No it's not. Good design in social software is about what best serves the group, not the individual. (read Clay Shirky's seminal work on the issue, if you don't

No it's not. Good design in social software is about what best serves the group, not the individual. (read Clay Shirky's seminal work on the issue, if you don't believe me http://www.shirky.com/writings/group_enemy.html http://www.shirky.com/writings/group_enemy.html). Now I want you to consider a couple of issues: 1) You give the user the ability to disagree by vote. What you're doing is that you're giving people the ability to click a button instead of writing a counter-argument. How does that add to the discourse? This doesn't benefit a group now, does it? 2) You have two voting metrics. How would the ranking algorithm work? Why is populist agreement important to how the comments are arranged? Does that not make groupthink more prominent, not less? If so, why clutter the interface? 3) Is there anything wrong with a fuzzy definition of how voting is used? Sure, people have been downvoted before because other people disagreed with them, but by and large this is a rare occurrence. Oftentimes comments are downvoted and then upvoted again, when more readers stumble across the comment and think: 'say, this isn't right, this shouldn't be downvoted even though I don't agree with it'. It's happened to me before. Why does this happen sometimes, and not other times? I have a theory about that: good comments run the risk of being downvoted when a) they disagree with the prevailing groupthink on HN ... AND b) they are badly written. (badly here is defined as: unclear, or disorganized, or just plain bad). You rarely run the risk of being unfairly downvoted if you take the time to write your comment, and write it well. My experience at HN seems to support this, I'm not sure if the theory would hold true all the time, but I suspect that it does work, most of the time. I think ambiguity about how comments are used is okay, and there really isn't anything terrible or wrong with the current system. At least, nothing to justify the addition of another metric.