Ask HN: Do posts by authors with more karma rank higher?
As the title states.. Is the HN ranking algorithm based purely on a post/comment's upvotes and engagement, or is the author's karma points an input into the algorithm as well? No. There's no documented karma input to ranking and I doubt any exists. Stories I submit are picked up as randomly as anyone else's. The same isn't true of stories I write; I have a pretty good track record of getting blog posts on the front page. But HN doesn't know when something I've submitted is something I've written; what gets stories from high-karma users ranked is mostly just name recognition. I doubt people are looking at the submitter name when choosing to upvote stories; I do think some subset of them look at the domain name of the story site though. I have never checked out how much karma anyone has in here, but after reading you I went on to have a look. First time I look at this, and it turns out you have "331331". That was a strange enough number that for a split second made me think I was looking at some random description field and you had put some magic number you like (a la 1337 or the likes) https://github.com/minimaxir/hacker-news-undocumented This might help answer some of your burning questions. Even this doesn’t stress how heavily HN is moderated. You can see it occasionally when there is some news cycle and you look an HN and everything is “Rust is cool. How we raised a round for a product we don’t have (yet). Tips for thinking effectively while in the open office.” This basically tells you that dang decided enough is enough and cleaned up the front page. > when there is some news cycle and [the front page does not reflect it] Political news is discouraged by the site guidelines and quickly flagged by users without mods' intervention (source: I'm one of these users). If I want the latest Trump nonsense, I can go to MSNBC or Fox News or Colbert or whatever. I'd rather have more stuff on the front page which we can have a (mostly) constructive conversation about. Some of us actually want to discuss politics with sane, smart people so I rarely flag thise stories (but I understand why you do.) I think this is a dicey area for discussion. Political discussions have always, and particularly so today, essentially consisted of ripping off bits of dogma and throwing at each other. While my family and I are generally on the same page, I avoid political discussions with them as well. I am often inclined to flag political threads here if they get egregiously bad. It very often gets to be the opposite of interesting. > Political discussions have always, and particularly so today, essentially consisted of ripping off bits of dogma and throwing at each other. Despite that I have actually learned stuff here from people that I originally disagreed deeply with. (Oh, and I am also fairly sure certain people have/could learned/learn something from me, but thats up to them to seize the opportunity. d:-) > I am often inclined to flag political threads here if they get egregiously bad. I wish they were better but often I wish people could stick to flagging bad behavior instead of the entire thread. That said, I am not here to challenge moderator decisions. Even with showdead on this place is fantastic compared to certain places I find it easy to compare with. > Despite that I have actually learned stuff here from people that I originally disagreed deeply with. Agreed. The incidence rate of people who can intelligently argue a controversial position is much higher here than in the average population. Agreed I use the hckrnews client which is better for seeing comments, and time of posting woah, this is crazy! The flamewar detector (# comments > upvotes) is pretty funny There was a large thread about that article 2 days ago: https://news.ycombinator.com/item?id=23439437 and https://news.ycombinator.com/item?id=19212822 in 2019. I have moderately high karma, and I haven't noticed any evidence that my posts are treated favorably as a result. I'm less sure on comments. At times I have suspected that comments from high karma users may get bubbled up, but I certainly couldn't swear to it. > I'm less sure on comments. Is your perception based on comments appearing near the top right after you post them? Because that's a feature. New comments are always at the top for a few minutes to get their initial set of votes before they move to their natural position in the stack. I‘m also quite certain that, with more karma, comments stay at the top longer. Cannot say much about submissions. I have seen anecdotal evidence of this too. I have noticed some of my comments have not been upvoted and yet have stuck higher in the rankings than comments that I know have received upvotes because I upvoted them myself. Although dang has stated that account karma doesn't impact the ranking algorithm so who knows. Either way, it seems obvious there is something more complex than a simple formula based on upvotes and post age. Fwiw, I saw this post on the front page, with no comments and you don't particularly have THAT much karma. Whatever that means. Haha good point FWIW, I just commented on the Ask HN about career advice and