3 ms·
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 mak
by minimaxir 13y ago
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.
- jflatow 13y agoI 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.
- hkmurakami 13y agoI'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.
- minimaxir 13y ago"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.
- jflatow 13y agoI 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?