Academic schedule and MetaFilter angst/happiness. Is there a chart? April 15, 2010 1:35 AM Subscribe
Academic schedule and MetaFilter angst/happiness. Is there a chart? It just seems like one day we're all spousing and hugging and then next day it's a call for flameouts and pitchforks.
Where do the meetups with tasty beverages and great face time fit in?
I work in an academic setting and, as I watch the cycles of MetaFilter mood, I wonder if that doesn't affect the tone here. For instance, this week I've been dealing with stressed out students who are staring down their last few weeks of the semester. As a result, sometimes their stress adds on to my own person stuff. BLAM!, I want to punch someone in the face for whatever reason. That isn't like my usual self. Once the mental thunderstorm passes, I realize it's all this extra BS stressing me out.
Has there been an infodump study of MetaFilter mood versus academic cycles? I don't just mean for folks in the college/university industry (and it sure is an industry, my workplace university had about 5 million in budget cuts this year... that'll stress even the lowest of the totem pole, of which I am one of many), but including parents, students, teachers and staff of all ages, too. As a mom with two kids in middle school and all that extra time, energy, worry, etc... I'm just wondering if there is a connection.
If this is just a dumb thing to ask here or whatever, please just delete it. Maybe send me a MeMail if it's convenient. I'm sure not trying to add fuel to the fire. It's just that my sociology minor is showing.
Extra hugs to all that need them.
Where do the meetups with tasty beverages and great face time fit in?
I work in an academic setting and, as I watch the cycles of MetaFilter mood, I wonder if that doesn't affect the tone here. For instance, this week I've been dealing with stressed out students who are staring down their last few weeks of the semester. As a result, sometimes their stress adds on to my own person stuff. BLAM!, I want to punch someone in the face for whatever reason. That isn't like my usual self. Once the mental thunderstorm passes, I realize it's all this extra BS stressing me out.
Has there been an infodump study of MetaFilter mood versus academic cycles? I don't just mean for folks in the college/university industry (and it sure is an industry, my workplace university had about 5 million in budget cuts this year... that'll stress even the lowest of the totem pole, of which I am one of many), but including parents, students, teachers and staff of all ages, too. As a mom with two kids in middle school and all that extra time, energy, worry, etc... I'm just wondering if there is a connection.
If this is just a dumb thing to ask here or whatever, please just delete it. Maybe send me a MeMail if it's convenient. I'm sure not trying to add fuel to the fire. It's just that my sociology minor is showing.
Extra hugs to all that need them.
philip-random, I've had almost the exact opposite impression, that things have been extra-fighty recently. The C word issue (both Catholic and anatomy), the recent (today) Buddhism grar, even police being added to the "Things MetaFilter is bad at list (along with Palin, I/P and others). I'd love some happy friendly MeFi.
Personally, my own fuse is rather short, as due to the Japanese academic cycle, I've just started a new job, and I'm tyring to find my way. On the other hand, I can't post much because I don't have a computer at work yet, so it balances out.
posted by Ghidorah at 2:32 AM on April 15, 2010
Personally, my own fuse is rather short, as due to the Japanese academic cycle, I've just started a new job, and I'm tyring to find my way. On the other hand, I can't post much because I don't have a computer at work yet, so it balances out.
posted by Ghidorah at 2:32 AM on April 15, 2010
Agree that there is something in the air - most everyone I know is super-cranky/crazy/hyper-unreasonable in some way or another at the moment.
Given that these people are not in academia, I figure it is seasonal.
Am planning to diarise to remind self to check on unhinged levels in Spring/Autumn.
posted by the.carol.baxter.experience at 2:50 AM on April 15, 2010
Given that these people are not in academia, I figure it is seasonal.
Am planning to diarise to remind self to check on unhinged levels in Spring/Autumn.
posted by the.carol.baxter.experience at 2:50 AM on April 15, 2010
My mercury is in centigrade. And this sounds like (dare I say it?) the other C word: confirmation bias.
posted by pracowity at 3:05 AM on April 15, 2010 [2 favorites]
posted by pracowity at 3:05 AM on April 15, 2010 [2 favorites]
Do you think rate of comment deletions could function as a proxy for the rate and extent of threads turning all GRAR-y? If so, it might be interesting to see that plotted against time; does anyone with l33t infodump skillz feel like trying it?
My impression from another online community (most of whose members were on an academic schedule) was that Forum Drama is cyclical, but not based on any obvious external factors. It just seemed to be that tensions between various users and factions gradually build up over time, then explode and set off a load of side-arguments as people either get sucked in to existing spats or are just feeling more edgy and defensive than normal, and end up taking offense at ambiguously-worded posts. Then it quiets down as people make up, leave or resolve to avoid their nemeses and a new era of occasionally strained civility reigns... until the next time.