my comment landed about 2/3rds down the page. Seems like a data point for the "ranking of comments does not look at karma" point of view. With 90556 in karma you're pretty close to 100k! I guess you get to find out whether HN sends you a Silver Upvote Button in the mail. I don't know if it is purely based on upvotes but what I can say for my account is that on average I used to get significantly less then one karma point per upvote. When I watched the upvote counter on the individual item[1] and compared it with my karma increase a view years ago it was pretty consistent at 0.5. Back then I shrugged this of as effect of the voting ring detector but it left me wondering: can the false positive rate be so high and so consistent ... IDK. Needless to say that I never participated in a voting ring and I'm not that well connected that many upvotes should be from the same people. An alternative explanation that crossed my mind was if it could depend on the story/comment ratio because back at the time I submitted much more stories than I wrote comments. I did this with good intentions, I often like to read what other think even if I don't get involved in the discussion, but I could understand if a platform incentivized a healthy ratio of submissions and discussion. [1] I can't remember if this was true for stories or comment but I think I could only see it for one type of submission. Not all upvotes count towards karma. This is related to HN's anti-abuse systems. It would be nice if it could just be simple, but one-to-one mappings turn out to be the most gameable things, so we've had to do fancy footwork over the years. Makes sense. Just to avoid the wrong impression, when I said I never participated in a voting ring, what I really meant was that I never tried to game the system im any way. Easy for me to say, I know, unfortunately impossible to proof. Anyways, I was just curious about my story/comment ratio theory, but I guess I will never know... Related question: do up/downvotes from users with more karma count more than those from users with little karma? No. You can create a new account, upvote someone and check how their karma changes. The latter would create a bubble of some people’s submissions flooding the site over time and not allowing others’ submissions to come to the front page. The ranking is affected to a good extent by when a post is submitted because of the times (and time zones) when more HN voters are online and active. Yeah this is really interesting. The "velocity" of upvotes/comments seems to be more important than the actual raw count. I have had a few of my post make the front page and I don't have a lot of karma and don't personally know anyone else who has a user account here. I've also posted a link that got buried without anyone looking at it, then came back later in the day to see someone else posted the same thing from a different source and it made the front page. If there are others gaming their posts here to get karma points they can have them. As far as I know there's no place to cash them in so it doesn't cost me anything. It's nice to see others find what I've submitted interesting and "karma" gives us a measure of that, but the reason I check into HN is to find things that interest me. I'm new here, but my limited experience so far has been good. I've only submitted a handful of things, of which I initially thought "this deserves much more up votes," but in retrospect I came about and acknowledged my bias because I'm the author or it fits my worldview. I don't think all my posts merit their current karma, but most do. Also, somehow I find myself being disincentivized to post unhelpful comments.
As a matter of fact, every time I share an article I found on Hacker News I recommend people reading the comments. This site definitely has the highest helpful to unhelpful comment ratio. I suspect even if it's not done deliberately it happens as a self-fulfilling prophecy. The people that have high amounts of karma tend to post better articles so people are more likely to look at them or follow the person. I've sometimes submitted some of my own content here, usually without much happening. Then have someone else (not always with more karma) submit it again 1 to 24 hours later and it got to the front page. I've also had content do fairly well on e.g. Reddit or Lobsters be pretty much ignored here as well. I call it the HN roulette. I think HN is just too large and a lot of good content (or at least, I'd like to think my own content is good, heh) gets missed. Quality of the content is definitely a factor, as are things like an enticing headline, but you also just need to be lucky. Things that I have seen that influence whether a story will end up being ranked: 1) Timing. Early in the morning Eastern US time 2) The uniqueness of the posting. When everyone is posting COVID or protest stories, a story about an interesting astronomical event (e.g. PHA) is slightly more likely to get a kick. 3) I suspect that there may be a tendency to browse /newest by some folks and if they see submissions by posters they know, there may be a slight liklihood of upvotes. Not 100% sure, but:
- Karma is indeed a good indicator.
