Said, thanks for doing this, and also for running GW!
It looks like #posts/day is propped up by attention seeking, so it stays stable; while #comments/post is a more genuine indicator of health, and shows a slow decline. Also it’s curious that the relaunch of LW didn’t affect metrics much. Wonder what the mods would say about this.
I miss the days when there were hundreds of comments per day and I had to write my own Javascript code to scrape and display all the recent comments (which I called LessWrong Power Reader) to keep up. It seems really strange that even the AI alignment posts these days are getting so few comments compared to for example the decision theory posts back then, even though there must be a lot more people working on or interested in AI alignment today than decision theory back then?
Thanks for running GW ought properly to go to clone of saturn, who both wrote the server code (e.g.) and administers the server. I’m just the front-end guy! :)
As for the graphs/data—you’re quite welcome! (And, by the way, if you, or anyone, have suggestions on what other relationships or patterns might be interesting to extract from this data and make a chart of, let me know and I’ll add it to the post.)
Median number of comments/post seems to have gone up a bit since the relaunch, at least according to the graphs I posted below. Average number probably has probably stayed roughly the same, which I think is mostly the cause of Scott moving to his own blog, and his posts continuing to create infinite amounts of comments.
Overall, I think the launch shows up relatively clearly on a bunch of the metrics I posted below, and it’s too hard to eye-ball scatter plots to judge on the plots above, though I agree that the overall effect of the relaunch on number of comments is relatively minor.
Said, thanks for doing this, and also for running GW!
It looks like #posts/day is propped up by attention seeking, so it stays stable; while #comments/post is a more genuine indicator of health, and shows a slow decline. Also it’s curious that the relaunch of LW didn’t affect metrics much. Wonder what the mods would say about this.
I miss the days when there were hundreds of comments per day and I had to write my own Javascript code to scrape and display all the recent comments (which I called LessWrong Power Reader) to keep up. It seems really strange that even the AI alignment posts these days are getting so few comments compared to for example the decision theory posts back then, even though there must be a lot more people working on or interested in AI alignment today than decision theory back then?
Thanks for running GW ought properly to go to clone of saturn, who both wrote the server code (e.g.) and administers the server. I’m just the front-end guy! :)
As for the graphs/data—you’re quite welcome! (And, by the way, if you, or anyone, have suggestions on what other relationships or patterns might be interesting to extract from this data and make a chart of, let me know and I’ll add it to the post.)
Median number of comments/post seems to have gone up a bit since the relaunch, at least according to the graphs I posted below. Average number probably has probably stayed roughly the same, which I think is mostly the cause of Scott moving to his own blog, and his posts continuing to create infinite amounts of comments.
Overall, I think the launch shows up relatively clearly on a bunch of the metrics I posted below, and it’s too hard to eye-ball scatter plots to judge on the plots above, though I agree that the overall effect of the relaunch on number of comments is relatively minor.
Maybe? I think the user habits are pretty different on the site now compared to then. But I agree that more comments would be better :)