Someone ran the numbers on hackernews comments over four years of data: http://blog.effectcheck.com/2011/05/31/do-social-news-sites-deteriorate/

Spoiler: yes.

The most hilarious thing about this study to me was how off-the-chart negative and depressed Paul Graham was!

Alex Payne suggests that Hacker News would regain its glory if it focuses on "content relevant to tech industry entrepreneurs": http://al3x.net/2011/02/22/solving-the-hacker-news-problem.html

Perhaps this would make PG less negative and depressed, too.

See also: The monkeys produced five pages, failed to type a word, and used the keyboard as a lavatory.

The first early but temporary dip in quality of articles submitted and comments was after a mention on techcrunch. Since then it has steadily gone downhill.

Which is interesting, because in general TechCrunch has quality readers who make quality comments.

I've also noticed a smattering of analyses of the Hunters and Gabberers of Reddit recently, e.g., http://www.dailydot.com/reddit/infographic-reddit-users-top-1000/

It's still the case with Reddit, as with HN, that a handful of people are responsible for most of the links and comments. So the state of mind of those few people as more people join the system shapes the tone of the discussions within.

I don't think most of the early HN crowd would agree with your comments about TC.

PG has an interesting solution with HN classic (http://news.ycombinator.com/classic) -- shows submissions by users who have been on the site for 2+ years. All online communities should incorporate some version of this feature, where the existing users don't feel like they have been overrun by new users with different core interests. Maybe one way to build a *classic* version would be to have snapshots of users at a certain point in time when there is a large bump in traffic/influx of new users and users get to see submissions from only a subset of the community.

Sub-reddits helped some of the early core users to stick around on Reddit.

Agreed, the early HN crowd does not like the quality of comments from TC readers, which is probably why they're unhappy those readers are now on HN.

HN classic is an interesting model to keep the originals interested. Is it fair to call it a sub-Reddit of HN?

Quality is indeed a big factor, marginally useful/mostly useless link bait type articles get a lot more traction. There is an interesting experiment going on where points for comments isn't shown.

I don't think classic is used widely. Colin Percival curates a daily best of HN: http://www.daemonology.net/hn-daily/ & for popular comments/articles there is Lists: http://news.ycombinator.com/lists

Fascinating! So the set of experiments Spanish the spectrum from manual techniques such as curating quality to automated techniques such as limiting privileges of community members.

I've been reading Theory of Reddit recently (thanks Josh!) and as every attempt to un-game Reddit seems to be met with people who try to game it: http://www.reddit.com/r/TheoryOfReddit

I know the guy behind EffectCheck. Let me invite him in here.

Hey all, I'm the EffectCheck guy. :)

Hello EffectCheck guy!

Well the conclusion wasn't really yes or no, it was both. :)

In fact, the fact the change in PG's statements is much more drastic than the community as a whole. The community also seems to possibly be trending back towards more positive impacts, though we'll need to wait to see if that trend pans out.

It's also possible that the mood of the community is tied to the overall prosperity of the economy.

Those are both good hypotheses.

I was thinking that overall the HN community seems more positive now and it made me wonder why PG is so grumpy.