The Water Cooler

What are the top two or three challenges facing the site as it moves forward, and what do you plan to do as a moderator to address them? What have you already done?

enderland

@RhysW yeah that's the trick eh? if you are wasting time anyways, might as well work :)

RhysW

so if i had control over my down time, id be really efficient

@enderland exactly

Monica Cellio

^^^ This question (which, based on current voting, seems likely to make the list) asks more generally about challenges. Hot questions are likely to come up there, right?

RhysW

if i could do 3 hour stints, with 1 hour inbetween each i would get so much done

Monica Cellio

So I'm just saying that another more-general question can address what you want to know while also providing room for people to talk about other issues.

RhysW

2:10 PM

no ones willing to give me that sort of work hours though :P

gnat

@MonicaCellio I see, thanks! Upon re-reading, you're right (missed that only because it didn't have word "hot" in it). It indeed makes better sense to remove in favor of your question

Monica Cellio

@gnat you don't have to remove it if you don't want to; let the community decide. I was just addressing the voting.

6 hours ago, by RhysW

@jmac Perfect! finds plane tickets

@RhysW hey if you're buying plane tickets anyway, they probably cost the same in both directions. So, y'know, if you're tired of jet-setting, you could buy me a ticket to come to you. :-)

gnat

@MonicaCellio we have a rule "don't repeat others", I want to follow it :)

Monica Cellio

@gnat fair enough. :-)

RhysW

@MonicaCellio but, but, the point of my flying out was to make you nominate yourself, if youre willing to come here, that's like saying 'i give up, ill come to you to make it easier for you to make me nominate myself' which really just sort of breaks all of the logic i have

Monica Cellio

2:21 PM

@RhysW think of it more like me coming for the on-site interview. Tell me why I should work for you, and all that. Besides, depending on what persuasion techniques you had in mind, remember that you're going to have trouble getting that baseball bat through airport security if you come here. :-)

RhysW

Oh good point, If you come to me I can have my persuasion tools all prepared

Monica Cellio

@RhysW exactly. And a nice lunch. And good beer. :-)

Sorry Chad, that was a mis-fire on my part.

Alex M.

nobody wants to visit me :(

RhysW

the removed messages around these messages makes this seem even funnier

Alex M.

it's not all vampires here in Transylvania you know

just... mostly

Chad

2:23 PM

@MonicaCellio lol its all good

RhysW

I never know where the line is on jokes with people i don't know well :P

enderland

oh c'mon its @Chad

;)

RhysW

who is that directed at enderland? :P

enderland

7 mins ago, by RhysW

I never know where the line is on jokes with people i don't know well :P

RhysW

ahh

im sure @Chad loves all my jokes :P

enderland

2:39 PM

@RhysW I thought you'd be intimately familiar with bad jokes by this point in your life ;)

RhysW

bah, you guys are just jealous of my hilarity!

@Chad you can keep your jokes around you know, i dont take insult to them :P

gnat

-3

Q: Should criminal activity be tolerated?

This belongs in Meta, but for some reason I'm allowed to litter up the main forum with off topic questions but not have concerns regarding site guidelines and enforcement addressed in the proper section. In fact, this isn't even a question, it is an order. Word your guidelines better. The member...

reputation

^^^ does Workplace want it? I can flag for migration :)

RhysW

I hope he realizes that the internet is about as lawful as international waters

Alex M.

criminal activity

bahahahahah

random mod pops in

"STOP! YOU VIOLATED THE LAW!"

RhysW

Oded is no mod, he is a community manager

and he feasts on posts like these

they are his life source

Alex M.

2:46 PM

I was not referring to Oded

it was just an hypothetical situation

ahh

Hi Will, welcome to The Workplace. While the resource you linked in your answer may be immensely helpful, please include the portions of it relevant to this question in your answer as well. Links may change, rendering your answer obsolete to future readers. — CMW 1 min ago

maybe a promotional account, maybe not. Let's see if they do something about their answer

gnat

@CMW at Programmers, ChrisF taught me: unregistered account + late answer + promotional content => flag as spam. I wouldn't say it's totally bulletproof heuristics, as I had spam flag submitted that way declined once... one time of 100-200 total flags like that :)

CMW

It's probably the case, and I tend to agree. Flagged it as VLQ anyway, as it's not a link-only answer

gnat

3:28 PM

"what a hot SO question! I can't stop myself from upvoting" stackoverflow.com/questions/22683634/hash-code-is-always-same

Chad

3:47 PM

Great question gnat... while I dont understand hash values all of those answers seem reasonable so deserve an upvote

CMW

4:07 PM

This question appears to be off-topic because it is about software recommendations and billing and not about navigating the workplace. — CMW 15 secs ago

gnat

@Chad wow, that sounds very much like a typical passer by coming to Workplace questions from hot list! :)

RhysW

6:37 PM

@Chad hahah

Monica Cellio

I think for hot questions (and non-hot ones that attract junk) we need to rally more people to downvote the crap into oblivion until it can be deleted, and for users with delete votes to cast them promptly.

