I've noticed people doing that (on days where I check new frequently), so I've been curious if that's addressed. It's good to know that it is. :)
The annoying thing about that tactic is that there's plausible deniability for the perpetrator to nonadmins. Well, unless someone is using a Python scraper, anyways.
You're not limited with a Python scraper. Just scrape the comments url, and pull out the id number. That's unique, even if the story url and title aren't.
The main problem with this analysis is that the time on the /newest is not useful without knowing how many users saw it. Maybe staying up for 5 minutes on a Tuesday afternoon is better than an hour on Saturday night.
I assume the main goal is to get onto the front page rather than just maximize views on the /newest page. I am willing to bet that getting on the front page at the least active time is still far better than being on the new page at the busiest time.
If I understand correctly, the determining factor is not the time spent on 'new' page but the number of upvotes. During less busy hours submissions spend more time on the 'new' page but are seen by less people per unit of time and therefore get less upvotes per unit of time.
The time itself means nothing. It's about ratio of time on new page to upvotes per unit of time which might be higher during busy hours. That means that it's not impossible that the correlation between time spent on new page and probability of getting to main page is negative.
However, to get on to the front page all that is required is that your 'hotness' score (approximately votes/time) is better than most of the competing stories.
More time on 'newest' -> fewer stories posted. The 'hotness' requirement is lower. It's also likely that as no major press releases are made during the off-times that the competition is lower quality.
Less popular posts that might not make it during the week have a better chance of making it. Now, on a Sunday morning, the top three stories have about 20 upvotes each.
I submitted my side project last Tuesday @6am and it drowned within 10 min. I found a better exposure when I submitted to ASK HN instead. It would be great if we get the same analysis on the "ask" page.
Thanks mayank for sharing your work.
Looking at the ask tab, this is actually somewhat of a loophole people are using inappropriately. Prime examples on the ask page right now that have absolutely no business being there:
A lot of these are actually Show HNs exploiting the longer lifetime of Ask HN posts. That said, I like Show HN projects. Having a separate tab for "Show HN" (similar to Ask) would be nice, giving those projects more exposure.
I thought that the title of the post had to start with 'Ask HN:' for it to end up in the Ask section. One of the posts you linked is actually my project and I definitely had no intention of being sneaky. Nor was I trying to bypass any rules. I initially posted it with a "Show HN" title, but it was immediately marked as spam (for reasons that I still don't understand because I've never posted to HN before), so I deleted it and asked a friend to submit the project for me. Neither of us knew that leaving out the url automatically classifies a post as Ask HN. I've just looked at the submit form again, and I see where it's written, but it's not immediately obvious that that's what will happen.
It's impossible to know the intentions behind every single post, but together they're a trend and they do tell a story. What is unclear to me is why would anyone post their Show HN in that gray, de-emphasized text field where URLs are not even converted into links? Visitors have to actually copy and paste the URLs into their address bar. That makes absolutely no sense unless there is another advantage to be gained from submitting text posts (instead of URL posts).
Many of us don't have influential networks. We consider HN as the de facto community for hackers and entrepreneurs. Submitting a Show HN is not an exploit. It is a proud move to showcase ones hard work to a fellow audience. I personally think that the community is healthy enough to balance between all types of submissions: articles, Ask HNs and Show HNs in terms of both the people submitting and the people consuming and upvoting. What I agree with you is the need for a separate Shown HN tab similar to Ask HN. There are third party crawlers that do that but we spend all out time here; might as well have the "Show HN" tab here as well.
Nope. I think we are on the same page overall. I just wanted to reassure the new users on this thread that the general practice of submitting a "Show HN" under "Ask HN" is not considered as an exploit. Add to that your point that it should not be abused and I think we have a solid rule of thumb.
Are you sure you're not allowed to post Show HNs through the Ask HN form? I'm finishing up a project that I'd like to post as a Show HN, and I was actually just about to post an Ask HN to find out if I should post my Show HN as a link or an Ask HN (I feel like I'm in Inception - We need to go deeper!!!)