Metafilter is far too big a community for me to keep track of, so I have no idea whether this applies here. Just a thought.
posted by metaBugs at 3:17 AM on April 15, 2010
My impression from another online community (most of whose members were on an academic schedule) was that Forum Drama is cyclical, but not based on any obvious external factors. It just seemed to be that tensions between various users and factions gradually build up over time, then explode and set off a load of side-arguments as people either get sucked in to existing spats or are just feeling more edgy and defensive than normal, and end up taking offense at ambiguously-worded posts. Then it quiets down as people make up, leave or resolve to avoid their nemeses and a new era of occasionally strained civility reigns... until the next time.
Metafilter is far too big a community for me to keep track of, so I have no idea whether this applies here. Just a thought.
posted by metaBugs at 3:17 AM on April 15, 2010
My contributions to the site are always friendly, cheerful and angst-free, as a careful reading of my posting history will verify. Perhaps this is because I never attended university.
posted by flapjax at midnite at 3:36 AM on April 15, 2010 [1 favorite]
posted by flapjax at midnite at 3:36 AM on April 15, 2010 [1 favorite]
Agree that there is something in the air - most everyone I know is super-cranky/crazy/hyper-unreasonable in some way or another at the moment.
...Wait. maybe the "something in the air" causing agita isn't temporal -- it IS allergy season in many parts of the world.
posted by EmpressCallipygos at 3:59 AM on April 15, 2010 [1 favorite]
...Wait. maybe the "something in the air" causing agita isn't temporal -- it IS allergy season in many parts of the world.
posted by EmpressCallipygos at 3:59 AM on April 15, 2010 [1 favorite]
It just seems like one day we're all spousing and hugging and then next day it's a call for flameouts and pitchforks.
We're just bipolar.
posted by octobersurprise at 4:13 AM on April 15, 2010
We're just bipolar.
posted by octobersurprise at 4:13 AM on April 15, 2010
Nah, multiple personalities.
posted by Brandon Blatcher at 4:19 AM on April 15, 2010
posted by Brandon Blatcher at 4:19 AM on April 15, 2010
Do you think rate of comment deletions could function as a proxy for the rate and extent of threads turning all GRAR-y? If so, it might be interesting to see that plotted against time; does anyone with l33t infodump skillz feel like trying it?
Deleted comments aren't in the infodump, so we can't study them that way. Neither is data about flagged comments in there. Both of these exclusions are good things, in my opinion. But it does mean it would have to be one of the MetaFilter staff analyzing comment deletion rates and comment flagging rates, if anyone were to do it at all.
posted by FishBike at 5:19 AM on April 15, 2010
Deleted comments aren't in the infodump, so we can't study them that way. Neither is data about flagged comments in there. Both of these exclusions are good things, in my opinion. But it does mean it would have to be one of the MetaFilter staff analyzing comment deletion rates and comment flagging rates, if anyone were to do it at all.
posted by FishBike at 5:19 AM on April 15, 2010
(though it occurs to me one could just look for gaps in the comment ID sequence to count deletions that way, as a raw number, though I'm not sure how reliable that is. Or how good an idea that is, either.)
posted by FishBike at 5:19 AM on April 15, 2010
posted by FishBike at 5:19 AM on April 15, 2010
I think it would be possible just count posts tagged "etiquette" on MeTa. I think most of them would be call-outs on alleged misbehavior, though some are just questions.
posted by nangar at 5:28 AM on April 15, 2010
posted by nangar at 5:28 AM on April 15, 2010
This calls for careful analysis -- if patterns can be determined, the mods could institute "controlled burns" under fake user names to bleed off the tension with minor spats before it gets to the "volcano under the glacier" stage. Perhaps judicious use of callouts could spark the necessary catharsis without triggering an event of 7.3 or higher on the GRAR scale.
posted by GenjiandProust at 5:31 AM on April 15, 2010
posted by GenjiandProust at 5:31 AM on April 15, 2010
Over at 750 words they use some external statistical dingus to evaluate what you've written each day and graph it along several axis(positive/negative emotional state, etc). It's neat. I'm too iphoned to look up right now, but I know the proprietor is a mefite.
It could be cool to run MetaFilter through that sort of tool, and match the resulting graph to academic schedules, lunar cycles, LOST episodes, etc.
Deleted comments present a blind spot to such analysis, but there's nothing that can be done about it.
posted by dirtdirt at 5:32 AM on April 15, 2010
It could be cool to run MetaFilter through that sort of tool, and match the resulting graph to academic schedules, lunar cycles, LOST episodes, etc.
