perhaps offsite) in a “List of Best LW Comments”, ordered by how they score in this thread.
I prefer Grognor’s way. Voting on comments in this thread should be based on the value added by comments in this thread. Basing voting on the value added by the original comments would just create an incentive for people to mass-repost the top comments feed.
Good idea for a thread. I have often thought that much important, insightful content is locked away in a deep comment threads that many LWers never see. In general, I think the current hierarchy is suboptimal for exposure of important content. Furthermore, past content should more often be revisited, updated and revised.
META
Discuss this thread in replies to this comment.
The ratio of meta comments to object comments on this post is about twenty to one right now. That’s a pretty good ratio but I think if we really tried we could push it up to 40:1.
It’s an impractical idea for a thread. People may occasionally bookmark good articles, but (generalizing from my own example) they rarely bookmark good comments. So I’m not surprised that this thread was something of a failure.
So far. Once people know there’s a place to put them, they could trickle in.
You’re right. Just in case a few others aren’t willing to play along, though, I hereby declare that non-meta comments on this thread are banned.
I don’t understand the rationale for this. It seems contrary to the usual practice in threads of this type, where items are separated in order to allow voting on them individually.
I support your reasoning. There is no good reason to group by recommenter.
Mind you there is only one non-meta comment in the thread. It seems to have flopped for some reason.
New terms: hcomment for comments in this thread; o-comment for comments suggested by this thread.
What I figure is that if you like the o-comments, you can vote on them instead of the hcomments. I don’t know if this is a good idea, since it multiplies the effect of selection bias, but I figure comments don’t need two sets of karma scores. Talk about information cascades.
On the contrary, it seems to me that information cascades would be better avoided by separating out voting on the comments “where they occur” from voting on them as “nominees” in this thread.
I had assumed the purpose of this thread was to identify these comments so that they could be collected elsewhere (perhaps offsite) in a “List of Best LW Comments”, ordered by how they score in this thread.
I prefer Grognor’s way. Voting on comments in this thread should be based on the value added by comments in this thread. Basing voting on the value added by the original comments would just create an incentive for people to mass-repost the top comments feed.
Would be useful if people quoted or summarised each comment so you can understand the context of comments on it without clicking through.
Bah, that rules out my first 5 candidates.
It also rules out how great this comment is (or could be, anyways).
Good idea for a thread. I have often thought that much important, insightful content is locked away in a deep comment threads that many LWers never see. In general, I think the current hierarchy is suboptimal for exposure of important content. Furthermore, past content should more often be revisited, updated and revised.
Look what I found!
(I still edit my non-meta post in this thread from time to time. Also, I still wish more than three people used this thread for its intended purpose.)