Hacker Newsnew | past | comments | ask | show | jobs | submitlogin
Effects of TechCrunch article (with News.YC traffic graph) (ycombinator.com)
50 points by pg on March 25, 2008 | hide | past | favorite | 29 comments


The worrying thing is that when things start to go bad they go bad fast because the system is self-reinforcing in two different ways:

1) When people see lame stories make the front page, they see that they can submit lame stories and have them make the front page.

2) When you submit a good story and it doesn't make the front page then it's kind of depressing. And what kind of people still bother to submit content when that happens? Sucky people.


There have always been lots of bad submissions. The real problem is voting, not submission. People aren't so much submitting worse stuff as upvoting worse stuff.

And the danger of that is worse than just encouraging bad submitters or discouraging good ones. It makes the site seem to be about dumber stuff. Seem because it is now about dumber stuff.


Can you put up a vote weighted page on an alternate URL, if the vote weighting implementation is ready to be turned on? In the past you've said that with vote weighting on the front page looks mostly the same, I think many of us are curious if that is still true.


So now you're finally admitting there's an effect-- any plans for fixing it?


Though we've never had an influx like this before, we're optimistic the same thing will happen this time.

That's a no.


"And what kind of people still bother to submit content when that happens?"

Entrepreneurs.

This forum is a metaphor for a startup itself. You have good days and you have other days. So what do you do? You KEEP ON SUBMITTING. You perservere.

Personally, I have reached "Hacker News Addiction, Phase II". Every time I read something somewhere else, I want to know what you guys think. So I submit it. And usually, no one comments. But I also understand that I don't really have much control over this, so I just keep submitting. And keep voting. And you should, too. Don't let the "lamemakers" take over your site. Just keep doing your thing and they'll be moving along soon enough.


I strongly agree with #2. I usually stick to contributing comments because I don't want some of my favorite stories I've come across (which I have not yet submitted) to get buried (and unable to be submitted later because of duplicates).


Did any stories about the end of the 700mhz spectrum auction make it up? It's not a good sign that something that big gets missed over by the community.


http://news.ycombinator.com/item?id=143955

Yeah, I submitted that a couple days ago and no one voted that up.

I think that if you come to depend on hacker news for your news, then you only visit the front page, and stop visiting the new pages. And given that you can upvote on the front page, people just do it there, rather than peruse through the new articles and upvoting them. So what you end up with is that an article just needs 2 or 3 votes to make it on the front page, and then there's a positive feedback effect. This makes it more of a hit or miss sort of thing for articles. I think if you can only vote when you're on the new scenes page, it might help.


Then submit it again. Maybe the people that would have been interested in that story were too busy working to visit the "new" page during the 4 hours it was there.


I am one of the new users and I'd like to comment on the average tone in comment threads is slightly less polite remark.

I am not entirely sure if polite meant not rude or if it meant a pat-in-the-back encouraging. However even a casual glance shows a lot of modded up comments that are of a latter kind. Especially in the Ask YC/HN articles. While reasonable encouraging is good, there's also a plenty of great job and looking good comments for pre-launch projects that are a plain junk .. with an emphasis on pre-launch. In my humble opinion these are extremely damaging comments regardless of how polite they are.

Perhaps this is a part of the site social culture. Or perhaps I'm misreading the meaning of "less polite". I just hope I'm wrong on both counts.

Now, mod me down if you agree :-p


I think he meant not rude. The idea that you should only say something here that you would be willing to say to someone face to face.


"the average tone in comment threads is slightly less polite "

This is a hard statement for PG to verify-- right on the heels of the TC spike was all of the crap that popped up about his recent essay.

I'm going to reserve judgment for a bit.


I have noticed there are much less programming/hacking stories recently and tried to upvote the few remaining even if they are not particularly interesting to me. If this site becomes exclusively about startups, I don't think it would be interesting to me.

Also much more stories where what is said is supposed to be important mostly because of who said it. I believe popularity contests are inherently against core hacker values.


