4 ms·

> an experiment we've been running to give good stories that fell through the cracks a second chance at getting attention on HN. Thanks for your efforts to imp

> an experiment we've been running to give good stories that fell through the cracks a second chance at getting attention on HN. Thanks for your efforts to improve HN. Some thoughts from a random user: I'm much more interested in having consistently high quality in 99% of the stories (and especially discussions) and don't mind at all if 1% or even 5-10% are overlooked. There's not much cost to missing a story or two; HN doesn't have to be complete record of tech news. I would greatly value a higher signal-to-noise ratio. That is HN's primary value to me now; while it's better than the competition, it's still low. A discussion forum with a high signal-to-noise ratio (informed, well thought through, rational, well-expressed, polite discussion) would win me over immediately; I'd pay $ hundreds a year to be a member.

dang 11y ago
Not just a story or two, but shockingly many good stories get overlooked by default. I wrote about this at https://news.ycombinator.com/item?id=9828818 https://news.ycombinator.com/item?id=9828818 if you're interested. > I would greatly value a higher signal-to-noise ratio. Fully agree! The question is how. It's a tough problem. Any ideas?
- hackuser 11y ago
  > Not just a story or two, but shockingly many good stories get overlooked by default I didn't realize that. Like every reader, I thought only my good stories were being overlooked! :) > Any ideas? I agree that signal-to-noise is a tough problem and I have no specific expertise. I suspect part of the solution is less democracy and freedom of speech: They are absolutely necessary in government and public life, to protect rights and ensure voices are heard, and I think we instinctively apply those ideals to online forums. But those tools don't apply well here; there are few rights to protect and much less risk to an unheard voice; nobody is going to jail, dying, or paying taxes. Also, democracy is a poor method for creating great products. The content, the linked stories and discussion, are your company's product [1]; what company designs their product/content by public vote? Does the NY Times? Do game studios? (I don't want to be Reddit or Wikipedia.) Even community-based open source projects have benevolent dictators. I suspect a good benevolent dictator (BD) editing/managing content, including front page links and discussions, could greatly improve signal-to-noise. Maybe the users and BD split votes 50/50, or the users control and the BD overrides them when it is valuable. Of course I might dislke that particular BD's judgement; as with strongmen/women in government, be careful what you wish for (though maybe the BD could reign and users vote up/down on the BD's overall performance). But I expect I'd much rather have the judgment of one or a few of the best HN members/admins than of the HN average. A couple more thoughts: 1) It would significantly differentiate HN from a million 'democratic' competitors. 2) It could be implemented experimentally: Have both a BD-edited front page (and maybe some disucssions) and the normal front page; see which draws traffic; experiment with different BD formulas and arrangements. I've talked more than enough. Thanks if you've read this far, and thanks for wrangling the community. ----- [1] I realize YC's profits come from elsewhere but for the sake of discussion let's focus on HN. I also realize YC may feel that engagement, not content, is the priority for HN. That may be your choice, but this one user is interested in signal-to-noise of content, and that will drive my engagement.
  - dang 11y ago
    What you describe isn't super far off from what HN does—it is a mixture of voting and curation. I wouldn't go so far as to call us a BD though. For sure the goal of HN is to optimize for quality content, though, not engagement per se (though engagement is important to deciding what counts as high-quality), nor raw numbers. That's because the value of the site is intellectual curiosity.

dang 11y ago

Not just a story or two, but shockingly many good stories get overlooked by default. I wrote about this at https://news.ycombinator.com/item?id=9828818 https://news.ycombinator.com/item?id=9828818 if you're interested. > I would greatly value a higher signal-to-noise ratio. Fully agree! The question is how. It's a tough problem. Any ideas?

hackuser 11y ago
> Not just a story or two, but shockingly many good stories get overlooked by default I didn't realize that. Like every reader, I thought only my good stories were being overlooked! :) > Any ideas? I agree that signal-to-noise is a tough problem and I have no specific expertise. I suspect part of the solution is less democracy and freedom of speech: They are absolutely necessary in government and public life, to protect rights and ensure voices are heard, and I think we instinctively apply those ideals to online forums. But those tools don't apply well here; there are few rights to protect and much less risk to an unheard voice; nobody is going to jail, dying, or paying taxes. Also, democracy is a poor method for creating great products. The content, the linked stories and discussion, are your company's product [1]; what company designs their product/content by public vote? Does the NY Times? Do game studios? (I don't want to be Reddit or Wikipedia.) Even community-based open source projects have benevolent dictators. I suspect a good benevolent dictator (BD) editing/managing content, including front page links and discussions, could greatly improve signal-to-noise. Maybe the users and BD split votes 50/50, or the users control and the BD overrides them when it is valuable. Of course I might dislke that particular BD's judgement; as with strongmen/women in government, be careful what you wish for (though maybe the BD could reign and users vote up/down on the BD's overall performance). But I expect I'd much rather have the judgment of one or a few of the best HN members/admins than of the HN average. A couple more thoughts: 1) It would significantly differentiate HN from a million 'democratic' competitors. 2) It could be implemented experimentally: Have both a BD-edited front page (and maybe some disucssions) and the normal front page; see which draws traffic; experiment with different BD formulas and arrangements. I've talked more than enough. Thanks if you've read this far, and thanks for wrangling the community. ----- [1] I realize YC's profits come from elsewhere but for the sake of discussion let's focus on HN. I also realize YC may feel that engagement, not content, is the priority for HN. That may be your choice, but this one user is interested in signal-to-noise of content, and that will drive my engagement.
- dang 11y ago
  What you describe isn't super far off from what HN does—it is a mixture of voting and curation. I wouldn't go so far as to call us a BD though. For sure the goal of HN is to optimize for quality content, though, not engagement per se (though engagement is important to deciding what counts as high-quality), nor raw numbers. That's because the value of the site is intellectual curiosity.