Deleted comments present a blind spot to such analysis, but there's nothing that can be done about it.
posted by dirtdirt at 5:32 AM on April 15, 2010
Do you think rate of comment deletions could function as a proxy for the rate and extent of threads turning all GRAR-y?
I'm waiting for one of you math folks to put some formula in here that starts with something like:
If G{GRAR}>dS{donut_supply/cortex}: then...
Only yours might actually make sense and have some cool symbols in it I won't understand.
posted by marxchivist at 6:02 AM on April 15, 2010
I'm waiting for one of you math folks to put some formula in here that starts with something like:
If G{GRAR}>dS{donut_supply/cortex}: then...
Only yours might actually make sense and have some cool symbols in it I won't understand.
posted by marxchivist at 6:02 AM on April 15, 2010
Seems plausible. My contributions on here have been much less contentious since I graduated.
posted by invitapriore at 6:05 AM on April 15, 2010
posted by invitapriore at 6:05 AM on April 15, 2010
if patterns can be determined, the mods could institute "controlled burns" under fake user names to bleed off the tension with minor spats before it gets to the "volcano under the glacier" stage.
Hell with that. The Pre-Mods could give time-outs to people who are Pre-Flagged.
posted by pracowity at 6:57 AM on April 15, 2010
Hell with that. The Pre-Mods could give time-outs to people who are Pre-Flagged.
posted by pracowity at 6:57 AM on April 15, 2010
if patterns can be determined, the mods could institute "controlled burns" under fake user names to bleed off the tension with minor spats before it gets to the "volcano under the glacier" stage.
Ok, so we're looking for people complaining about cycling, stats or libraries, got it.
posted by Brandon Blatcher at 6:59 AM on April 15, 2010
Ok, so we're looking for people complaining about cycling, stats or libraries, got it.
posted by Brandon Blatcher at 6:59 AM on April 15, 2010
(though it occurs to me one could just look for gaps in the comment ID sequence to count deletions that way, as a raw number, though I'm not sure how reliable that is. Or how good an idea that is, either.)
Yeah, that's the best you could do with the Infodump on that front. It would indeed be reliable as far as detecting rate of deletions—I don't think there's any other mechanism that can cause gaps—so as a quick one-off attempt it might be fun. All it would tell you is that A Comment Was Deleted From Subsite X though, not by whom or from which thread.
It might be interesting to graph some additional stuff against time that I'd have to do myself, though, with db access: analysis of raw deletions-per-day rate vs. users-with-deletions (to control for one person having a freakout) vs. threads-with-deletions (to control for an outlier thread where a lot gets removed), correlate that additionally against flagging data in similar fashions.
I'm suspicious about it turning out anything strongly conclusive about a specific cycle; like metaBugs argues, I suspect any build-up-and-crest cycle is more aperiodic than that, if it can be characterized as a proper cycle at all rather than just a random walk on the Tufteian sparkline that is GRAR.
But it could be fun to go looking.
posted by cortex (staff) at 7:05 AM on April 15, 2010
Yeah, that's the best you could do with the Infodump on that front. It would indeed be reliable as far as detecting rate of deletions—I don't think there's any other mechanism that can cause gaps—so as a quick one-off attempt it might be fun. All it would tell you is that A Comment Was Deleted From Subsite X though, not by whom or from which thread.
It might be interesting to graph some additional stuff against time that I'd have to do myself, though, with db access: analysis of raw deletions-per-day rate vs. users-with-deletions (to control for one person having a freakout) vs. threads-with-deletions (to control for an outlier thread where a lot gets removed), correlate that additionally against flagging data in similar fashions.
I'm suspicious about it turning out anything strongly conclusive about a specific cycle; like metaBugs argues, I suspect any build-up-and-crest cycle is more aperiodic than that, if it can be characterized as a proper cycle at all rather than just a random walk on the Tufteian sparkline that is GRAR.
But it could be fun to go looking.
posted by cortex (staff) at 7:05 AM on April 15, 2010
patterns and seasonality, always a fascinating topic. pity we're throwing a 'blowing off steam' party starting now for those preparing for the end of term next week ;p else I'd stay on and blather
posted by infini at 7:16 AM on April 15, 2010
posted by infini at 7:16 AM on April 15, 2010
Everyone does seems a bit sensitive lately. It’s like going home for the holidays but I get to do it every day and from the comfort of my own home.
posted by victoriab at 7:22 AM on April 15, 2010
posted by victoriab at 7:22 AM on April 15, 2010
I work in an academic setting and, as I watch the cycles of MetaFilter mood,
I work in a retail setting, so I'm just angry all the time.