Short-term this depends on mods, but long-term it shouldn't. And we should do everything we can as users now to help out.

Chad

@MonicaCellio It depends if the question get 10 crap answers in those few hours then yes i think there is damage.

Monica Cellio

@Chad does it? I mean yeah, we don't want crap, but if the crap is downvoted does it have to be deleted right this minute?

Chad

@MonicaCellio Maybe I should have qualified that as 10 crap answers that are feel good so they end up highly voted

Monica Cellio

Could you elaborate on the damage? Again, I agree they need to be flushed; I don't want crap on the site any more than you do. I'm trying to understand the 24x7-urgency angle.

@Chad ahh.

Chad

6:46 PM

"Your boss is a Jerk you should quit!" +1 I agreee

Monica Cellio

And with hot questions the theory is that those votes come from outside our community?

Chad

alot of them do... i am sure some are here too

When they get 10 votes in under an hour then I think yea they probably do

Monica Cellio

This seems like it would help a lot:

18

Q: Auto-protect questions that get more than N answers from new users in a 24-hour period

Expanding the criteria for auto-protection has been discussed in the past: Should we automatically protect all questions with more than N answers? But I think that discussion addresses the wrong problem: Protecting a question with a large number of answers doesn't do anything to fix the probl...

feature-request new-users protected-questions

Shog9

@Chad open to analysis that would prove otherwise, but so far what I've seen is that outsiders tend to vote along the same lines as insiders

IOW, stuff ends up getting outside attention because the folks on-site liked it a lot.

Chad

@Shog9 Maybe you could fix it by only counting votes by people with more than 500 rep :p

Shog9

6:53 PM

heh

yeah, that'd get rid of a lot of votes

Monica Cellio

@Shog9 well, one factor is that people who aren't already active here don't have downvotes, only upvotes, so if they vote at all it'll be up.

Shog9

@MonicaCellio well, a good portion of outsiders don't have either

Monica Cellio

I've sometimes found myself really wanting a downvote on a site where all I have is the association bonus but I can tell an answer is crap.

Shog9

but yeah, those that do will tend to have upvotes

Monica Cellio

@Shog9 true outsiders don't, but people coming from other sites (which is probably a lot of the hot traffic, no?) are more likely to have 101.

Shog9

6:56 PM

← 23 messages moved from Workplace 2014 Election

@MonicaCellio some of them will. A pretty big chunk of folks on other sites don't have any rep though either

RhysW

@MonicaCellio yes, i agree with you there!

enderland

@MonicaCellio might as well just UV anyways

Monica Cellio

@Shog9 do we have any data on hot-question traffic -- what sites they come from, of those with local accounts what rep category, etc?

Shog9

@MonicaCellio mmm... Some

Monica Cellio

So (thinking out loud here) there seem to be two types of concerns with hot questions: people posting crap answers, which protection would impede (at some cost), and what we'll call inappropriate upvotes (Chad's "my boss sucks - +1 I agree" scenario).

The votes are a problem because (a) they distort the page view and (b) they prevent delete votes. We could look at existing data to address the "how much would protection have helped?" question, but votes are secret so we can only look at who could be voting.

That means wanting to break hot-question link-followers into a few groups: no account (or doesn't log in) so can't affect anything, account but can't vote, can upvote but not downvote, and can vote freely. If we knew that, would it actually help us?

(Those first two groups might be equivalent for our purposes.)

gnat

7:16 PM

@MonicaCellio you need to better define "oblivion". To stop an additional answer of pushing question up the list, one needs Qscore/5 votes, in a +20 question this means less than 4 downvotes will keep lemming ride. Also, account for sympathy upvotes from passers by (shuffling questions at sidebar tamed that but not completely) - 2-3 upvotes to crappy answer ("just for fun" "just for trying") will motivate answerer to keep it rep-wise

5

Q: Do answers that attract both positive and negative votes benefit the community?

