Crunching last posts. August 27, 2010 2:23 PM Subscribe
Riding off into the sunset on my pony.
Could someone please crunch the infodump and get some stats about people's final contributions to MeFi?
If we set some sort of standard (e.g., last post/comment was over a year ago,) could we get some stats about users' final posts? Like, average age of the poster, length of post/comment, etc?
Could someone please crunch the infodump and get some stats about people's final contributions to MeFi?
If we set some sort of standard (e.g., last post/comment was over a year ago,) could we get some stats about users' final posts? Like, average age of the poster, length of post/comment, etc?
...or we could only crunch those with disabled accounts?
posted by griphus at 2:25 PM on August 27, 2010
posted by griphus at 2:25 PM on August 27, 2010
We'd need to put somebody in charge on the crunching. A captain, if you will.
posted by mmmbacon at 2:30 PM on August 27, 2010 [64 favorites]
posted by mmmbacon at 2:30 PM on August 27, 2010 [64 favorites]
It's the ratio of Good Posts : Lady Gaga Posts isn't it?
posted by mccarty.tim at 2:37 PM on August 27, 2010
posted by mccarty.tim at 2:37 PM on August 27, 2010
Every final comment in MeTa is a variation on "Whatever happened to free speech? I guess you people just can't stand someone who's not afraid to disagree with the political correctness police."
posted by Horace Rumpole at 2:40 PM on August 27, 2010 [4 favorites]
posted by Horace Rumpole at 2:40 PM on August 27, 2010 [4 favorites]
Yes, why?
posted by nomadicink at 2:40 PM on August 27, 2010
posted by nomadicink at 2:40 PM on August 27, 2010
Because, like a woman trying to get impregnated, I want to choose the optimal time to flame out, call y'all a bunch of butterheads and take my ball and defect to Reddit.
Seriously, though, I've noticed that a lot of people I am linked/linked to are disabling their accounts and I just got to wondering about the statistics. I don't actually want to know specific last comments. Well, I do, but I'm not going to request automation of my morbid curiosity.
Also, I believe this guu has seniority.
posted by griphus at 2:44 PM on August 27, 2010 [1 favorite]
Seriously, though, I've noticed that a lot of people I am linked/linked to are disabling their accounts and I just got to wondering about the statistics. I don't actually want to know specific last comments. Well, I do, but I'm not going to request automation of my morbid curiosity.
Also, I believe this guu has seniority.
posted by griphus at 2:44 PM on August 27, 2010 [1 favorite]
...or we could only crunch those with disabled accounts?
Fuckin' ableist.
posted by gman at 2:48 PM on August 27, 2010 [1 favorite]
Fuckin' ableist.
posted by gman at 2:48 PM on August 27, 2010 [1 favorite]
Yeah, but it's only two people that you're linked to that have disabled their accounts and finding their last comment is easy.
posted by nomadicink at 2:48 PM on August 27, 2010
posted by nomadicink at 2:48 PM on August 27, 2010
I'm going to cut off my hand.
posted by kaibutsu at 2:49 PM on August 27, 2010 [5 favorites]
posted by kaibutsu at 2:49 PM on August 27, 2010 [5 favorites]
Of the people who link to you, only 1 account is disabled.
Goddamn it, if you're going to be stalker at least don't get lazy about it.
posted by nomadicink at 2:50 PM on August 27, 2010
Goddamn it, if you're going to be stalker at least don't get lazy about it.
posted by nomadicink at 2:50 PM on August 27, 2010
I'm pretty sure cortex and/or jessamyn are going to pop in here and drop a big ol' NUNYA on this request.
posted by carsonb at 2:51 PM on August 27, 2010
posted by carsonb at 2:51 PM on August 27, 2010
Er, again, I don't actually want to know specific last comments. I just want to know, site-wide, what the stats are on people who disable. Average age, length of last post. No specific info.
posted by griphus at 2:52 PM on August 27, 2010
posted by griphus at 2:52 PM on August 27, 2010
Perhaps being linked by you was the last straw.
Sorry, I'll just go start drinking now. Don't link me bro.
posted by vapidave at 2:54 PM on August 27, 2010 [3 favorites]
Sorry, I'll just go start drinking now. Don't link me bro.
posted by vapidave at 2:54 PM on August 27, 2010 [3 favorites]
I don't think the infodump notes whether an account is disabled or not.
posted by jedicus at 2:57 PM on August 27, 2010
posted by jedicus at 2:57 PM on August 27, 2010
Can we sort flameouts by pissiness? That would be cool.
posted by msalt at 3:04 PM on August 27, 2010
posted by msalt at 3:04 PM on August 27, 2010
Every post is my last post.
posted by 2bucksplus at 3:06 PM on August 27, 2010
posted by 2bucksplus at 3:06 PM on August 27, 2010
msalt: Can we sort flameouts by pissiness? That would be cool.