posted by jonmc at 7:24 AM on April 15, 2010 [7 favorites]
I work in a retail setting, so I'm just angry all the time.
posted by jonmc at 7:24 AM on April 15, 2010 [7 favorites]
I work in a maintenance setting, so I just steal the candies out of the dish on your desk after you've left for the day.
posted by Alvy Ampersand at 7:30 AM on April 15, 2010 [2 favorites]
posted by Alvy Ampersand at 7:30 AM on April 15, 2010 [2 favorites]
I work in a corporate setting but it's too far to cycle, even if I was in the mood.
posted by tellurian at 7:31 AM on April 15, 2010
posted by tellurian at 7:31 AM on April 15, 2010
Bipolar disorder is not "moodiness." It's a serious illness.
Yes, I know. However much I may appear to be, I am not a complete idiot. </moodiness>
posted by octobersurprise at 7:32 AM on April 15, 2010 [2 favorites]
Yes, I know. However much I may appear to be, I am not a complete idiot. </moodiness>
posted by octobersurprise at 7:32 AM on April 15, 2010 [2 favorites]
Cortex, throw in account closings as another factor. Maybe offset in time a bit, since GRAR proceeds Big Red Button punching. A trailing indicator, if you will.
Perhaps tracking a rise and fall in the relative frequency of Carlin's Seven Dirty Words might also be good. Do people start cussin' up a storm just prior to a GRAR?
posted by adipocere at 7:34 AM on April 15, 2010
Perhaps tracking a rise and fall in the relative frequency of Carlin's Seven Dirty Words might also be good. Do people start cussin' up a storm just prior to a GRAR?
posted by adipocere at 7:34 AM on April 15, 2010
Clearly it's time for the Great MetaFilter Back Tagging Project (rev 2.0)!! We need thousands of volunteers to pore through the archives, read every comment, and tag each thread with a scientifically accurate mood tag for infodump posterity.
posted by The Winsome Parker Lewis at 7:42 AM on April 15, 2010 [2 favorites]
posted by The Winsome Parker Lewis at 7:42 AM on April 15, 2010 [2 favorites]
ponders deeply : could it be a side effect of volcanic activity?
posted by infini at 7:44 AM on April 15, 2010
posted by infini at 7:44 AM on April 15, 2010
I thought people were being unusually negative about everything earlier this week but now it seems to have calmed down. Maybe we all just needed a nap.
posted by The Whelk at 7:47 AM on April 15, 2010
posted by The Whelk at 7:47 AM on April 15, 2010
ponders deeply : could it be a side effect of volcanic activity?
That one thread did erupt into a glorious meltdown of cheese, creamy rice, and toppings.
posted by The Whelk at 7:47 AM on April 15, 2010 [1 favorite]
I thought people were being unusually negative about everything earlier this week but now it seems to have calmed down. Maybe we all just needed a nap.
it's all the Muppetry.
Wokka Wokka.
posted by zarq at 8:08 AM on April 15, 2010 [2 favorites]
it's all the Muppetry.
Wokka Wokka.
posted by zarq at 8:08 AM on April 15, 2010 [2 favorites]
Rate of flagging? Could the overall flagging rate - as opposed to the hard deletion rate - be a good indicator of community mood? Or does flagging spike when everyone is all schmoopy also? Does flagging spike at all? Or is it just a constant trickle/stream of yay/grar?
posted by greekphilosophy at 8:09 AM on April 15, 2010
posted by greekphilosophy at 8:09 AM on April 15, 2010
it's all the Muppetry.
Well stick a hand up my ass and call me Fozzy Bear!
wait ...no
posted by The Whelk at 8:12 AM on April 15, 2010 [1 favorite]
Well stick a hand up my ass and call me Fozzy Bear!
wait ...no
posted by The Whelk at 8:12 AM on April 15, 2010 [1 favorite]
Well stick a hand up my ass and call me Fozzy Bear!
That's Rrowlfing.
Or so I've heard.
We know what Furries are. Are Muppet Fetishists known as Fuzzies?
posted by zarq at 8:15 AM on April 15, 2010
That's Rrowlfing.
Or so I've heard.
We know what Furries are. Are Muppet Fetishists known as Fuzzies?
posted by zarq at 8:15 AM on April 15, 2010
I'll give that comment ID number gap analysis stuff a try tonight, and see what sort of hideous charts I can produce. In the mean time, maybe we should decide what the SI Metric unit of GRAR should be. We could name it after the GRARiest person in history and define their GRAR level as 1.0, and use that as a calibration standard for our GRARometers.
No?
posted by FishBike at 8:33 AM on April 15, 2010
No?
posted by FishBike at 8:33 AM on April 15, 2010
But I've been in a spectactularly good mood all week!
dammit, I can't do anything right!
