When I think of high quality, I tend to think of a high signal to noise ratio. This got me thinking, why isn’t karma [net upvotes / number of posts and comments]? Upvotes are relatively good measure of signal, but I don’t only care about lots of signal, I also care about an absence of noise to wade through.
The existing system—rewards writing a lot of content that is barely worth reading.
Your proposal—if you happen to write something extraordinarily good at your first attempt, such that you don’t trust yourself to achieve such high bar again on average… it incentivizes you to never post anything again.
.
I spent some time thinking about this, writing, and deleting again… I think the key problem is that for an optimal equation you should also somehow include how many people saw that comment. Like: “three people saw that comment, three people upvoted” sounds great; “hundred people saw that comment, five upvoted” sounds like mostly a waste of time. You can’t derive this from votes alone.
I think that acts as a slight incentive against demon threads, which seems good to me. I do think it would be valuable if you could easily get the ratio of people upvoting to people seeing a comment and not upvoting. But you are also not counting against it the ratio of people downvoting to people seeing a comment and not downvoting, so maybe it balances out?
Robust communication requires feedback. Knowing you received all the packets of information, and checking whether what you received matches what they sent.
Building ideas vs breaking ideas. Related to Babble and Prune, but for communities. Shortform seems like a good place for ideas to develop, or babble. For ideas to be built together, before you critique things. You can destroy a half built idea, even if it’s a good idea.
If they don’t tell you how to hold them accountable, its a Chaotic intention, not a Lawful commitment
When I think of high quality, I tend to think of a high signal to noise ratio. This got me thinking, why isn’t karma [net upvotes / number of posts and comments]? Upvotes are relatively good measure of signal, but I don’t only care about lots of signal, I also care about an absence of noise to wade through.
Thoughts?
The existing system—rewards writing a lot of content that is barely worth reading.
Your proposal—if you happen to write something extraordinarily good at your first attempt, such that you don’t trust yourself to achieve such high bar again on average… it incentivizes you to never post anything again.
.
I spent some time thinking about this, writing, and deleting again… I think the key problem is that for an optimal equation you should also somehow include how many people saw that comment. Like: “three people saw that comment, three people upvoted” sounds great; “hundred people saw that comment, five upvoted” sounds like mostly a waste of time. You can’t derive this from votes alone.
Seems like that’s easily fixable by having the numerator be most upvotes on a single post/comment, or your total upvotes, whichever is higher.
I think that acts as a slight incentive against demon threads, which seems good to me.
I do think it would be valuable if you could easily get the ratio of people upvoting to people seeing a comment and not upvoting. But you are also not counting against it the ratio of people downvoting to people seeing a comment and not downvoting, so maybe it balances out?
Robust communication requires feedback. Knowing you received all the packets of information, and checking whether what you received matches what they sent.
Building ideas vs breaking ideas. Related to Babble and Prune, but for communities. Shortform seems like a good place for ideas to develop, or babble. For ideas to be built together, before you critique things. You can destroy a half built idea, even if it’s a good idea.