This is becoming more of a problem now. It is impossible for new stories to get as much attention now as when the site first started, but historically there really haven't been many good stories missed.

This Ted talk failed to really make the front page twice. It must have briefly been on the front page but it takes so much more weight now to actually hold that position. http://news.ycombinator.com/item?id=474875 and http://news.ycombinator.com/item?id=473821 I would encourage you to look for more examples of good stories that have not made the front page; this is the evidence that pg has said is needed to show that there is actually a problem here.

Time limiting submissions may help a little, though I haven't noticed people mass submitting middling articles. What we need is some time of cultural incentive to read the new page, and pages 2/3/4 of the new page.

> What we need is some time of cultural incentive to read the new page, and pages 2/3/4 of the new page.

Definitely. I'll admit that I never spend the time to go through the queue, let alone a few pages deep. How does one encourage this behavior? I don't care much about karma, but I guess if others do perhaps you could give extra karma for being one of the early upvoters?

PG should ask YCombinator applicants to help on the new page. If he's got people he funded helping flag spam and rewrite lame headlines, he could probably get candidates to step up a bit on this one.

If I may briefly interject, one really nice resource for this is the greasemonkey hacker news toolkit script

http://userscripts.org/scripts/show/25039

It can add the 5 newest submissions to the bottom of the front page. This is not a perfect solution, I agree it is a band aid to the larger issue of reducing junk submissions. However, it makes it easier for me to come to the front page and I definitely vote up more from the new page than I used to now (though I always made a conscious effort to go to the new page and vote there as well, to try and get articles I enjoyed onto the front page)

I did not write the script, but I do like it =)

might not be perfect -- but its really great for me - thanks :)

Stuff falling off the end of the New list before it has time to get voted onto the front page is a perennial problem. It doesn't seem any worse now than it was a year ago. Which is not surprising, because the factor driving the rate of new submissions, the number of users, is also the one driving the rate at which things on the New page get upvoted.

Are you suggesting, by extension, that if there are 300 submissions an hour (instead of per day) and 24 times more users that it will "all work out" and a set of similarly interesting things would pop out?

Quite possibly. When you submit something that's already been submitted, it counts as an upvote. So although things would fly off the New page too fast for most people to see most of them, interesting things would still get enough upvotes to get a toehold on the front page.

Actually I think the front page would be marginally more interesting at the consequence of many more stories going unnoticed. I'm also not convinced that it is possible for there to be 24 times more stories -- is there really that much relevant news out there? With 24 times more users I think we would see 5-6 times more submissions.

> is there really that much relevant news out there?

I was trying to use hyperbole to make a point (personally I think dilution's already occured). But yeah, look at digg.

Have you experimented with, instead of a queue, a randomized subset of the links posted over the last 24 h?

The random strategy may give an even chance to new submissions, regardless of what hour of the day they were posted. Would remove the ability to wait for a specific time to make a submission for the purpose of making it score high faster.

I don't know what possible negative consequences the randomized subset could have.

You would lose the sequence of submissions, which is itself quite interesting information.

My only thought is to think about limiting submission rights to a certain/higher karma threshold.

Please make it rate limiting that is relaxed as karma increases. Let someone with zero karma submit a story per 2 days or something, not never.

I think story submission should require 50-100 points or karma and/or a week or two of membership. It's a kind or rite of passage, participate in the community and you can contribute stories. Be around long enough to observe the culture.

and what about the person who has an AMAZING story but just joined?

I realize that we have these issues, but anything that's not going to make it take way too long for people to be able to submit isn't going to help the signal/noise ratio much.

It's kind of presumptuous and pompous to assume that just because someone hasn't been on HN, they don't know hacker culture through some other means. HN may be "hacker news," but plenty of hackers aren't on HN (even though it may seem as if everyone is on HN now).

And the more emphasis we put on karma, the more people will do things with the intent of enhancing their karma rather than enhancing the discussion.

It would certainly cut down on a lot of garbage. Most of the posts I flag (mostly spam links) are from users who have only been registered for a few minutes.

On the other hand, you could have long-time lurkers with no karma that are unable to submit and that could be a problem. You couldn't base it on length of time registered instead of karma either because then spammers would register accounts and sit on them before submitting garbage.

As a long time lurker, the reason I even fit in that category is because I don't submit stories but rather enjoy reading the conversation.

> you could have long-time lurkers with no karma that are unable to submit and that could be a problem

I think think that's a problem. If someone is just a lurker they aren't really fully a member of the community the way a poster is. And when has posted plenty of comments, gotten up-arrows for some and down-arrows for others, they have a better idea of what community standards are.

As a lurker, I sort of agree. That is, I don't think that I'm a real part of the community - being more the type to hang around the edges, listening to the conversation - but I think that that is a valid way of figuring out community standards; watching other people try, and seeing what works. So yes, I'm sort of biased against restricting article submission based on points and comments.

On the other hand, neither do I expect to submit anything very soon, so I suppose that not being able to wouldn't matter much to me (until I want to submit something, of course). It would be quite interesting to see statistics for who submits articles, and how many comments / points they have, and how long they have been a member.

Karma >= 2 would probably be sufficient.

It rules out people who have not interacted with the community at all, without turning HN into some exclusive "members only" club.

I think it should be at least karma >= 50

25 would be good. 50 is a long way. I'm was at 221 4 weeks back and I'm still at 232. (I don't comment on threads that I don't understand) and there are many like who are doing it this way. So keeping the posting threshold as low as 25 would be nice.

What happens if this guy heard from someone that people here give good feedback abt his new idea or startup? Should he wait for long (like 2-3 weeks to get 25 or 50 points) to get feedback? So I'm a bit biased about having posting thresholds for new threads.