It seems to me that it has more to do with tradeoffs than an actual policy against it. Ask HNs are much less likely to be ignored outright, and they linger on the Ask HN page, which is great. However, they tend to drop off the front page much more quickly than a link submitted with a url and no text. I've never seen officially stated rules on how to submit Show HNs, so let me know if you've seen something different.
The ask section was probably not created to give people an additional opportunity to promote themselves extra hard, it was probably created for putting actual questions in front of the HN community. People who abuse this facility know this and do it anyway because they think their links are deserving of more visibility than those poor schmucks who just post a Show HN URL on /newest and watch it disappear with 0 upvotes within 30 minutes.
Posting an Ask HN to query whether you should do a Show HN strikes me as incredibly superfluous at best, and at worst it comes across as a pretty obvious scheme to post your link twice.
The answer to "should I post my project on Show HN" will always be "yes", so there is absolutely no legitimate need for an extra ask thread.
> It seems to me that it has more to do with tradeoffs than an actual policy against it.
At no point was I talking about policy. I was talking about decency, humility, fairness, and restraint.
> Posting an Ask HN to query whether you should do a Show HN strikes me as incredibly superfluous at best, and at worst it comes across as a pretty obvious scheme to post your link twice.
Easy tiger, I'm sure if I was looking to spam HN I would have figured out a way to worm a link into my last comment. Anyway, I assure you that's not the case. It's a perfectly legitimate question, given the debate you and I are engaged in right now, and the fact that there's almost no guidance on this from PG in any of the site docs.
Furthermore, the text form clearly isn't intended just for Ask HNs because it's not labeled as such, and because there are Tell HNs, Warn HNs, Ask PGs, Tell PGs, Dear HNs, I'm sure a few other variants I can't recall now, and posts with no labels, and tons of Show HNs that no one (except you) complains about. It's a reasonable question.
This is about people who like to post Ask items simply because they get more exposure (and I linked several examples above). I don't believe it takes guidance from the site owner to see why this practice is bad. It should be very clear to just about everyone that it's sneaky and does not enhance overall HN content quality. Not every example of sneaky, degrading behavior should need to be explicitly addressed in the FAQs.
As for the question if it's a good idea to Ask HN whether you should Show HN, I don't see any legitimate reason to do this. If your project is bad and unworthy (or just unlucky), it will get ignored and/or flagged. I know, because it happened to me, too. Several times, actually. It's a great signal to stop working on a project.
But chances are, your project is actually pretty nice and it will end up on the front page. There is absolutely, positively no need to ask whether you should show - other than to increase your visibility and circumvent the duplicate link mechanism.
> tons of Show HNs that no one (except you) complains about
I realize that in this, like in many other things, I might be utterly alone and that posting fake Ask HN items is actually considered a great hack to improve your exposure and everybody thinks it's just awesome.
That's OK. I'm just a random guy stating his opinions. I tried to explain why I think it's bad and why I think people who are doing that are abusing the system. While I would never personally upvote or engage with such an entry, it's totally possible that nobody else sees a problem here. In fact, I'm pretty sure pg and admins don't see a problem there, or measures would have been taken to prevent it. Nevertheless, I feel the need to say what I believe is right anyway.
It's up to you to decide whether I'm right or not.
It's the same submit page, just leave 'url' blank and write your post content in the 'text' box. The title can be whatever you want, though the common prefixes are helpful.
For those confused about why you would want to optimize the amount of time on the /newest page:
I believe it comes from a comment that pg(?) made at some point that once an article falls off the first page of /newest, it's pretty much impossible to get on to the front page. I'm not sure if that's part of the algorithm, or just an observation he was making though.
So, spending more time on /newest means that you have more opportunity to make it on to the front page.
I think a real problem is not enough emphasis is put on newest. I suspect not that many people (relative to the active user base) look at it frequently, so people end up "encouraging" others to vote up their posts, etc. It'd be cool if there was a way to encourage people to be exposed to these posts more (I use the firehose account on Twitter which helps a little, but not much.)
