4 ms·
I would love to see a kernel density plot of comments by point value, hypothesis being that there are a lot more low-point submissions than there were before. I
by andreyf 17y ago
I would love to see a kernel density plot of comments by point value, hypothesis being that there are a lot more low-point submissions than there were before. I see a lot of un-thought-out off-the-cuff comments lately: knee-jerk anti-Chinese sentiments, top comments with obvious flaws in reasoning, speculation, or just plain immaturity.
I know people have been saying this for years, but this is the first time I'll say it: I've been avoiding reading news.YC because I don't feel like I learn anything from the comments anymore. I'd really welcome a return of PG's orange-names experiment, but with a gradient instead of a binary grouping (to avoid tribal sentiment).
- jacquesm 17y agoAverage karma is a meaningless metric.
- andreyf 17y agoMeaningless how? In some absolute sense? It's a metric that, like any public metric, will guide behavior. The question to ask is "how will it guide behavior?" The most quick/intuitive explanation is probably the best explanation - I think it'll make people think twice "is this more insightful than my average comment?" before hitting Submit. As the community grows, this is exactly what we need: incentives for commenting less, and reading more.
- jacquesm 17y agoAverage karma is meaningless because there are many modes of use of HN that are very useful and that will inherently lead to a very low average karma.
- andreyf 17y agothere are many modes of use of HN that are very useful and that will inherently lead to a very low average karma Such as? Also, I'm claiming that as a community grows, it's vital for people to post less often - preferably only on topics regarding which they have something truly insightful to share. Your counter-claim would need to be stronger: you need to show that the "modes of use" you speak of are vital to the community, that there is harm in disincentivising them.