Is HN Changing? - Growth

http://jacquesmattheij.com/is+HN+changing+-+growth

by jacquesm • 17 years ago

37 40 17 years ago

Nice data collection. Conclusions make sense.

SamAtt• 17 years ago

Nice data. I'd like to see it updated for a few months. I suspect if you do that you'll see the numbers are even higher than this shows. I think a lot of people go in and out of activity as their off-line life becomes more or less complicated.

tialys• 17 years ago

The karma inflation says a lot I think. There seems to be a rash of new users posting less relevant 'link-bait' stories that get voted up for being edgy/funny, but have little interest to hackers. I've noticed that i've been far less interested in the front page of HN lately due to most of the links on the front page being non hacker news or links I already saw on Reddit(!). This is purely anecdotal however.

btipling• 17 years ago

The community will do as it will. If you want to keep something on topic you need moderators, like subreddit.

seldo• 17 years ago

I for one have got a lot more strict about flagging things that are interesting but off-topic.

iamelgringo3• 17 years ago

I've tried to do that quite a bit, but since PG changed the rules for flagging last year, it hasn't been nearly as effective.

ubernostrum2• 17 years ago

There seems to be a rash of new users posting less relevant 'link-bait' stories that get voted up

When I first encountered HN, not long after its launch, it appeared to function largely as a syndicator of TechCrunch articles. So I find your characterization of a "rash" of new "link-bait" rather amusing...

jaf656s• 17 years ago

Did you wait a while before creating an account?

If you did, then you can ignore the rest of my comment.

If not, you could have joined during one of HN's "popularity peaks" when it was mentioned on another popular website, creating a wave of new users. Those peaks always seem to correspond with a rush of Techcrunch (and similar) articles.

grandalf• 17 years ago

I think the problem is that most HN readers don't bother flagging the link bait stories... many even vote them up.

I would be OK with a "no linkbait headlines" policy where any article that had a sensational headline such as "Why I will never use another Apple product again" or "Shocking lessons from my Google interview" should just be considered inappropriate for HN.

electromagnetic2• 17 years ago

I'd second this, I personally don't hesitate flagging something I think is inappropriate. I've been around HN for a fair while, and lurked a lot in the beginning too, so I will often flag an article that may have been appropriate sans sensationalism.

CWuestefeld• 17 years ago

Nice. Good questions in this specific instance, as well as a good example of what everyone running a web site should be doing.

pg4• 17 years ago

Here are the actual traffic stats as of today: http://ycombinator.com/images/hntraffic-19jan10.png

andreyf6• 17 years ago

Paul, have you considered re-implementing the orange-names experiment as a gradient of gray-to-orange, to avoid a tribal sense of "grays versus oranges"? Or is that not-obviously-scalable?

iamelgringo3• 17 years ago

Am I right in reading the graph, that when HN started, there were several thousand users?

pg4• 17 years ago

IIRC there were around 1500 uniques the first day.

smokinn• 17 years ago

Any idea what caused that huge spike in late August/early September?

pg4• 17 years ago

The thread on _why.

mortenjorck• 17 years ago

Any community that gains a critical mass of users will change. The interrogatory form of the headline isn't really necessary.

HN will change as it grows, just as did Slashdot, or Reddit, or Digg. It doesn't have to be a bad thing. YC can try to shape this growth by introducing features and changing algorithms, and the userbase can try to shape it through submissions and voting.

As with many things, it's all a question of balance. You don't want to tip things too far in favor of preserving the old culture lest you scare off newcomers and stagnate (perhaps Slashdot at some periods in its history), nor do you want to tip the other way and lose your core users (Digg).

Only time will tell how HN navigates these waters.

Alex3917• 17 years ago

My perception is that submission quality is roughly the same, but average comment quality has been dropping. Maybe the solution is to make comments work more like submissions, e.g. new top-level comments start out dead and need two upvotes before they go live. This is essentially how all new stories start out, and it seems to be working reasonably well.

Also, part of the issue with scaling is that there are some types of comments that are fine to post when the userbase is small, but that aren't ok to post as the userbase grows-- even if they add value in some way. So the problem isn't just coming from new users posting bad comments, it's coming from old users posting the same comments. So maybe some sort of radical experiment to try to refocus the discussions wouldn't be such a bad idea, at least to see how it goes.

andreyf6• 17 years ago

So the problem [is] old users posting the same comments.