Finally, a real world use for numbers approaching infinity.
posted by MCMikeNamara at 3:29 PM on August 27, 2010
Finally, a real world use for numbers approaching infinity.
posted by MCMikeNamara at 3:29 PM on August 27, 2010
I'm not sure this would be a good idea, but it's not possible anyway. Here's the first few lines of
posted by Monday, stony Monday at 3:43 PM on August 27, 2010
usernames.txt from the infodump:
Sat Aug 21 02:27:37 2010 userid joindate name 1 2000-01-27 20:16:57.367 mathowie ... ... ...As jedicus said, the infodump does not include the information on whether an account is disabled or not.
posted by Monday, stony Monday at 3:43 PM on August 27, 2010
What do you mean by 'age' -- time elapsed between joining and last post/comment? Also, the infodump does not indicate whether an account is disabled, so you only have time since last post/comment to go on. And the infodump only has length data for comments, not posts, so if you really mean comments where you're saying posts you might want to clarify that.
posted by Rhomboid at 3:43 PM on August 27, 2010
posted by Rhomboid at 3:43 PM on August 27, 2010
Could someone please crunch the infodump --- I demand royalties.
posted by crunchland at 4:04 PM on August 27, 2010
posted by crunchland at 4:04 PM on August 27, 2010
I think a better use of this would be "average time of site participation", rather than some sort of flameout Thunderdome, but hey - doesn't look like this is technically possible anyway (at least not in an automated sense).
posted by djgh at 4:09 PM on August 27, 2010
posted by djgh at 4:09 PM on August 27, 2010
Can someone go make sure fishbike is ok?
posted by cjorgensen at 5:02 PM on August 27, 2010 [1 favorite]
posted by cjorgensen at 5:02 PM on August 27, 2010 [1 favorite]
i told you i was hardcore
posted by DU at 5:08 PM on August 27, 2010 [1 favorite]
posted by DU at 5:08 PM on August 27, 2010 [1 favorite]
We'd need to put somebody in charge on the crunching. A captain, if you will.
*in charge of.
I assumed The Crunching would be the name of the ship.
posted by mannequito at 5:34 PM on August 27, 2010 [1 favorite]
*in charge of.
I assumed The Crunching would be the name of the ship.
posted by mannequito at 5:34 PM on August 27, 2010 [1 favorite]
In a world... where data filled drives, untabulated and unrefined, one man must face the chaos with only his wits, his nerve, and his SQL. Coming this Fall to a server near you...
The Crunchening... This time it's personnel.
posted by Babblesort at 6:37 PM on August 27, 2010 [1 favorite]
The Crunchening... This time it's personnel.
posted by Babblesort at 6:37 PM on August 27, 2010 [1 favorite]
Yeah, I don't think this is such a terrific idea. There's already way to much "WHY DID THIS USER LEAVE?" going on here. That said, the infodump does not belong to me so maybe someone else can do this?
posted by jessamyn (staff) at 7:08 PM on August 27, 2010
posted by jessamyn (staff) at 7:08 PM on August 27, 2010
Can someone go make sure fishbike is ok?
I ATEN'T DEAD
Yeah, I don't think this is such a terrific idea.
I thought it sounded OK actually, as long as we stick with the high-level summary kinds of stats that griphus asked for in the post, and don't go looking at individual user activity questions like "who made the longest final comment" and that sort of thing.
That said, the infodump does not belong to me so maybe someone else can do this?
Looking it it now, yeah, but was your comment above more of a "meh, not interesting" or a "please don't do this" sort of thing?
posted by FishBike at 7:25 PM on August 27, 2010
I ATEN'T DEAD
Yeah, I don't think this is such a terrific idea.
I thought it sounded OK actually, as long as we stick with the high-level summary kinds of stats that griphus asked for in the post, and don't go looking at individual user activity questions like "who made the longest final comment" and that sort of thing.
That said, the infodump does not belong to me so maybe someone else can do this?
Looking it it now, yeah, but was your comment above more of a "meh, not interesting" or a "please don't do this" sort of thing?
posted by FishBike at 7:25 PM on August 27, 2010
more of a meh, if it's happening at stats level, totally OK with me.
posted by jessamyn (staff) at 8:02 PM on August 27, 2010
posted by jessamyn (staff) at 8:02 PM on August 27, 2010
If we set some sort of standard (e.g., last post/comment was over a year ago,) could we get some stats about users' final posts? Like, average age of the poster, length of post/comment, etc?
Yep, that works pretty well actually. As others have already mentioned, the Infodump doesn't tell us if an account is disabled or not. But a list of userids whose last post or comment happened more than 1 year ago is easily generated, and picks up people who simply stopped posting or commenting here without pushing the big red button.
So, the average "age" of such users is 1.43 years (from signup date to the datestamp of their last comment or post).
The average comment length, for those whose last contribution was a comment, is 461.1 characters. For comparison, the average length of all comments is 421.0 characters. The Infodump doesn't contain post length data, although I think cortex has mentioned this might get added one day.
There are 14,757 users whose last comment or post was over 1 year ago, and for 13,006 of them, their last contribution was a comment. That implies 1,751 last contributions were posts, which seems like rather a lot, except that I suspect a fair number of those are spammers of various kinds. Maybe I should exclude deleted posts?
