You're looking at Less Wrong's discussion board. This includes all posts, including those that haven't been promoted to the front page yet. For more information, see About Less Wrong.

TRManderson comments on [Short, Meta] Should open threads be more frequent? - Less Wrong Discussion

3 Post author: Metus 18 December 2014 11:41PM

You are viewing a comment permalink. View the original post to see all comments and the full post content.

Comments (26)

You are viewing a single comment's thread.

Comment author: TRManderson 19 December 2014 02:48:36AM *  6 points [-]

So I did what you suggested and plotted the number of top level posts and total posts over time. The attached graph is averaged over the last 20 open threads. Code available here: https://gist.github.com/TRManderson/6849ab558d18906ede40

I don't trust myself to do any analysis, so I delegate that task to you lot.

EDIT: Changed GitHub repo to a gist

Comment author: gwillen 19 December 2014 03:32:43AM 5 points [-]

For those who don't want to guess or dig into the source, the missing x axis unit is "minutes".

Comment author: solipsist 19 December 2014 03:39:34AM 0 points [-]

In other words, open threads die after about 4 days.

Comment author: Viliam_Bur 20 December 2014 07:48:49PM *  0 points [-]

That's probably when people are like "I will rather wait for another open thread, so my comment has more visibility". If we will have open threads more frequently, that moment might just come sooner.

However, I am in favor of an experiment.

Comment author: gjm 20 December 2014 10:48:33PM 2 points [-]

Posting near the end of an open thread's lifetime is (maybe) like cooperating near the end of an iterated prisoners' dilemma: others benefit (by having interesting discussions available more of the time) but you lose (by having what you post seen by fewer people).

One way one avoids this in IPD tournaments is by making the end uncertain (e.g., each turn after the 100th there's a 1% chance that the game ends). It would be interesting (though, frankly, probably not worth the effort) to randomize the starting of open threads in a similar way: e.g., once 3 days have elapsed, each hour a new thread gets started with probability 1%.

Comment author: FrameBenignly 19 December 2014 03:35:28AM 4 points [-]

Could you add labels so stupid people like me could figure out what that graph means? I'm guessing the blue line is top level comments and the green line is total comments; both averaged from all the weekly open threads. The y axis would be number of posts and the x axis should be time but I'm not sure what unit it's in. Also, why does the blue line suddenly stop?

Comment author: TRManderson 19 December 2014 04:09:15AM 3 points [-]

The blue line suddenly stops because the last comment is posted at that time. I was kind of lazy about this graph and did have labels and a legend, but apparently I was too out of it to realise they didn't show on the png.

As said by gwillen, x axis is minutes.

Comment author: TRManderson 19 December 2014 04:07:28AM 3 points [-]

Sorry about not having units, I added code to set them but apparently it was the wrong code and I wasn't paying enough attention.

Green line is total comments, blue is top level comments. X-axis is minutes, y axis is number of comments.