> Last month, I posted two links a day apart. The content of the articles was exactly the same. Granted, the titles and links were a bit different, but it’s nonetheless fascinating to see the 8X difference between the two.
The point here was that all else being (nearly) equal, sometimes posts take off, while others don't. Although you can optimize, at the end of the day, it can be often be a game of chance!
There's a difference between Hacking Hacker News and spamming Hacker News with low-quality posts, which the OP had done but conveniently failed to mention in the post. (the "Uber vs. Lyft" is the only post that is not disguised content promotion for his startup: https://news.ycombinator.com/submitted?id=gregmuender)
This is incredibly hand-wavy. For one thing, speculation about which days of the week are "better": Every single day, seven days a week, for any given moment in time, there are the same number of items on the front page. Does anyone have data on (for example) average numbers of upvotes varying in a statistically significant manner by days of the week?
For another, data about how many characters in HIS titles correlated to x number of upvotes: Maybe that says something about how he writes. But, again, is there any data about average title length for all articles that hit the front page? That would be more meaningful.
Not that I haven't done similar pieces on my own blog, but the point of one of my write-ups was: Gee, this completely shocked me and looks a lot like random chance to me.
I submit articles regularly. Some hit the front page. Some are ignored. Most are not my own articles. HN is a very large forum. The most consistent thing seems to be that certain authors (like Paul Graham) get their work submitted by people other than themselves and then get wildly upvoted and intensely discussed.
That isn't to say that analysis is meaningless, but I personally didn't get anything useful out of this article and another it linked to other than the fact that he now gets traffic from HN and from blogging at a level that, for a previous project, he spent up to $15K/month to get. But I have no idea what kind of money that is making him or how. Just that he now generates traffic without spending scads of money on it.
That's nice information. I am still trying to figure out how to get that piece working better for various projects of my own. But, for me, the single biggest thing is that I have other stuff to work out so content is updated very regularly and is good quality. Until I have that worked out, traffic will continue to be anemic -- which is fine. Working that out is a complex process and I am okay with the progress I am making.
Since I have a database of all HN submissions, I spent a few minutes to check. Turns out that title length (for non-1-word submissions) has no significant effect on submission score at all: http://i.imgur.com/1zthOwS.png
Nice chart! Yes, my sample size was rather small, so it's difficult to make absolute conclusions, but it was interesting to see a definitive correlation between post length and upvotes!
10 comments
[ 3.2 ms ] story [ 34.9 ms ] threadPeople don't like to read the same news twice?
For another, data about how many characters in HIS titles correlated to x number of upvotes: Maybe that says something about how he writes. But, again, is there any data about average title length for all articles that hit the front page? That would be more meaningful.
Not that I haven't done similar pieces on my own blog, but the point of one of my write-ups was: Gee, this completely shocked me and looks a lot like random chance to me.
I submit articles regularly. Some hit the front page. Some are ignored. Most are not my own articles. HN is a very large forum. The most consistent thing seems to be that certain authors (like Paul Graham) get their work submitted by people other than themselves and then get wildly upvoted and intensely discussed.
That isn't to say that analysis is meaningless, but I personally didn't get anything useful out of this article and another it linked to other than the fact that he now gets traffic from HN and from blogging at a level that, for a previous project, he spent up to $15K/month to get. But I have no idea what kind of money that is making him or how. Just that he now generates traffic without spending scads of money on it.
That's nice information. I am still trying to figure out how to get that piece working better for various projects of my own. But, for me, the single biggest thing is that I have other stuff to work out so content is updated very regularly and is good quality. Until I have that worked out, traffic will continue to be anemic -- which is fine. Working that out is a complex process and I am okay with the progress I am making.
Since I have a database of all HN submissions, I spent a few minutes to check. Turns out that title length (for non-1-word submissions) has no significant effect on submission score at all: http://i.imgur.com/1zthOwS.png