- Having a 'non'-anonymous HN profile also seems to make your posts more credible. Otherwise, good luck! Karma is a meaningless number. It depends more on luck than good management. I've put well-written posts out there with no karma given. I've also written throw-away lines that gain oodles of karma. Go figure. 20 million karma points and five bucks will buy you a coffee. I always thought it would be neat to apply something like pagerank to reddit (or HN, or whatever) comments. Surprised nobody has done this. (Maybe it would turn into an echo chamber pretty fast, but I'm not 100% certain, which is like I'd like to see someone try it.) Technically, I suspect there's a very strong correlation. You get a lot of karma for popular posts. But never seen any evidence that the system gives you preferential treatment if you have good karma. I've got a fairly decent karma score and everything I post sinks like a stone. :) No, but the only way to get a post to the front page of HN these days is to discreetly ask a bunch of your friends to upvote it. And people with high karma are probably people who tend to have other friends who are also HN readers and who they can ask for upvotes. As much as this probably isn't the fair system we all want, it's what the system's optimization encourages today. Try posting something SUPER interesting, staying quiet and not telling anyone. It's almost guaranteed not to make it to the front page. Although I don't encourage this, with the current algorithm, you could probably even prevent others' from getting their work on the front page by posting it ahead of time. When they try to post it they'll get a duplicate link but it's already stale and past its upvote-to-front-page life, which is probably about 30-60 minutes. Why do you say these things? I'd be fascinated to know. Almost everything here is wrong. HN's anti-abuse software isn't perfect, but if you ask friends to upvote your article, there's a high chance it won't help and you'll get your (and their) accounts penalized in the process. We actually go through the penalized posts looking for things to rescue because people tank their own good work in this way so reliably. The dupe detector does not work the way you described. If you post something super interesting that doesn't get attention, you (or anyone) can always email us at hn@ycombinator.com and we might put it in the second-chance pool (https://news.ycombinator.com/item?id=11662380), in which case it will get a random placement on HN's front page. Thanks! This is helpful to know. I've seen lots of friends (including dozens of YC founders) ask for upvotes, usually via FB in ways that wouldn't necessarily constitute a ring (e.g. asking strangers in founder-friendly facebook groups), and they're usually of things worthy of publicity, and they do make it to the front page pretty quickly after the upvote requests. I've also seen a lot of instances where posting once didn't do anything, and then posting a second time with an equivalent but different URL + asking a few friends to help upvote made it to the front page pretty quickly. For example, I posted https://github.com/dheera/rosshow on 2019-Mar-27 -- no dice. Then posted it again with a "www.", i.e. https://www.github.com/dheera/rosshow -- then asked a couple friends to upvote -- and then BAM front page (https://news.ycombinator.com/item?id=19519165). So the only thing I did differently between trial A and B was asked people to upvote, and trial B succeeded. Assuming I'm reading the data correctly, your friends' votes were rejected by HN's software on that second submission. Randomness is the biggest factor on /newest, which is why we allow a small number of reposts to begin with: to give good submissions multiple cracks at the bat. Thanks, helpful to know! I would have thought reposts would be penalized. Is there anyway to potentially mitigate the randomness factor though? For example by measuring things like impression time, bounce rate, and so on. HN has a pretty good voting ring detector. And while folks have emailed me and asked me to upvote their submissions, my response is always "I upvote things I think are interesting and bring something new to the discussion, period." I would gather other "high karma" individuals have similar positions. > the only way to get a post to the front page of HN these days is to discreetly ask a bunch of your friends to upvote it This is not true. I don't do this and 5/13 of my submissions hit the front page in the last 12 months. The most important thing is to find an article that already has a suitable title so that you don't have to editorialize it against the site guidelines just to get the crucial first upvotes. It's a shame that often the best article on a topic doesn't have the best title (often when it is a primary source). Sometimes the mods change the link later. This is a difficult subtlety to get across publicly, but it can be ok to edit the title to grab attention when an article is particularly good for HN. That sin is venial. If the article isn't particularly good, and especially if someone is just trying to promote something, woe betide. Not true. It has much more to do with timing - both time of day and submitting a post when the audience is receptive to it, not necessarily being the first to do so. I have been here a long and have relatively high karma; about 3/4 of my posts get no traction, some get a small amount of attention, some go to the front page and stay there all day. The posts that are most successful in terms of getting attention are often not the ones I think most deserving of it, eg scientific journal articles with the meatiest content usually fare poorly. I've never asked anyone to boost a post and being a rather nati-social creature I don't have an audience of devoted fans, if anything the opposite. I could get more karma if I was selective about the time of day I posted (eg lunchtimes are usually good) but I just post stuff as I discover it. I don't know where my karma ranks, but I've had a few front page submissions - nothing original. Just interesting reviews of brand new tech, that sort of thing. I simply posted the article first and lots of people were interested in it. To me, it seems obvious that the value of the content and timing are the two biggest levers. I've gone to post things that are new and interesting, but someone else posted it, my submission counted as an upvote, and the submission ended up on the front page. I have never asked anyone to upvote anything. I'm sure you're right that people occasionally leverage a friend network to upvote content to the front-page. I get the sense this happens a lot on Product Hunt. But I know for a fact this isn't the _only_ way to reach the front-page of HN because I didn't ask anyone to upvote this post. ¯\_(ツ)_/¯ For what it's worth, Ask and Show posts are less likely to fall into the abyss than regular posts because they both have their own front pages where the threshold is lower than the main FP. So they have a longer exposure time to gain the upvotes to make it to the FP. Isn't "super interesting" relative/subjective? And if it doesn't make FP its therefore not super interesting? Heck, I'm low karma, posted mildly interesting things that have been FP. Maybe FP is the measurement of interesting? Relative to the general audience. Ideally, whether or not something makes it to the front page should depend on whether the readers of HN think it's interesting. However, the current (sub-optimal IMO) situation is that as long as it meets some pretty low bar for "interesting", it depends MUCH more on the submitter's friend network than the readers.