A while back I put together a doodad to collect and plot this sort of data over reddit (both as a whole and at a subreddit level), see http://reddalyzr.com/#/all The banner is there as I built it as a demo/dogfooding app.
Right now it only has a few days worth of data in it as the database backing it got blown up earlier this week.
I too often see better results when my submissions are in /new/ longer, but when you do get a "hit" article during the day to hit the front page it gets a lot more eyes.
As far the delete-resubmit thing, I don't see that a lot, and I highly doubt it works on a site as technically centered as this one is. Sounds like a great way to get banned.
one problem is, though, that what people like here may not match what you want people to read. it's easier to get a rant on the front page than some more interesting, but technical, article...
This doesn't seem like very meaningful data. So you want to stay on /newest for the longest amount of time? Submit Sunday night. Well I would venture not a lot of people read HN on Sunday nights, so what is the purpose of hanging around their?
The best trick I've seen to staying on the homepage is writing about interesting things, making novel points, or providing insight to currently relevant material.
Except that many a times (may I say most of the times actually), things that are good are still lost without being noticed. A clear proof comes from someone submitting something that does not get anywhere followed by someone else submitting nominally the same URL and making it to the front page.
Off topic:
The <div id="content"> has a fixed width, a left:50% and a margin-left:50%, making it difficult to read on smaller screens. In my case, that is my screen divided in half. It is also difficult to adjust with dev tools (I sometimes have to do this for readability).
( The post is something I'm very interested in, but I haven't had time to read it yet :P )
Nice analysis. I would love to see an analysis of the title changing behavior. E.g., the original title, and the title after being edited by the mods. (Sometimes this happens more than once). In fact, I think I've observed mods apparently fighting (with each other) over the title.
You should be able to examine that by looking at the MySQL data dump. I save titles as they appear on /newest, indexed by href, so you should be able to see title editing in action.
There is a better tactic, submit the same page with different random querystring params. Means, unlimited amount of submissions. Frequently used as well.
You can only flag submissions. Flagging is meant to be reserved for spam or other egregious violations, as far as I know, though the "rules" on HN are not hard and fast. I'll sometimes flag really egregiously link-baity submissions.
AFAIK the site itself, and mods (unknown who they are besides pg really) kill posts at their own whim. Flagging is one signal for that. Of course, you can just wait for the passage of time, which will "kill" any post on the front page eventually.
That's what I meant by "the site iteslf", sorry I wasn't clear. There's definitely some automated spam-detection, and voting circle detection, which auto-bans some posts.
Posts don't need upvotes to live, and don't need downvotes to die. I imagine there's some ranking done based on the number of upvotes or comments in the last X period against the amount of time it's been up at all.
> Please don't submit comments complaining that a submission is inappropriate for the site. If you think something is spam or offtopic, flag it by going to its page and clicking on the "flag" link. (Not all users will see this; there is a karma threshold.) If you flag something, please don't also comment that you did.
I generally understand "flag" to mean "Dear Mods, you should really look at this". /shrug
As per the rules I often flag things that are "off topic", especially things that violate the would-it-be-on-TV-news guideline. TBH, I think a downvote would be better than flagging for that but.. guidelines.
Look the best time to submit a article or idea is when you are ready to share the thought.
If you are relying on extra time on HN front page as your marketing strategy you have other problems.
If evenings in California is over-represented that's pg's problem as he tries to grow HN.
Reading this I was tempted for a micro-second to write some script that will submit when I am asleep - but then I realised if I did that, so would everyone else and I would have nothing interesting to read when awake.
I agree with this 100%. I share stuff from my website and stuff from a lot of other websites, and I notice when something does well or flops, but since I have no real vested interest in it, I don't care and don't want to let upvotes dictate what I share. If I think the HN crowd will like it, and they don't that's fine, move on. I'm sure those trying to sell something see it quite differently.
