"Rats, top comments will be impossible to beat. I can probably piggy-back off a top comment, though, those comment threads aren't long yet..."
actually, this text in the reply box highlights a huge problem with threaded (and voted) discussions in general: there is little incentive to reply to any already-huge thread, even with valuable content.
i've been thinking of how to solve this issue for the past week and have some ideas, working on a prototype.
I'm curious what you came up with.
I used to work on Google Moderator, which was a project created to solve this problem for non-threaded discussion (which is even worse), and we had some neat solutions. We displayed comments to users to be voted on, and ended up using a variant of wilson confidence intervals not just to rank, but to figure out which comments we should display to users to be voted on.
where i ended up was, it cannot be solved within a single view. there needs to be some form of split side-by-side view for the discussion, each prioritizing replies differently - by freshness, votes, weight or other metrics. another idea was to create a heat-map based on reply velocity to give a third dimension to the whole thing.
replies need "simmer" time at the top to aggregate enough votes. if there is one huge reply thread that dominates, few will scroll through pages looking for new content.
just the ability to fold comments on HN would help a lot already.
Why not push really long threads to the bottom? People will find the long discussions by scrolling to them. Indeed, they will know that they are there, just like now we know that the long discussions are at the top. The effect though is that now readers have to get past other discussions first. Basically, the order should be short-good, long-good, all-crap.
Also, while this may be a problem here, HN still has the best comment sorting system I've seen (but not measured). Comments have time to stay at the top and simmer down and if they are good they gain enough momentum. The only time this seems to be come a problem is when you get to a really popular thread after hours/days of discussion. Then you cannot get a word in edgewise.
An alternative solution could be that for every comment you make, you give up a bit of karma and you gain it back through comment upvotes (but not article postings). That way people will watch what they say. This may give some positive feedback to posters that harvest huge amounts of karma, but it will also prevent bad comments. On top of that, you could make your post stick up top longer by paying a blood price: for every 50 points of karma you get another minute guaranteed at the top of your discussion thread, or some such. Or just free market FTW: the person that bids the highest amount wins the top spot for that many minutes.
Another site that I was on had the ability to fold up everything that you'd already read, and only show you the new content. It made keeping up with active discussions utterly trivial. Of course a first time user still sees everything.
I generally agree, however, your comment would be more valuable if you would include verifiable information instead of anecdotes and personal opinion, and you would actually analyse the problem instead of throwing around vague criticism.
My take on the subject is:
This effect can be mostly attributed to the fact that a top comment means more exposure, and given that the upvotes outweight the downvotes on the response (which is true, since that's why it is at the top), by deductive reasoning we arrive to the obvious conclusion that more exposure will further cement the position of a given post.
Also we must not forget about the effect of peer pressure[1]
I disagree with the author. This isn't exactly a parody thread. A parody is "an imitative work created to mock, comment on or trivialise an original work" [1]. For one thing, I would hardly call Hacker News comments "original work", and a simple creation like this hardly captures the full scope and breadth of comments on here.
Furthermore, I'm not really sure why this belongs on HN, because it's not very technical, and frankly, not very intellectually stimulating. People on here don't appreciate humor, so those who upvoted this should have known better. I've flagged the article.
I want to laugh, but the grammarian in me is cringing too much. If you don't understand the cause of my cringing, please read the Their/They're/There section of http://theoatmeal.com/comics/misspelling
EDIT: saraid216's response was funny, without being cringe inducing
I'm sorry, you seem to believe you can just throw around these opinions of yours without evidence. I demand you provide me with multiple peer reviewed scientific papers from journals of note to backup your so called claims.
Frankly there's no point in even continuing this conversation until you do and we should just assume that I'm right and you're wrong.
I don't understand why you're making this so complicated. This problem can be easily solved using unix pipes. There's no need to complicate your technology stack with extraneous definitions of "parody".
There's surprising amount of hidden jokes in there too - feel free to read the OP's blog post [1] for many of the details (assuming you like spoilers)!
As in every HN thread where hellbans are mentioned, someone has to be the person who has to explain that you can see hellbanned posts by enabling showdead in your preferences. I guess that's me this time.
Good god this was funny, right down to the usernames.
But now its time to be very HN about it.
> Rats, top comments will be impossible to beat.
Actually I think I have a solution to this, though its just guesswork.
I'm not particularly well known here (or anywhere), but either people actually like the words I say, or I simply have yet to make a complete ass out of myself. (I think one of these is more plausible, but who is to say.)