I think that would help reduce the volume and increase S/N. To address the hidden-jems problem, another idea would be to randomly serve up the second page (Page 2 of the front page), giving the hidden-gems a better chance to make it to the first page.

Better moderation of the blog spammers may help as well. For instance:

http://news.ycombinator.com/submitted?id=ccraigIW has 11 submissions in the past 24 hours, none with a score > 2. That's a third of the "new" page filled up with blog posts from one source.

Karma threshold, average submission score or signal-to-noise ratio may be a good way to rate limit submissions.

An interesting idea would be to make a submission cost 1 Karma point. So people would only submit a story if they think it's reasonable that someone might upvote it.

There are two issues with that. One is it makes karma into even more of a competition:

1) Keynesian beauty contest http://en.wikipedia.org/wiki/Keynesian_beauty_contest

2) I don't think most spammers care about karma anyway, and will likely just continue to submit junk. I think well meaning newbies might be either turned off by it, or try to game it (e.g., make accounts with 3 friends, constantly vote up each others submissions so they always get 2 votes, and always have enough karma to keep submitting.)

Personally, I like the idea of rate limiting. Seems totally reasonable to me that until you have interacted with the community, you shouldn't clog up the news page. How do you get around people making 50 accounts a day to submit 50 stories a day? I don't think we have scale yet to worry about that sort of problem. At that point, we'll have to ban users, websites, ip addresses in some fashion

I think the issue right now though is not so much that it is overrun with spammers - just overrun with relatively well intentioned people who don't know what's appropriate. Case in point: I know I have submitted a couple stories that I thought were HN worthy, but that people called me on in the comments (Alabama Fat Tax, for instance). I still thought it was relevant, but I had the self restraint to only submit occasionally anyway. Someone with similar taste to me without such self restraint is the issue that I think we need to solve, and rate limiting does that job admirably (I think).

I agree with issue #1, but the other solutions (minimum karma of 100 or something like this) would have the same problem. Regarding your issue #2, I think users with negative karmas would have to be disallowed submitting in order to stop spammers.

"just overrun with relatively well intentioned people who don't know what's appropriate."

Basically my idea was that you have to prove you know about HN-culture by writing good comments, and then by gaining karma you can start submitting.

I think we are both on the same page as to what the issue is, but we disagree about the solution =)

"1) Keynesian beauty contest http://en.wikipedia.org/wiki/Keynesian_beauty_contest"

This is so true. Most of the other sites degenerated into karma wh*ring.

Another problem is that the behaviour of a dominant group attract others from the group and puts pressure on those that differ.

the thing is, any news submission site with voting already is a keynesian beauty contest

Agreed, but requiring karma be wagered on a submission I think turns it into even more of a game

A method that should be minimally visible while eliminating rapid fire spam posting:

Allow each account to have one item at a time in queue that has fewer than x points with x being a global threshold (perhaps start at two and increase if needed).

This allows anyone to post but prevents any one account from dominating the queue without the group's consent. It's still possible to game by making multiple accounts but that should be fairly easy to see in the logs if it becomes a problem.

How many of you actually Police the new section of HN? I will flag maybe a couple stories per week, anyone with a much higher rate?

Maybe the site should encourage those with higher karma to spend more time policing (the put up or shut up method).

I read the new page rather than the "front" page, because so many interesting submissions never make it to the front. Likewise, there are so many duplicates that won't make it that ought to be flagged.

I'm getting tired of bothering. I rarely, rarely get any karma for this, there is no reward, and the problem is getting worse. I'm going to stop bothering soon, and write a Bayesian-based filter to auto-detect the things I might have an interest in.

The mechanisms here are failing to find things of interest to me any more.

I've seen sites that solve some of the spam problem by requiring an account to have existed for at least 24 hours before submitting.

A lot of the proposed solutions here seem pretty complicated. The one that makes the most sense to me, while definitely not perfect, is to start by limiting new posts to accounts that have existed for a certain interval (maybe even just 30 minutes). That should filter quite a bit of spam and even though it's still easy to get around, at least it will require more effort. It's also very easy to implement.

Then you can experiment with other methods, but I doubt we'd miss many meaningful posts with this kind of limitation.

The other bucket of accounts created for spamming would need a little more sophistication, like maybe a queue limit per day. Still not too crazy.

How about a karma marketplace?

1. Upvoting costs an amount of karma points proportional to the number of points the article has at that time. Hence, it costs more to upvote an article on the home page than it costs to upvote an article in the New section.

2. If I choose to spend karma by voting on an article and it starts to rise, I get proportional "returns" on my karma.

This incentivizes me to upvote New articles I think HN will like. But then again if I lose all my karma points, I might need a bailout.

Unless this is having an impact whereby the home page is filled with bad stuff, I'm afraid there's little you can do. There are only so many slots on the front page and bringing more "new good stuff" on it will only push old good stuff out faster.

This "problem" is a reflection of the fact that there's higher competition for a limited resource (front page slots) because there are more stories to choose from, rather than a problem with the submission process, imho.

Personally, I wouldn't mind a bit of high-quality churn on the front page.

It isn't that hard to get karma...thresholds could go far in preserving the culture if set to something really high (I am thinking 1000+). I reached 500+ fairly effortlessly though and I was not always a good citizen- I have 5-10+ submissions that were killed and several of my comments have been downmodded or killed.

This is usually when some blue eyed optimists calls on all the good upstanding old timers to vote up the good stories!

Bitter realists like myself suggest it may be time for a new new new social news site.

I nice simple improvement might be to put the 5 newest stories at the bottom of the front page. A little more visibility could make a big difference.