[mood plummets...]
posted by MexicanYenta at 8:36 AM on April 15, 2010
dammit, I can't do anything right!
[mood plummets...]
posted by MexicanYenta at 8:36 AM on April 15, 2010
We could name it after the GRARiest person in history and define their GRAR level as 1.0, and use that as a calibration standard for our GRARometers.
Maybe go in the other direction: identify the atomic level of GRAR, the smallest unit that can be identified, and proceed from there in kilo- and mega- and so forth modifiers to describe the larger episodes.
posted by cortex (staff) at 8:43 AM on April 15, 2010
Maybe go in the other direction: identify the atomic level of GRAR, the smallest unit that can be identified, and proceed from there in kilo- and mega- and so forth modifiers to describe the larger episodes.
posted by cortex (staff) at 8:43 AM on April 15, 2010
Since it's SI Metric I'm after, maybe derivation from base units is the way to go? Define 1.0 GRAR as the amount of annoyance caused by having 1 kg of water poured on your head from a height of 1 meter in 1 second? Well, not your head, specifically, cortex. Maybe, like, Newton's head.
posted by FishBike at 8:50 AM on April 15, 2010 [1 favorite]
posted by FishBike at 8:50 AM on April 15, 2010 [1 favorite]
We could name it after the GRARiest person in history and define their GRAR level as 1.0
You'll need to form a GRAR board and implement GRAR standards.
posted by Brandon Blatcher at 8:57 AM on April 15, 2010
You'll need to form a GRAR board and implement GRAR standards.
posted by Brandon Blatcher at 8:57 AM on April 15, 2010
We need thousands of volunteers to pore through the archives, read every comment, and tag each thread with a scientifically accurate mood tag for infodump posterity.
Why stop there? If we're going to read every comment, they should be color-coded. Done correctly, this would remove the need for most comment deletions as the more spiteful/contentious they get, the harder they'd be to read.
posted by philip-random at 9:10 AM on April 15, 2010
Why stop there? If we're going to read every comment, they should be color-coded. Done correctly, this would remove the need for most comment deletions as the more spiteful/contentious they get, the harder they'd be to read.
posted by philip-random at 9:10 AM on April 15, 2010
Once you guys figure out the metrics, pony request: Auto-permaban anybody whose GRAR quotient exceeds a predetermined excessive-GRAR threshold and sustains said GRAR level for seven days. On the user's profile, instead of saying "this account is disabled" label it "this plate of beans has spoiled."
posted by The Winsome Parker Lewis at 9:10 AM on April 15, 2010
posted by The Winsome Parker Lewis at 9:10 AM on April 15, 2010
man, you guys are skittling the hell out of this car.
posted by The Whelk at 9:19 AM on April 15, 2010
posted by The Whelk at 9:19 AM on April 15, 2010
I think we should use known gold standards as the base units for measuring user's posting history mood.
You comment history analysis ranks you as:
.44 Optimus Chymes of snark
.36 Mikos of astuteness
.55 Whelks of whimsy
.03 Sixcolors of freakout
.05 DirtyNumbAngelBoys of GRAR
.13 ParisParamuses of troll
posted by Babblesort at 9:29 AM on April 15, 2010 [2 favorites]
You comment history analysis ranks you as:
.44 Optimus Chymes of snark
.36 Mikos of astuteness
.55 Whelks of whimsy
.03 Sixcolors of freakout
.05 DirtyNumbAngelBoys of GRAR
.13 ParisParamuses of troll
posted by Babblesort at 9:29 AM on April 15, 2010 [2 favorites]
Hmm. Let me check the Phil Collins Index...
Yep. I can definitely feel something.
posted by It's Raining Florence Henderson at 9:29 AM on April 15, 2010 [2 favorites]
Yep. I can definitely feel something.
posted by It's Raining Florence Henderson at 9:29 AM on April 15, 2010 [2 favorites]
You're no statistical analysis collaborator, you're no statistical analysis collaborator of mine.
posted by cortex (staff) at 9:30 AM on April 15, 2010
posted by cortex (staff) at 9:30 AM on April 15, 2010
Whelks of whimsy
This sounds vaguely like a Lite Psyche band circa '66.
posted by jonmc at 9:36 AM on April 15, 2010 [1 favorite]
This sounds vaguely like a Lite Psyche band circa '66.
posted by jonmc at 9:36 AM on April 15, 2010 [1 favorite]
Ha! Babblesort attempts to measure comment cromularity without defining standard fishpants units! Rookie!
posted by Mister_A at 10:30 AM on April 15, 2010 [2 favorites]
posted by Mister_A at 10:30 AM on April 15, 2010 [2 favorites]
conversely, having things to do in college keeps you busy and from making posts in metafilter. I think my rate of contributions has gone down in the past 5yrs and mefi is probably better for it : )
In terms of Grar - I have much Grar to give - an art college degree show is not the most mellow of things to prepare for.
Fortunately there are many fabulous stars in my class and completely unbiased tutors I get to interact with every day. Otherwise, it would be so much more difficult.
posted by sgt.serenity at 11:29 AM on April 15, 2010
In terms of Grar - I have much Grar to give - an art college degree show is not the most mellow of things to prepare for.
Fortunately there are many fabulous stars in my class and completely unbiased tutors I get to interact with every day. Otherwise, it would be so much more difficult.
posted by sgt.serenity at 11:29 AM on April 15, 2010
In terms of Grar -
GRAR and GRAR-y are such cutesie wootsie Hello Kitty foofy lil' ol' expressions.
Just sayin'...
posted by y2karl at 12:01 PM on April 15, 2010
GRAR and GRAR-y are such cutesie wootsie Hello Kitty foofy lil' ol' expressions.
Just sayin'...
posted by y2karl at 12:01 PM on April 15, 2010
The Effects Of Leaning Like A Cholo On My MetaTalk Posting Habits
posted by Damn That Television at 12:19 PM on April 15, 2010
posted by Damn That Television at 12:19 PM on April 15, 2010
Personally, flapjaxatmidnite's comment, thought I have the impression it's semi-tongue-in-cheek, rings pretty true to me. Now that I'm out of school I am a severely less horrible person in a lot of ways. Obviously you give up a lot of other worthwhile dimensions though; I'm not knocking academics, honest. Just. For certain kinds of Type-A overwrought human beings (raises hand) wow, adding high-stakes school to the mix leads to major fucked up personality disorder. Whoo boy. I'm still celebrating how much more at ease I am, and nicer/less self-absorbed now that it's been years since I was in school...(and yes, I do also miss it terribly at times, so it evens out).
posted by ifjuly at 12:22 PM on April 15, 2010
posted by ifjuly at 12:22 PM on April 15, 2010
GRAR and GRAR-y are such cutesie wootsie Hello Kitty foofy lil' ol' expressions.
He is pretty cute
posted by The Whelk at 12:22 PM on April 15, 2010
He is pretty cute
posted by The Whelk at 12:22 PM on April 15, 2010
SEE! See right there. Total whimsy. HA! Try to deny it now.
posted by Babblesort at 12:25 PM on April 15, 2010
posted by Babblesort at 12:25 PM on April 15, 2010
*this* close to live-blogging a saffron risotto people
posted by The Whelk at 2:03 PM on April 15, 2010
posted by The Whelk at 2:03 PM on April 15, 2010
funny, I think I'm a more horrible person now that I'm out of school. But I'm not sure school is a factor, I think it's paying more attention to politics. Rest of comment deleted for excessive GRAR.
posted by jacalata at 2:53 PM on April 15, 2010
posted by jacalata at 2:53 PM on April 15, 2010
Has there been an infodump study of MetaFilter mood versus academic cycles?
Which academic cycle? Yours, mine, someone else's? You're assuming everyone here is doing the same thing at the same time, we're not.
The one thing that's going to affect how grumpy people seem to be here more than anything else is your own mood. When I'm having a bad day I see GRAR everywhere and react accordingly. But when I'm not I give people the benefit of the doubt and assume they're joking or being awkward rather than bitchy, and when I'm cheerful I'm often totally confused by people mentioning how irritable everyone else is. Plus if I'm cheerful I'm much more likely to skim over or ignore the shitty threads, for example I had a really productive week so just rolled my eyes at the buddism callout and didn't bother reading it, leaving me with a much more positive view of the site in general.
So yes each user will see correlation between how grumpy the site is and how difficult their life is at that time but it doesn't mean there's any causation from life's difficulties to the wider mood here, particularly when your life difficulties aren't shared by everyone in the first place.
posted by shelleycat at 3:39 PM on April 15, 2010
Which academic cycle? Yours, mine, someone else's? You're assuming everyone here is doing the same thing at the same time, we're not.
The one thing that's going to affect how grumpy people seem to be here more than anything else is your own mood. When I'm having a bad day I see GRAR everywhere and react accordingly. But when I'm not I give people the benefit of the doubt and assume they're joking or being awkward rather than bitchy, and when I'm cheerful I'm often totally confused by people mentioning how irritable everyone else is. Plus if I'm cheerful I'm much more likely to skim over or ignore the shitty threads, for example I had a really productive week so just rolled my eyes at the buddism callout and didn't bother reading it, leaving me with a much more positive view of the site in general.