You guys seem to forget the fact that YC applications are due soon. Perhaps some of the new influx is users trying to cash some quick karma?


I had exactly 1 karma point when we got accepted into YC. :)


it's good to know that some of the somewhat/to outright discourteous comment threads aren't the norm...


"There has been some dilution as a result. The stories that get voted up are not quite as good, and the average tone in comment threads is slightly less polite."

The problem is voting. As you might guess, I think a more meritocratic approach to social news (e.g., intelligence testing through puzzles) would give you better results when you have more users.


I'm not so sure. That's not so much meritocratic as a test of someone's determination or the amount of time they have to waste.

True meritocracy in online voting situations would be something like giving more importance to votes from people with higher karma. So, the people who get voted up more, then get more power in voting situations themselves. That way, the site becomes whatever the "top users" want, a true meritocracy.

I'm also a fan of giving precedence depending on the "age" of the user. That is, how long they've been a member of the site. This means the site may not evolve as quickly, but it appears that's something people want to avoid anyway. It tends to be the new "influx" of users that comes after popularity that drags a site down, so letting older users have more say could make sense.


"I'm not so sure. That's not so much meritocratic as a test of someone's determination or the amount of time they have to waste."

That depends on the nature of the puzzle. Not many people can solve Rubik's Cube no matter how determined they may be.

"True meritocracy in online voting situations would be something like giving more importance to votes from people with higher karma. So, the people who get voted up more, then get more power in voting situations themselves. That way, the site becomes whatever the "top users" want, a true meritocracy."

That may not lead to a meritocracy. I think intelligence testing is more reliable.


A good argument, but do you think a user's raw intelligence could correlate to the quality of their links and link preferences?

I don't. The people who post all the political links to Reddit seem quite intelligent, as do many of the people starting pointless arguments. The ability to troll and incite flame wars might even go UP with intelligence.

Perhaps if the puzzles you proposed measured "emotional intelligence" or domain specific knowledge we'd stand a chance, but raw intelligence doesn't prove much. Many of the nicest, most "savvy" people I know are not of high intelligence in an IQ sense, but are still very successful in their fields, and vice versa.


Yes, you can have different sorts of puzzles to measure different sorts of intelligence. It would make a fascinating experiment to see the impact of the puzzle type on link and discussion quality.

In fact, my next puzzle-based social news site will involve a word-based puzzle, which may reward users more who have high "verbal intelligence".


Just a guess, but I think domain specific knowledge testing would work best. I know enough grouchy "intelligent" people to go for the raw intelligence option, but if people were tested on their general knowledge of the sector that the site is based on, that's going to result in, hopefully, more informed discussion and links.

Indeed, perhaps this is why sites like Reddit and Hacker News were / are such high quality sites at the start of their lives, because only the most informed people knew about them. Once the less informed get involved, it all goes to pot. So testing how "informed" someone is might, effectively, resolve it all.. but I can't help but feel a lot of people would be turned off by such "testing"!


So testing how "informed" someone is might, effectively, resolve it all.. but I can't help but feel a lot of people would be turned off by such "testing"!

So why do people visit this site?

http://mightyquiz.com/

Imagine turning MightyQuiz into a social news site.


Puzzle based social news only appeals to people that like solving puzzles. That's fine and I'm glad your site exists, just don't expect your model to supplant vote based story ranking.


>That depends on the nature of the puzzle. Not many people can solve Rubik's Cube no matter how determined they may be.

Sure they can. It's algorithmic. Anyone can be taught the algorithm. Learning to solve it without any help would be different.


I mean discovering an algorithm to solve it on your own is very hard.


It's possible (probable, even) that every one here has read Clay Shirky's "A Group Is Its Own Worst Enemy", but here 'tis for posterity: http://www.shirky.com/writings/group_enemy.html


sorry but intelligence testing alone isn't meritocratic. perhaps when combined with other metrics it could be a factor, but far from the only one.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: