I guess i'll hijack this thread to ask anybody else if they have noticed a slip in quality of front page? I used to check HN first thing and read 50%+ of frontpage stories, now I barely click one or two.
When there is something interesting, I've often read it elsewhere, or its an Ars re-hash of a story from yesterday. When I submit stories I find interesting and HN-worthy, I find them at 3-4 points having been submitted 12 hours ago.
My perception is that the HN frontpage is being gamed more by social marketing teams, journalists who work for re-spamming blogs and growth hackers. The types of stories are more soft news, marketing from co's etc. than much more hard-tech.
As an aside, right now the homepage is unusually good - must be because it's a weekend and the social media hackers are at home.
It's not as good as it has been in the (far) past, but it's better than it was at its nadir, when the front page was saturated by stories from a few obviously baiting sites on a very few topics.
One thing that does seem to have gotten persistently worse is the process of getting a good story onto the front page. Mediocre stories with immediate or simple appeal do seem to be doing a better job of outcompeting Rilke than they ever have in the past.
It's a good site for halfway-decent posts about Golang or CSS compilers. Thankfully, it's become less hospitable for nakedly ambitious Bitcoin cheerleading or comically inept copyright law analyses. It would be great if the site could come up with some way to let the kinds of posts 'gruseom's submission log models get some space on the front page, but it would be bad if that attempt reverted us back to the worst of the Venturebeat and Techdirt days.
These are stories I quickly recalled from the last ~week that I enjoyed, all just happen to not register on HN.
If I go through my queue on pocket i'm certain there would be many more examples. I'd estimate less than 10% of what I now save on pocket for reading is sources from HN.
My take is that there was an influx of non-tech people wanting to find out what tech people were reading about. If you subscribe to the usual tech blogs you'll find (or would have found) that a lot of their stories are/were sources from here or are reblogs[0]. This mass of new users has now affected what is selected and how.
[0] Example, Notch tweets that they are canceling their Oculus deal. His tweet gets 8 points:
Ok, now you have me worried. This is a problem. We want the best stories on the front page as well as the best version of each story. We certainly don't want knockoffs.
One thing you can do when you know of a better version of a story is post its URL in the thread along with a clear explanation of why it's better. We see most of those, and often use them. (Edit: and maybe we should have a textually unique string, à la hashtags, that people can insert in such comments. Then I can write code to make sure I see them all.)
That, however, will do nothing to help HN get stories that it's missing completely, since there won't be any such thread to comment in. You seem to be saying we're missing most of those. How can we fix that?
That, however, will do nothing to help HN get stories that it's missing completely, since there won't be any such thread to comment in. You seem to be saying we're missing most of those. How can we fix that?
One idea is to change the rules so that users are allowed to resubmit stories that received fewer than four upvotes after 24 hours. Right now you risk getting banned for doing that.
It might be a bad idea since the resubmissions will push more legitimate stories out of the new submission queue more quickly, though.
Another idea is to increase the number of stories displayed on https://news.ycombinator.com/newest from 30 to 120. At this point, if your submission drops off the first page of /newest then it's pretty hopeless that it'll be seen by anyone. Showing 120 stories at once will give good content more time to get onto the front page where more people will see it.
> Another idea is to increase the number of stories displayed on https://news.ycombinator.com/newest from 30 to 120. At this point, if your submission drops off the first page of /newest then it's pretty hopeless that it'll be seen by anyone. Showing 120 stories at once will give good content more time to get onto the front page where more people will see it.
Or: show a random 30-subset of the newest 120 stories on /newest, and refresh this 30-subset per view/per time interval. Set this up to get the best of both options: at any point in time the newest 120 stories would be on /newest for someone, and there would only be 30 stories to look at for any one person. This will also give people more incentive to go check /newest once in a while, since they would, more often than now, find many more as-yet-unseen stories there each time.
Yeah. I often encounter really amazing stories, then go to post them on Hacker News only to find that they were submitted a few days ago and got little attention. This is an example:
Part of the problem might be all these sources are widely read. I don't read through hacker news, what I can read anywhere else (or have already read). If you read business week/nytimes/bbc news anyway, your not going to read them on hacker news or upvote them. I'm much more interested in lesser know sources and well written commentry.
I really think that ultimately not all users deserve to be treated equally. Much as I tend to disagree with many of your viewpoints in comments, your own submission list is impressive and would mostly merit front-page posting.
I've played with various karma systems in the past, and it's a really hard thing to get right. In particular, I believe systems need to be responsive to recent activity, while reflective of legacy (this also tends to allow newcomers to rise the karma heap fairly quickly, if deserving). Tuned sufficiently well, this needn't breed apparent favoritism (and contempt).