Interesting hypothesis.

experiment to try to refocus the discussions

Expound?

iamelgringo3• 17 years ago

I've noticed that my comments have been typically getting less points than they used to. What I've noticed is that when I don't get as much positive reinforcement (points) as I used to, I'm less inclined to take 30 minutes and writing and editing a thoughtful comment. It's a downward cycle.

akkartik• 17 years ago

Same here. The problem is not more upvotes, but upvotes spread out over a larger set of comments.

Perhaps we need a study of comments in the different percentiles now vs a year ago vs 2 years ago. Filter out comments that don't tell you something new. How does the remainder look?

robryan2• 17 years ago

This may be because there are more comments on each submission so your comment may be glanced over instead of being read properly like it would be if there were only a few comments on something.

The comment system is usually very good but for things like the Google/China post it is very hard to come back a few hours later and see what's new because of how things are ordered, I'm sure there would have been a lot of valid points in that thread which didn't get upvotes.

JacobAldridge• 17 years ago

Maybe the solution is to make comments work more like submissions, e.g. new top-level comments start out dead and need two upvotes before they go live.

Is that also related to the way comments are now weighted against time, as well as value? Every new comment is momentarily - or longer - placed at the top of the comments page, so something trite will live above a 32 point well-reasoned post for a period of time, depending on how old the 32 pointer happens to be.

When the new comments went straight to the pile of 1 point comments at the bottom, there was less incentive to contribute late in a discussion as your input could be overlooked; in fixing that problem, there's now an increased noise-to-signal ratio as low-value comments filter their way down, rather than build their way up.

andreyf6• 17 years ago

I would love to see a kernel density plot of comments by point value, hypothesis being that there are a lot more low-point submissions than there were before. I see a lot of un-thought-out off-the-cuff comments lately: knee-jerk anti-Chinese sentiments, top comments with obvious flaws in reasoning, speculation, or just plain immaturity.

I know people have been saying this for years, but this is the first time I'll say it: I've been avoiding reading news.YC because I don't feel like I learn anything from the comments anymore. I'd really welcome a return of PG's orange-names experiment, but with a gradient instead of a binary grouping (to avoid tribal sentiment).

jacquesm3• 17 years ago

Average karma is a meaningless metric.

andreyf6• 17 years ago

Meaningless how? In some absolute sense? It's a metric that, like any public metric, will guide behavior. The question to ask is "how will it guide behavior?"

The most quick/intuitive explanation is probably the best explanation - I think it'll make people think twice "is this more insightful than my average comment?" before hitting Submit. As the community grows, this is exactly what we need: incentives for commenting less, and reading more.

pg4• 17 years ago

Interesting is the start of the real growth roughly a year after the launch, coinciding with a broadening of the scope of HN.

This isn't true. What you're seeing there is the usual holiday dropoff, followed by the huge bump that happened when Arrington outed us. Actually HN's growth rate has been pretty consistent throughout. A graph of x^n generally looks as if it's nearly flat initially and then takes off at some point, but this is an illusion.

It also isn't true that the change in focus happened then. We changed the site from "Startup News" to "Hacker News" after 6 months.

electromagnetic2• 17 years ago

There appears to be a major dip every Mid-December to January. There also appears to be a moderate, but much longer, dip starting in May and ending in July.

I'm guessing this speaks more of the average age of users. I expect users with children tend to get less online-time around the summer holidays and especially the Christmas break. I know personally, without children, that I had much less time to access HN around Christmas.

pchristensen• 17 years ago

Plus college kids

blasdel• 17 years ago

> Possibly it would be better to extend the length of the 'new' page because in practice once something moves to page 2 it is lost. It is very rare that something makes it to page 2 without enough upvotes and after that still collects enough votes to make it to the homepage, no matter how topical or interesting.

I don't think that's true at all. I'm not sure if pg logs enough data to test this, but I am quite sure that a lot of the popular stuff that makes it to the front page got there via multiple-submission, not off the new page. It's fucked if the first submitter is more than a day before the second, but the usual course is for multiple people to see something their feedreaders around the same time. This doesn't require fresh content, fresh references to old content work too.

Because of this, news.yc generally ignores any links that aren't presently being talked about elsewhere -- it's not the locus of discovery. Metafilter is the only site like this that manages to really handle it well.

Is HN Changing? - Growth

Related Stories