My average karma on HN is 20.59, which seems to be a lot. Specifically, on the Leaderboard[1] that puts me in fourth place among the HN big names for average karma (though I have nowhere near the total karma to appear there).
I've noticed that when I reply to a post, even if there's already 30 comments, my post usually ends up at the top. And it stays there, even if no one replies, and even if I don't get many or any upvotes for it, for a few hours sometimes.
Alas I don't know for certain, but my guess is that if you have a high average karma then your posts are automatically weighted higher, so you can inject your opinion into almost any topic at any time. This affords me the luxury of being a contradictory whine even if I come late to the party!
> Alas I don't know for certain, but my guess is that if you have a high average karma then your posts are automatically weighted higher, so you can inject your opinion into almost any topic at any time. This affords me the luxury of being a contradictory whine even if I come late to the party!
One thing I dislike about the 'average karma' is that it discourages engaging in an extended discussion. For example, if you reply to this post, you may get a number of upvotes, but probably not as many as your original post. This creates an incentive just to ignore replies (when possible) so as not to bring down your average comment score.
Well, I've been banned on this acccount for long periods of time for one. I don't know if people keep written accounts of being banned for stupid reasons.
Well, random sorting might be a bit of an annoyance due to HN's other annoyance, the "Unknown or expired link" problem, which seems to crop up more often if I do actual work and then pop over to HN when I have to wait for tests/compilation/etc. Locating the comment you were replying to with random or date-based sorting (so that you can click "reply" again, paste your previous reply, and then resubmit) would be a serious pain in combination with that.
But you're correct in that the current algorithm (or any sufficiently transparent algorithm) does lend itself to gaming. It might be harder to game if you didn't mind penalizing comments based on the quality[1] of replies they generate, then sorting based on the thread's average quality would mean that there's less incentive to post in the top thread just to get exposure, since your new, karma-less post will drag down its average and subsequently could drag down the thread. That could get prohibitively expensive to calculate for huge threads, but would be less prone to gaming, I think.
[1] If we can hand-wave "karma" to mean "quality"; I get way more upvotes when I am flippant but technically correct than I do when I try to be thoughtful.
Well, no, right now average score is meaningless. But if having a higher average score means that you have a stronger voice, as OP suggested, then people will start to care about those numbers.
I assume the goal is to encourage high-quality discussions (including extended discussions/back-and-forth), in which case average comment score will work against that.
I tend to think that better discussion in a forum like this comes about with less back and forth. That's not to say that alternating comments on opposite sides of a viewpoint can't be enlightening, but I think it's stronger if those viewpoints come from multiple people on both sides.
The metric which would be a tad closer to ideal would be some average of u/t where u := the number of users who upvoted a given comment, t := total number of users who saw that comment. Where "saw" can mean either "the comment was already there when the user loaded the comments page", or, if you want to get fancy, "the comment was within the browser's viewport long enough to be read" (which some clever JavaScript can tell you.)
You are also forgetting your groupies. People will upvote your post because its you and not due to its content. tptacek has mentioned this in the past, and I personally suffer from it.
Well, pg isn't on that leaderboard. Not sure what karma means or what it's worth other to the user than giving one the ability to down-vote (which should require posting a reason for having done so, but that's a different thread).
As a tool or "quality score" that aims to control and keep a forum from derailing this kind of a scoring system has probably done OK. Not sure if it is solely responsible for these effects on sites like HN and SO, but there's not denying that the signal to noise ratio is far better than for example USENET S/N or some forums out there (vBulletin or phpBB type).
Having operated a forum before I can tell you that it can be an absolute nightmare (I was using vBulletin) on many fronts. HN seems to be able to maintain decent quality.
Yeah, he was definitely at the top before, and now he doesn't appear at all. I'm pretty sure you're right, and he coded in a rule to remove himself from the list.
Piggybacking off a piggyback comment to take the discussion back on topic:
There is further parody content in the URLs of the 'reply' links
Didn't realize until stalking through to the author's commentary on his personal blog: http://bradconte.com/hacker-news-parody-thread.html, because the relevant replies (including the author's own: ctrl+f "B-Con") were too far down this thread to ever see without foreknowledge.
I've certainly noticed it's become easier to get replies as my average karma rises. It's also easier to get posts modded up, creating a positive feedback loop (I'm guessing more people vote on posts that are near the top of the page?). Don't have any concrete evidence though.