Has there ever been a point when Rilke would make the front page? The only comparable moment I remember was when Tarkovsky made the front page for a bit, but that was an extraordinary aberration.
You did. As a fellow Dostoevsky fan, that was one of the things that made me really like this site. I remember I even scoured your account for other Dostoevsky posts that didn't make it.
> Mediocre stories with immediate or simple appeal do seem to be doing a better job
Do you have some criteria by which you classify stories as mediocre other than your own personal preferences? If there was a decision procedure for this, we could just implement it and have it scrape the web for objectively great stories. Lacking that, we can build a site that aggregates the opinions of its users.
I did not have the impression that it has gotten worse recently, but it's hard to know. I take your opinion seriously, though. I'd appreciate your help in identifying posts that are "[gaming] by social marketing teams, journalists who work for re-spamming blogs, and growth hackers". It's my job, and my fervent desire, to make the front page inhospitable to such dreck.
HN has been in an arms race for a long time against artificially promoted content, and we (mostly PG) have developed a lot of defenses against it over the years--a large part of why HN is still standing. I've written software to combat such gaming and I intend to write more. One thing I'm pretty sure of, though, is that the problem of bad stories is easier than the problem of bad comments, and it is the latter that poses the greater risk to HN just now.
Incidentally, this past week was a good one for the HN front page. These two threads, in particular, were IMO as good as any in recent memory:
Thanks - I will send along any examples I find. A suggestion might be to have another way to signal that a story is being artificially boosted besides 'flag'. For instance, a 'report' button where a brief description could also be attached (ex. 'This person posted this link to Facebook and asked their friends to upvote it').
The second part is good stories not ranking (see comment above), which doesn't have a straightforward solution, although it might sort itself out with softer, re-blog, and promoted stories being moved out of the way.
My comment may be lost into the aether. But if anyone is out there, I think Hacker News could really use a story merge feature. That would significantly reduce noise around "major" events, and likely prevent second order noise as people try to piggy back on with tangentially related "analysis".
It drives me crazy that "growth hackers" have become what they are now. My job could be pretty accurately described as growth hacking, but I'd never try Reddit/HN gaming as a growth strategy. To me, good growth hacks (the ones I would go after, at least) are similar to PayPal's method of letting you send people $10, provided they sign up for PayPal. Something that has the potential to grow virally, but also simply demonstrates why the product is useful to the people encountering it.
My current job doesn't have HN-style content. If it did, I'd just submit it once, and leave it alone.
Actually we're pretty sure there is less of that now than a couple years ago, because Daniel has written some amazingly effective software for detecting it.
There are some sources which are used that perhaps shouldn't be. Sadly lots get posted from those sources and people up ote them so there's not much to be done.
As always, visit /new and upvote the great stuff and flag the flagable stuff.
If you notice weird vote patterns a short email with links usually gets results.
Sites like HN can't escape Parkinson's law of triviality: Complex or mixed-conclusion content will naturally get drown out by easily "understood" (at least in the perception of the reader) and debated material.
Titles play a large part as well, as they allow people to support or reject a notion without the effort of even following the link.
This isn't meant to be grumpy or conspiratorial, nor is it negative about HN, it's just the way these things often work. HN does not measure or demand that you follow a link to click the arrow (nor does it require it for "flag", which is a used in practice to downvote). It doesn't maintain a history of recently visited items to ease voting on content you may have read earlier in the day. These naturally lend themselves to cheap votes.
HN has always been rather intensively curated. You may be right about the statistical trend, but conscious intervention can reverse this "entropy" somewhat in our limited context. The question is how we can do this better.
Fascinatingly, the number of stories posted to the site has not been increasing:
That might be a good thing. The number of substantive HN-appropriate stories in the world is not growing by leaps and bounds; neither, therefore, should our story stream. We want quality, not quantity.
There's a lot we can do about story quality, because the front page is so limited, there are many fewer stories than comments, and because we manage the front page for quality in ways we could never manage the threads. However, there's no way we can scan the universe of stories to find all the best ones. We rely on /newest to be our universe of stories. If /newest is broken, we'll fix it. But we're going to focus on the problem of bad comments first.
I'm happy to see these are the priorities. To emphasize comments > submissions, my HN app for android opens all articles directly to the comments. it takes 2 clicks to see the submitted link.
When there is something interesting, I've often read it elsewhere, or its an Ars re-hash of a story from yesterday. When I submit stories I find interesting and HN-worthy, I find them at 3-4 points having been submitted 12 hours ago.
My perception is that the HN frontpage is being gamed more by social marketing teams, journalists who work for re-spamming blogs and growth hackers. The types of stories are more soft news, marketing from co's etc. than much more hard-tech.
As an aside, right now the homepage is unusually good - must be because it's a weekend and the social media hackers are at home.