I gave this answer to a question. It rapidly attracted a significant number of negative votes and negative comments, to the point where my first reaction was to delete the answer. I could improve it of course, and I did make some effort. What surprised me is that despite getting so many negative...

discussion reputation answers voting down-votes

12

A: The association bonus should not enable users to vote on every site

In order to tame the damage of drive-by votes in hot questions, I would consider more targeted approach. I think the most straightforward way to prevent over-voting from newcomers who just arrived at the hot question would be to simply delay granting them upvote privilege for a day or two. This...

Monica Cellio

@gnat that's assuming that rep is a significant motivation for the people posting these answers. It is for people who are (or want to be) here, but do drive-by posters care? Anyway, by "downvote into oblivion" I mean grayed-out and below (at least -3 net). This means more people need to downvote.

gnat

> What surprised me is that despite getting so many negatives, I also seem to be getting a number of positives, to the point where my net reputation on this answer is +66. Deleting the question would be a net loss.

^^^ you see totally reasonable user hesitates to delete because of rep brought in by drive-by upvotes

Monica Cellio

@gnat that Shirky article has a lot of good observations in it.

@gnat I see an anecdote, and maybe a common one but we can't tell. I'm not saying it's not a factor; I'm just saying that since SE isn't going to overhaul the voting system (though the "delay" idea has appeal), we have to work with what we have. So what can we do to either reduce the bogus upvotes or get more offsetting downvotes? Again, for obvious crap this is for a few hours until a mod can delete, and in six months I hope that'll be "an hour or so until the community can delete".

gnat

@MonicaCellio agree. my theory is, SO functioning is so stable because back then, Atwood thoroughly studied it. Conversely, younger sites suffer because they do their own trial and error instead of studying Shirky and applying it for their specific context

@MonicaCellio well, that very dreaded idea of 24/7 watch is the only thing that comes to mind, if one assumes usual indifference of SE team. I do something like that at Programmers, and as far as I can tell, this more or less works. It's not reliable of course, as single-person dependency. Bus factor and such...

Bus factor

In software development, a software project's bus factor (also known as truck factor, or bus/truck number or lottery factor) is a measurement of the concentration of information in individual team members. The bus factor is the total number of key developers who would need to be incapacitated (for example, by getting hit by a bus/truck) to send the project into such disarray that it would not be able to proceed. In this event, the project would retain information (such as source code or other systems) with which no remaining team member is familiar. A high bus factor, where many individuals...

the way I do is sort of an art and I doubt it'll scale. I watch the posts and use down/close votes, comments, flags the way that is likely to keep bad things from happening per my past experiments

the closest thing to what I do is 24/7 mod watch

Monica Cellio

@gnat I very strongly doubt that you will get any volunteer mod team to agree to maintain a rotation so that there is somebody manning the desk every instant of every day. The best we can hope for is that everybody pays attention and that we achieve some schedule diversity. The site is important, but 2 or 3 questions attracting crap answers between 4 and 6 AM local mod time isn't critical. This isn't "the service that everybody depends on for critical functions is down" critical.

gnat

7:29 PM

sorta perverse as they are supposed to be exception handlers, not over the clock watchers. But oh well

Monica Cellio

@gnat I very much appreciate your dedication to the site and to keeping quality up. Please keep doing that. It really helps. Thank you.

We need more people doing more of that. Delete votes are hard but more people have close votes and many more have downvotes.

And if someone can make an edit that fixes a problem, please make that edit! (Generic "you", not gnat in particular.)

gnat

@MonicaCellio next closest thing to 24/7 support duty is more or less regular schedule. Per my observations at Programmers, things may get troublesome when there's 4-hour gap. 2h gaps feel okay

can't tell how that will scale at Workplace though, due to higher exposure at hot list

but it somehow feels 2h gaps would be okay here too

Chad

@gnat I think that all we really need is a limited time of the watch, where the mods come out consistently closing and posting notices on answers that need improvement, and you will get the majority of the community falling in step...