The blog's text looks fine for me on a 95DPI monitor using Firefox on Windows 7. It looks quite bad on a recent Chromium build but that's to be expected for Webfonts rendered in Chrom* running on Windows.
you should have checked github first. I'm using dotfiles that allows me to hjkl by tea-bagging, which requires less precision than my "stylus". Its called "dotmynuts," and its really active - 400 pull requests in the last 10 minutes.
See, dotfiles like yours are why there's not more women in technology. As a feminist, I demand that your dotfiles be made inclusive by also accepting input from boobs.
Also, who uses h and l anymore? It's 2013, you should be using w and b.
At the risk of a) being wrong and b) breaking the 4th wall by not conforming to said stereotype, looks like both my username and my tendency to ramble just within the limits of OT have been parodied there.... Oh well, if that's actually the case, then I'm honored :)
EDIT: btw, hate my username and emailed PG to get it changed; he said it's not currently possible in the code. Wouldn't want him poking around the DB for me! Made my username back when HN first started as an anonymous account, but kept using it thereafter... CamelCase usernames suck!
I think a parody of something is one of the best ways to draw insight, and in my opinion I'm kinda happy that this is what the parody of HN is: links to wikipedia/cited sources, debates about whether or not an article was correct, nitpicks and views from different positions/experiences, and the off hand XKCD comic.
Probably much more civil than a parody of a slashdot/reddit/4chan post would be.
Let's not be too self-congratulatory. Hacker News today is in many ways like reddit of a few years ago, and Slashdot of many years ago. And this is a somewhat bowdlerized parody anyway.
I think all social news networks will degrade over time, but it's still good to have these health checks every once in a while to see what the state of it currently is. It lets us know when we need to move on.
I have a related question: what is HN's stance on parody?
I've seen many parodies posted to HN that have been flagged to hell because some don't notice that it's a parody. (excellent example: http://news.ycombinator.com/item?id=5126318) I also have my own parodies that I've stopped posting because they've met the same fate.
201 comments
[ 4.4 ms ] story [ 343 ms ] threadYou obviously misinterpreted the author's intentions and you missed:
* The fact that it wasn't made to make you laugh
* Why do we laugh, what makes things funny?
* What is laughter, really?
(tongue in cheek, of course)
[1]: http://en.wikipedia.org/wiki/Laughter
[2]: http://science.howstuffworks.com/life/laughter.htm
actually, this text in the reply box highlights a huge problem with threaded (and voted) discussions in general: there is little incentive to reply to any already-huge thread, even with valuable content.
i've been thinking of how to solve this issue for the past week and have some ideas, working on a prototype.
replies need "simmer" time at the top to aggregate enough votes. if there is one huge reply thread that dominates, few will scroll through pages looking for new content.
just the ability to fold comments on HN would help a lot already.
Also, while this may be a problem here, HN still has the best comment sorting system I've seen (but not measured). Comments have time to stay at the top and simmer down and if they are good they gain enough momentum. The only time this seems to be come a problem is when you get to a really popular thread after hours/days of discussion. Then you cannot get a word in edgewise.
An alternative solution could be that for every comment you make, you give up a bit of karma and you gain it back through comment upvotes (but not article postings). That way people will watch what they say. This may give some positive feedback to posters that harvest huge amounts of karma, but it will also prevent bad comments. On top of that, you could make your post stick up top longer by paying a blood price: for every 50 points of karma you get another minute guaranteed at the top of your discussion thread, or some such. Or just free market FTW: the person that bids the highest amount wins the top spot for that many minutes.
I think they changed the underlying code, but http://forum.iwethey.org/forum/main.iwt still looks the same. So you can experiment with the idea there.
My take on the subject is:
This effect can be mostly attributed to the fact that a top comment means more exposure, and given that the upvotes outweight the downvotes on the response (which is true, since that's why it is at the top), by deductive reasoning we arrive to the obvious conclusion that more exposure will further cement the position of a given post.
Also we must not forget about the effect of peer pressure[1]
[1]http://en.wikipedia.org/wiki/Peer_pressure
For those with a lack of humour, I include a smiley here: :-)
Furthermore, I'm not really sure why this belongs on HN, because it's not very technical, and frankly, not very intellectually stimulating. People on here don't appreciate humor, so those who upvoted this should have known better. I've flagged the article.
Also, I've never even heard of Brad Conte.
[1] http://en.wikipedia.org/wiki/Parody
>People on here don't appreciate humor
How could you possibly come to such a conclusion? I appreciate humor, and I am sure others do as well.
* You omitted that HN is also a place to talk meta.
* You misunderstand that a parody is not supposed to capture the full scope of something.
* It was funny to read, that's all that matters.
I'm only replying to you comment because it's the top voted one.