So yes each user will see correlation between how grumpy the site is and how difficult their life is at that time but it doesn't mean there's any causation from life's difficulties to the wider mood here, particularly when your life difficulties aren't shared by everyone in the first place.
posted by shelleycat at 3:39 PM on April 15, 2010
Right then, as promised, here are some hideous charts of comment deletion rates, as determined by gaps in the comment ID sequence. I made two charts, one which shows the number of missing comments per month, and a second that shows the percentage of comments which are missing. Both charts have one line per sub-site.
I'm using the word "missing" rather than "deleted" because, as I kind of expected, there are other reasons for gaps in the comment numbering sequence besides deletion. For example, the biggest gap I could find was almost 200 consecutive missing comments, and it coincides with the technical difficulties mentioned in this MeTa post.
Similarly, there are some really wacky missing comment counts in earlier years, like in some cases 8,000 missing comments in a month, or 75% of all comment IDs missing for a month. I let the lines for these early years bust right through the top of the chart to avoid compressing the more recent data down to a flat line at the bottom.
This doesn't really seem right. So I'm not sure the correlation of missing comment IDs to deleted comments is a very good one after all.
posted by FishBike at 6:01 PM on April 15, 2010 [1 favorite]
I'm using the word "missing" rather than "deleted" because, as I kind of expected, there are other reasons for gaps in the comment numbering sequence besides deletion. For example, the biggest gap I could find was almost 200 consecutive missing comments, and it coincides with the technical difficulties mentioned in this MeTa post.
Similarly, there are some really wacky missing comment counts in earlier years, like in some cases 8,000 missing comments in a month, or 75% of all comment IDs missing for a month. I let the lines for these early years bust right through the top of the chart to avoid compressing the more recent data down to a flat line at the bottom.
This doesn't really seem right. So I'm not sure the correlation of missing comment IDs to deleted comments is a very good one after all.
posted by FishBike at 6:01 PM on April 15, 2010 [1 favorite]
Yeah, there's a lot of crazy noise in there. The horrid spikes on meta and askme between 2003 and 2005 is almost certainly an artifact of the shared-table nature of those two sites in the early part of askme's life; I think you could repair that weirdness manually by taking that problem into account (search the union of metatalk and askme comment ids in that weird period for gaps, I guess), but you may still lose info about which site the deletions were coming from.
I should have been more precise earlier; there's no routine, current mechanism that would lead to skipped non-deleted comments, but going back in time there will no doubt be problems. That's one big one. Another one, early on, is Matt tweaking the site and doing a fair amount of test-and-delete stuff, which probably accounts for a lot of the crazy spikiness in the rate-of-deletion chart early on.
posted by cortex (staff) at 6:09 PM on April 15, 2010
I should have been more precise earlier; there's no routine, current mechanism that would lead to skipped non-deleted comments, but going back in time there will no doubt be problems. That's one big one. Another one, early on, is Matt tweaking the site and doing a fair amount of test-and-delete stuff, which probably accounts for a lot of the crazy spikiness in the rate-of-deletion chart early on.
posted by cortex (staff) at 6:09 PM on April 15, 2010
I think you could repair that weirdness manually by taking that problem into account (search the union of metatalk and askme comment ids in that weird period for gaps, I guess), but you may still lose info about which site the deletions were coming from.
What I'll probably do with this is represent the shared table period as a 5th sub-site, something like "MeTa+AskMe". Then we will be able to see that line split in two when the data for the two sub-sites was separated. Do we know exactly when that happened? I suppose it's easy to figure out by looking for the first MeTa comment ID that also exists in AskMe.
It's also pretty clear there can be technical problems that result in ID gaps. If the transaction to write a comment to the database has to be rolled back after the next ID counter has already been incremented, the counter itself doesn't get reset, but the comment doesn't appear in the table either. It pretty much has to work that way or else all database transactions would get serialized by the counter update.
posted by FishBike at 6:29 PM on April 15, 2010
What I'll probably do with this is represent the shared table period as a 5th sub-site, something like "MeTa+AskMe". Then we will be able to see that line split in two when the data for the two sub-sites was separated. Do we know exactly when that happened? I suppose it's easy to figure out by looking for the first MeTa comment ID that also exists in AskMe.
