Hacker News new | comments | show | ask | jobs | submit login
Ask HN: Make the front page bigger?
10 points by jflatow 950 days ago | hide | past | web | 15 comments | favorite
Hacking votes has long been a problem on HN. It's well known that if you can collude to get enough votes in the first 10 minutes or so of your post, you can make the front page and continue to get votes based on interest. Without a hack, it's likely your post will barely get seen by anyone to even get voted on.

It'd be eye-opening to see the breakdown of traffic to the front page vs. the newest page vs. the more pages.

It seems like if you want your post to be seen, and actually voted on, you have little choice but to try and game the system. Maybe this is what YC wants.

Why not consider making the front page a little bigger to reflect the increase in submissions over the past several years? 33%? 50%? 100%? There are many other ways to relieve the pressure to game, but this one is a relatively small change.




"It seems like if you want your post to be seen, and actually voted on, you have little choice but to try and game the system."

There's an alternative. Posting good content. Sure, good content doesn't guarantee upvotes and collusion does. But colluding doesn't provide the feedback that teaches a person how to distinguish good content from mediocre, or good submission timing from poor submission timing.

That's assuming of course that someone thinks those things matter. I am not sure that they really do, except in so far as they matter more than karma points - particularly those gleaned from submitting other people's work rather than writing one's own comments.

I'll admit I'm probably out of touch, since my mental model doesn't really include the elements that fret over submissions making the front page. That's probably because I focus on writing, which is less about luck and more about skill.

It's not that I think the mechanics of HN are perfect: It's that I think HN is not broken.


I agree with you. I'm just suggesting that the amount of good content at any given time might be a little bit larger than 30 links. The community has grown a lot, but we still focus on the same number of links, I think that means that a lot of otherwise good content is getting missed, because if people are anything like me, they usually don't go past the front page.


What is the rationale for believing that there are more than 30 good links?

Why is this a function of the size of the HN community?

Why wouldn't a larger front page create a situation in which it was easier for collusion to move a post to a prominent position?


Hacking votes has long been a problem on HN. It's well known that if you can collude to get enough votes in the first 10 minutes or so of your post, you can make the front page and continue to get votes based on interest. Without a hack, it's likely your post will barely get seen by anyone to even get voted on.

Not exactly.

1) If there is voting collusion, the point counter will increase, but the points won't "count" to getting to the front page. (upvotes to posts which come from a non-HN referrer are automatically discounted)

2) Flagging is much, much more powerful than the effect of upvoting. If a terrible/advertorial article makes it to the HN front page despite collusion, it will die quickly as the increase exposured draws flags.


I guess it depends how the collusion actually occurs. The more 'sophisticated' voting rings can easily beat the referrer trick.

Flagging is good for demoting bad content, but not for promoting unseen stuff.

The idea here is to change very little, just relieve a little pressure - I've seen a number of posts recently about how to game HN, so it seems anecdotally to be getting worse.


"Promoting unseen content" won't be solved by increasing the size of the front page because the amount of traffic and attention a post receives is directly proportional to the absolute ranking (anedotally, posts from my blog which hit the Top 5 got 3-5x the concurrent traffic than posts at the bottom of the front page).

Adding more entries to the front page won't help at all.


I believe you that #30 < #5. Partly because if you ever hit #5 the total amount of time you spent on front page is greater.

I'm really curious about the jump from #30 -> #31 though.

a) what's the total drop in activity? b) how much has this changed in the past year? c) is there a 'better' cutoff to choose?


I've definitely seen "hey, please look for [article X] on the Newest page of HN and upvote, thx! (link to the HN Newest page)" on social networks before.


Not sure how useful meta discussion are regarding HN, but...

Have you tried some of the alternative interfaces? I hardly ever browse by the default front page any more. Try hckr news [0]. You can see at a glance the most popular posts of the day, and filtering by top 50% gives a good overview of the posts that could be on the front page modulo the time they were posted, what they are competing against, whether they quickly gained votes or not, etc.

Regarding diving up HN, I would love to see HN/programming, HN/business, HN/general, etc., but I would hate to see it get carved up too much. I like Lobsters [1] tagging system, and perhaps that would be a better fit for HN. Keep the front page general, but allow us to filter by interest. For instance, the compsci tag on Lobsters filters back to 11 days, and there are some interesting posts in that archive. Really eases the pain of thinking you'll miss something great by not checking HN for a couple days. You can easily dive into the past posts on hckr news as well, but you have to filter a lot more general stuff when doing so.

[0] http://hckrnews.com/

[1] https://lobste.rs/

Edit: I forgot to get my most relevant point across. I'm not sure increasing the front page would be all that helpful. We would still have people talking about good posts getting knocked off the front page no matter what size it is, and making it too large just means few will venture to the depths where some good posts land.


Using alternate interfaces such as HckrNews is part of the problem. Using interfaces like that means you don't visit the "newest" page. That means you don't vote for good stories from the newest page, and good stuff falls off, never to be seen.

If people don't visit "newest", good stuff won't reliably make it to the front page. People who care about contributing to the quality of the front page should visit newest, skim it, and upvote stuff that's interesting. It's doesn't take long, it's not onerous, and it would improve the odds of good stuff getting noticed.


I've seen some of them, always happy to learn about alternatives. Personally I like the HN interface though, which is why I'm proposing something that otherwise preserves it.

The issue for me with the alternative interfaces is that they seem to discourage voting / propagating points back to HN. I might be wrong about this but I've never fully understood how they all work.


Better to begin to divide up Hacker News, so that there is not a single front news page. Can 10,000 community forums be wrong?

At present there are already several front HN pages: "new", "ask" "jobs", "comments" and "news".

It is past time to divide up the "news" page to reduce the focus on the front news page.

HN is actually many communities of interest, and it is desirable to reflect that in HN.


Can 10,000 community forums be wrong?

Yes, because quantity != quality. And even worse, it causes fragmentation.

Reddit is a start, but requires a lot of manual effort to make Reddit an effective community tool.


Yeah, maybe, but that's also a much bigger change.

I guess it depends if HN is still growing much bigger or if it's somehow bounded. If it's bounded, then just increasing the size of pages might be good enough because you won't have to keep doing it to keep up.


We can call such sections and smaller communities sub-hns. :)




Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | DMCA | Apply to YC | Contact

Search: