A much more accurate portrayal is everyone second guessing whether or not their comment will please the HN group-think. People don't filter themselves when they think their idea is stupid, they filter themselves when they think their idea will be unpopular.
I think in practice that 'popular comment = good comment' works well enough that it seems to beat all competing systems for internet commenting so there must be some kind of truth in the statement. Maybe it has a low enough false positive rate to make a small false negative rate acceptable?
But perhaps this is the problem of moderating comments on a binary scale of "good" and "bad".
A comment judgement/rating system that is more nuanced will not only help the readers to choose from but also help the commenting user to understand why his comments are being downvoted. Examples of nuanced downvote reasons may include:
Possibly ... 1. Non-sequitur 2. Racist/Bigotted/Sexist 3. Gratuitous Profanity ... etc
If we really want to get clinical about this, we can simply use this list: http://en.wikipedia.org/wiki/List_of_fallacies
This of course is only for the downvote.
I personally would be interested in seeing whether, over time, my comments (and therefore my thinking) have been victim to certain patterns of thinking that I haven't noticed in myself.
> If we really want to get clinical about this, we can simply use this list: http://en.wikipedia.org/wiki/List_of_fallacies
That struck me as somewhat non-parallel. Racism, sexism, bigotry, and profanity aren't fallacies.
I don't worry so much about being up voted or down voted (but I dig getting karma points as I know I'm giving someone else some value). I do however care about end up in a disagreement with someone else that turns into a superiority match. I personally don't want to be that person.