EDIT: saraid216's response was funny, without being cringe inducing
Frankly there's no point in even continuing this conversation until you do and we should just assume that I'm right and you're wrong.
How much you pay to read a blog post on why that isn't important?
I'm honestly really curious about this. Could you elaborate?
[1] totalfratmove.com
[1] http://www.paulgraham.com/relevantessay.html
There's surprising amount of hidden jokes in there too - feel free to read the OP's blog post [1] for many of the details (assuming you like spoilers)!
[1] http://bradconte.com/hacker-news-parody-thread.html
Also, you forgot at least one archetype: the enthusiastic rant from an oblivious hell-banned poster.
There should also be a subthread complaining to the moderators about the title change.
;)
But now its time to be very HN about it.
> Rats, top comments will be impossible to beat.
Actually I think I have a solution to this, though its just guesswork.
I'm not particularly well known here (or anywhere), but either people actually like the words I say, or I simply have yet to make a complete ass out of myself. (I think one of these is more plausible, but who is to say.)
My average karma on HN is 20.59, which seems to be a lot. Specifically, on the Leaderboard[1] that puts me in fourth place among the HN big names for average karma (though I have nowhere near the total karma to appear there).
I've noticed that when I reply to a post, even if there's already 30 comments, my post usually ends up at the top. And it stays there, even if no one replies, and even if I don't get many or any upvotes for it, for a few hours sometimes.
Alas I don't know for certain, but my guess is that if you have a high average karma then your posts are automatically weighted higher, so you can inject your opinion into almost any topic at any time. This affords me the luxury of being a contradictory whine even if I come late to the party!
Can anyone confirm or deny my suspicion here?
[1] http://news.ycombinator.com/leaders
Darn you ;)
One thing I dislike about the 'average karma' is that it discourages engaging in an extended discussion. For example, if you reply to this post, you may get a number of upvotes, but probably not as many as your original post. This creates an incentive just to ignore replies (when possible) so as not to bring down your average comment score.
Done. :P
This algorithmic craze is too clever for its own good. It is particularly evident in the false-positive bans.
(And I swear, it's scary how HN joke comments are indistinguishable from the real ones.)
Source?
It's not like I'm a menace to the community.
http://www.google.com.au/#hl=en&output=search&sclien...
Edit: Why am I being downvoted????
But you're correct in that the current algorithm (or any sufficiently transparent algorithm) does lend itself to gaming. It might be harder to game if you didn't mind penalizing comments based on the quality[1] of replies they generate, then sorting based on the thread's average quality would mean that there's less incentive to post in the top thread just to get exposure, since your new, karma-less post will drag down its average and subsequently could drag down the thread. That could get prohibitively expensive to calculate for huge threads, but would be less prone to gaming, I think.
[1] If we can hand-wave "karma" to mean "quality"; I get way more upvotes when I am flippant but technically correct than I do when I try to be thoughtful.
Locally perturbing and simulated annealing sound interesting. Maybe we could convince pg to reticulate the splines as well.
I assume the goal is to encourage high-quality discussions (including extended discussions/back-and-forth), in which case average comment score will work against that.
I never would have guessed.
I wonder if this thread could be turned into a parody of the parody?
~ The matrix revolutions
Crisis averted.
As a tool or "quality score" that aims to control and keep a forum from derailing this kind of a scoring system has probably done OK. Not sure if it is solely responsible for these effects on sites like HN and SO, but there's not denying that the signal to noise ratio is far better than for example USENET S/N or some forums out there (vBulletin or phpBB type).
Having operated a forum before I can tell you that it can be an absolute nightmare (I was using vBulletin) on many fronts. HN seems to be able to maintain decent quality.
http://contrastrebellion.com/
I mainly do my programming by SSHing to it and using Vim on my ipad.
I've both actually done this, and told people about it on HN. I'm ashamed.
Also, who uses h and l anymore? It's 2013, you should be using w and b.
EDIT: btw, hate my username and emailed PG to get it changed; he said it's not currently possible in the code. Wouldn't want him poking around the DB for me! Made my username back when HN first started as an anonymous account, but kept using it thereafter... CamelCase usernames suck!
Probably much more civil than a parody of a slashdot/reddit/4chan post would be.
The good content can stand on its own. No need to look down on other sites to look good by comparision.
I've seen many parodies posted to HN that have been flagged to hell because some don't notice that it's a parody. (excellent example: http://news.ycombinator.com/item?id=5126318) I also have my own parodies that I've stopped posting because they've met the same fate.