It's also pretty clear there can be technical problems that result in ID gaps. If the transaction to write a comment to the database has to be rolled back after the next ID counter has already been incremented, the counter itself doesn't get reset, but the comment doesn't appear in the table either. It pretty much has to work that way or else all database transactions would get serialized by the counter update.
posted by FishBike at 6:29 PM on April 15, 2010
It looks like a big spike happens on Ask/Meta right around the 2004 elections. 9/11, too.
posted by invitapriore at 6:55 PM on April 15, 2010
posted by invitapriore at 6:55 PM on April 15, 2010
I don't know the exact timing, but I know I've nailed it down in the past. It may be noted on the infodump page of the wiki or on the timeline page.
posted by cortex (staff) at 7:05 PM on April 15, 2010
posted by cortex (staff) at 7:05 PM on April 15, 2010
Ok, I've updated the charts to take into account the weirdness with comment IDs when AskMe and MeTa were sharing a table. There's a "pre-AskMe" line for MeTa before AskMe even exist, a short "AskMe+MeTa" line for the period when they were combined, and then the usual separate AskMe and MeTa lines after they split.
This definitely de-weirded the charts somewhat, and I changed the scale on the percentage one since most of it now fits under 10%. There's still some suspiciously big spikes. invitapriore's observation of the timing of these makes me think they might be right, though. If so, what happened in January 2010 in AskMe?
posted by FishBike at 7:07 PM on April 15, 2010 [1 favorite]
This definitely de-weirded the charts somewhat, and I changed the scale on the percentage one since most of it now fits under 10%. There's still some suspiciously big spikes. invitapriore's observation of the timing of these makes me think they might be right, though. If so, what happened in January 2010 in AskMe?
posted by FishBike at 7:07 PM on April 15, 2010 [1 favorite]
I don't know the exact timing, but I know I've nailed it down in the past. It may be noted on the infodump page of the wiki or on the timeline page.
I just queried for the first AskMe comment ID, and the first non-unique comment ID between AskMe and MeTa, to get the start and end of the table sharing range of comment IDs.
posted by FishBike at 7:08 PM on April 15, 2010
I just queried for the first AskMe comment ID, and the first non-unique comment ID between AskMe and MeTa, to get the start and end of the table sharing range of comment IDs.
posted by FishBike at 7:08 PM on April 15, 2010
(Maybe answered my own question: in January there was this AskMe question about infidelity and a nearly 700 comment MeTa thread about it. I guess you guys probably deleted a lot of comments from there, yeah? If so, there's a bit of anecdata for the deletions=GRAR hypothesis.)
posted by FishBike at 7:23 PM on April 15, 2010
posted by FishBike at 7:23 PM on April 15, 2010
Don't forget there was a server hiccup earlier this year that led to some Ask MeFi comments going missing - I don't know if those comment IDs were recycled?
posted by subbes at 8:16 PM on April 15, 2010
posted by subbes at 8:16 PM on April 15, 2010
Yeah, there may have been some weirdness in late January with the SQL hack thing, but I don't remember exactly what happened as far as the indices. But that's not a bad lead.
There wasn't much removed from the infidelity askme, fwiw.
posted by cortex (staff) at 9:04 PM on April 15, 2010
There wasn't much removed from the infidelity askme, fwiw.
posted by cortex (staff) at 9:04 PM on April 15, 2010
Morale of the story: We need to start getting more comments deleted in Music. It looks so _dead_ there lying all flat at the bottom like that...
posted by kaibutsu at 10:04 AM on April 16, 2010
posted by kaibutsu at 10:04 AM on April 16, 2010
Morale of the story: We need to start getting more comments deleted in Music. It looks so _dead_ there lying all flat at the bottom like that...
Ukuleles! Dumb instrument or dumbest instrument? Also, covers! Isn't that just lazy?
posted by The Whelk at 10:11 AM on April 16, 2010
Ukuleles! Dumb instrument or dumbest instrument? Also, covers! Isn't that just lazy?
posted by The Whelk at 10:11 AM on April 16, 2010
I could do an analysis of which tags are most frequently associated with deleted posts, if people are looking for some data to guide their efforts to maximize their deletion-worthy comment count.
On the other hand, this is an interesting problem in that if your comments are excessively deletion-worthy, you might just be banned instead. So it's a non-trivial problem to maximize total GRAR contributed to the site.
Perhaps someone should fund a study to investigate this further. Actually... the prior existence of such a study might explain the behavior of a small subset of the site's users.
posted by FishBike at 10:39 AM on April 16, 2010
On the other hand, this is an interesting problem in that if your comments are excessively deletion-worthy, you might just be banned instead. So it's a non-trivial problem to maximize total GRAR contributed to the site.
Perhaps someone should fund a study to investigate this further. Actually... the prior existence of such a study might explain the behavior of a small subset of the site's users.
posted by FishBike at 10:39 AM on April 16, 2010
You are not logged in, either login or create an account to post comments
posted by philip-random at 1:41 AM on April 15, 2010