Oh, and there are 30,969 unique userids who have ever posted or commented, so although quite a lot of people have been inactive more than 1 year, more than half the contributors here are still active by this definition. That's pretty encouraging, isn't it?
posted by FishBike at 8:03 PM on August 27, 2010 [2 favorites]
Yep, that works pretty well actually. As others have already mentioned, the Infodump doesn't tell us if an account is disabled or not. But a list of userids whose last post or comment happened more than 1 year ago is easily generated, and picks up people who simply stopped posting or commenting here without pushing the big red button.
So, the average "age" of such users is 1.43 years (from signup date to the datestamp of their last comment or post).
The average comment length, for those whose last contribution was a comment, is 461.1 characters. For comparison, the average length of all comments is 421.0 characters. The Infodump doesn't contain post length data, although I think cortex has mentioned this might get added one day.
There are 14,757 users whose last comment or post was over 1 year ago, and for 13,006 of them, their last contribution was a comment. That implies 1,751 last contributions were posts, which seems like rather a lot, except that I suspect a fair number of those are spammers of various kinds. Maybe I should exclude deleted posts?
Oh, and there are 30,969 unique userids who have ever posted or commented, so although quite a lot of people have been inactive more than 1 year, more than half the contributors here are still active by this definition. That's pretty encouraging, isn't it?
posted by FishBike at 8:03 PM on August 27, 2010 [2 favorites]
So, the average "age" of such users is 1.43 years (from signup date to the datestamp of their last comment or post).
What's the average "age" of everyone?
posted by cj_ at 8:25 PM on August 27, 2010
What's the average "age" of everyone?
posted by cj_ at 8:25 PM on August 27, 2010
What's the average "age" of everyone?
Going from signup date to date of last post/comment, it's 2.45 years. Or if you go from signup date to today's date, for everyone who ever posted or commented, then you get 4.52 years. And you get 5 years, almost exactly, if you go from signup date to today's date regardless of whether or not they posted or commented.
posted by FishBike at 8:34 PM on August 27, 2010
Going from signup date to date of last post/comment, it's 2.45 years. Or if you go from signup date to today's date, for everyone who ever posted or commented, then you get 4.52 years. And you get 5 years, almost exactly, if you go from signup date to today's date regardless of whether or not they posted or commented.
posted by FishBike at 8:34 PM on August 27, 2010
Of the people who link to you, only 1 account is disabled.
Goddamn it, if you're going to be stalker at least don't get lazy about it.
Uh, it takes one to know one?
posted by joe lisboa at 9:30 PM on August 27, 2010
Goddamn it, if you're going to be stalker at least don't get lazy about it.
Uh, it takes one to know one?
posted by joe lisboa at 9:30 PM on August 27, 2010
Well, the infodump doesn't include data on whether an account is disabled; I just ran the stats on accounts that haven't posted in the last 12 months, and I can tell you we actually haven't had any of those in the past year.
posted by Mike1024 at 4:06 AM on August 28, 2010
posted by Mike1024 at 4:06 AM on August 28, 2010
"Oh, and there are 30,969 unique userids who have ever posted or commented, so although quite a lot of people have been inactive more than 1 year, more than half the contributors here are still active by this definition. That's pretty encouraging, isn't it?"
No way to exclude switching of sockpuppets/screen names.
posted by klangklangston at 8:27 AM on August 28, 2010
No way to exclude switching of sockpuppets/screen names.
posted by klangklangston at 8:27 AM on August 28, 2010
No way to exclude switching of sockpuppets/screen names.
True, but my gut feeling is that the number of people who have been active serially under multiple accounts is a small enough fraction of the userbase that that won't fundamentally skew any analysis.
I've never really tried looking hard at those numbers; using some of the data not in the Infodump I could (or could work with pb to) try and take a high-level look at the landscape of serial/sock accounts some time, though for obvious privacy reasons that'd only ever be high level stuff. But it might be interesting to supplement my gut that way.
posted by cortex (staff) at 9:28 AM on August 29, 2010
True, but my gut feeling is that the number of people who have been active serially under multiple accounts is a small enough fraction of the userbase that that won't fundamentally skew any analysis.
I've never really tried looking hard at those numbers; using some of the data not in the Infodump I could (or could work with pb to) try and take a high-level look at the landscape of serial/sock accounts some time, though for obvious privacy reasons that'd only ever be high level stuff. But it might be interesting to supplement my gut that way.
posted by cortex (staff) at 9:28 AM on August 29, 2010
"I die, that France may live!"
posted by planetkyoto at 7:08 AM on August 30, 2010
posted by planetkyoto at 7:08 AM on August 30, 2010
I thought most last posts were related to the "is this safe to eat" tag.
posted by arcticseal at 1:22 AM on August 31, 2010
posted by arcticseal at 1:22 AM on August 31, 2010
"FREEEEEEEDOM" *crunch*
posted by WalterMitty at 5:28 PM on September 1, 2010
posted by WalterMitty at 5:28 PM on September 1, 2010
« Older We hold their wonderfulness to be self-evident... | Megafiller.com Makes it to Time Magazine Newer »
You are not logged in, either login or create an account to post comments
For now.
posted by Plutor at 2:24 PM on August 27, 2010