The same thing happens to every subreddit. It's good, people like it, it grows very large, the quality of submissions goes down (or the same content gets reposted constantly). People are incentivized by their visible "karma" score to post mass amounts of low-effort content.

I feel like Hacker News is one of the only online communities I've participated in where the score is handled appropriately. It's a good personal measure of how well my posts and comments were received by the community, but it's not bragging rights. No one else can see my vote tallies here, so I have no real incentive to increase it.

I've seen this affect sites like Stack Overflow as well. At the bottom of any popular question are usually very low-effort answers from posters who really shouldn't have answered at all. I wonder if it's an example of their otherwise well-executed reputation system creating noise that might have been avoided.

Out of curiosity, does any one know of any research or articles into reputation systems in general?

Randy Farmer's "Building Web Reputation Systems" is a good start:

http://buildingreputation.com

There's a blog, book, and wiki.

Some keyword searches: http://www.worldcat.org/search?q=kw%3A+online+reputation+sys...

https://scholar.google.com/scholar?q=online%20reputation%20s...

This can be solved by moderation. Not voting, old school approach of moderators removing things. https://www.reddit.com/r/AskHistorians/ still has quite a high standard, and that's because they have strict moderation.

You can't have "democracy" and "voting" and "algorithms" and then have high content.

Depending on the subreddit, it's sometimes interesting. but I haven't found a NNTP-level client for it, yet.

It does support RSS quite pervasively. Tag ".RSS" to the end of any URL (before the arguments), and you'll get a feed.

Subreddits, users, posts, searches, and a number of other elements.

A quick search finds at least two GNUS backends for reddit.

Same problem as Twitter/Facebook: little to no barrier to entry, endless bots and politically motivated low-value posters.