Monica Cellio

@gnat I haven't been tracking Programmers. If their mods agreed to a "helpdesk rotation" sort of schedule, that's interesting but I suspect very unusual on the network. I don't think you'll get that here. We all, mods and non-mods, do our best, but sometimes there will be gaps and we'll just have to clean it up in the morning. That's life.

gnat

@MonicaCellio I think they didn't, but there's less need for that over there...

in Workplace 2014 Election, 28 mins ago, by gnat

@Shog9 well yeah, as far as I can tell they don't coordinate explicitly / intentionally, it's more like organically grown repeated pattern, somewhat loose. But then, Programmers are less (I think about twice less than Workplace) exposed at hot list, which makes it less important

in Workplace 2014 Election, 27 mins ago, by gnat

...and there are 5 mods at Programmers, not 3, which makes it even less important

Monica Cellio

7:38 PM

@gnat I think we disagree on the degree of the need. But no matter what happens with mods (who gets elected, their timezones, their typical schedules, etc), we need to help and encourage the community to act more on these so that it'll be less critical if it takes a mod a few hours to get to it.

Trust me, even with an ideal distribution, there are going to be times when it'll take a few hours to get to it -- people get sick, or go on vacation, or are all busy with the same holidays, or whatever.

gnat

sorry for sounding like broken record, but all we talk about here is just why I am not comfortable about having only 3 mods. I would rather have 4 and have someone step down later if things eventually get better (though I doubt this will happen anytime soon)

Monica Cellio

@gnat only the CMs can decide that and we've heard what sounds like a pretty firm ruling on that. It doesn't seem worth spending energy on wishing for what we don't have, just now. I strongly suspect that we'll be allowed another election to add to the team in a year; I've seen that on other sites. Maybe sooner than a year, depending on what happens.

gnat

@MonicaCellio I know. Next best thing would be then to have more than 3 (4, better 5) candidates so that next runner up after 3 elected can quickly step up if the need is proven. At Programmers, mapleshaft was promoted that way iirc

Monica Cellio

@gnat yeah, I expect we'll have more candidates before nominations close.

gnat

15

Q: Introduction to new community moderator

you may be familiar with me, but in case you aren't I am maple_shaft and I was asked to be the new community moderator. I am very excited to take this responsibility and I hope that I do not let anyone down. I believe in the purpose and scope of Programmers and wish to help guide it in the righ...

discussion moderators

Shog9

8:50 PM

@gnat if you want to stop sounding like a broken record, then stop talking and LISTEN. I'm telling you, based on several years of experience working with scores of different mod teams now, that NO ONE does what you're suggesting needs to happen.

It doesn't happen with 3 mods, it doesn't happen with 6 mods, and it doesn't happen with 16 mods.

It's a frickin' volunteer position - you do it when you have the time to do it.

3

Throwing more bodies at the wall doesn't change that.

If this site requires 24/7 vigilance in order to not fall apart at the seams - and I personally do not believe it does - then it either needs a thriving population of community moderators, or it needs a large, paid staff.

#1 will happen eventually. #2 won't.

gnat

9:13 PM

I totally don't understand what you are talking about, sorry

RhysW

@gnat 24/7 vigilance isnt needed

2

(is the heart of the speech i think)

and if 24/7 vigilance Is needed, then the site's in bad shape!

gnat

@RhysW sure, no problem. I didn't expect this to happen, my understanding is, we briefly discussed this in theory with Monica and moved on to more realistic things. Windmills, fighting?

RhysW

honestly i wasnt here for the whole conversation, im just summarising what shog said as you said you weren't sure

@MonicaCellio That last part makes quite a good metaphor!

Shog9

10:05 PM

@MonicaCellio let's take a step back here...

Defining "outsider votes" as votes cast by someone who hasn't yet earned even 15 reputation on the site at the time the vote was cast

And looking at all votes cast during the past 180 days

We'll define "outsider agreement" as a case where outsider votes on a post are > 0 and insider (folks who've earned >= 15 rep on the site) upvotes - insider downvotes are also > 0

We'll define "outsider disagreement" as a case where outsider votes on a post are > 0 and insider votes are < 0

Cases of outsider agreement: 2,680

Cases of outsider disagreement: 277

Monica Cellio

@Shog9 excellent! Thank you so much for bringing some real data to this discussion!

Shog9

Wait, there's more:

Cases where the insiders' votes would've brought a post's score < 0, but outsider votes were present in sufficient quantity to make the post's score positive: 147

Total posts voted on by outsiders in the past 180 days: 3870

Total posts voted on by insiders in the past 180 days: 7186

Total posts voted on by both groups in the past 180 days: 3053

Draw your own conclusions from this

Monica Cellio

Digesting -- thanks so much!

What is the difference between outsider disagreement (277) and insider negative but for outside votes (147)?

enderland

thanks for posting this @Shog9

Shog9

@MonicaCellio So... Imagine you & I see an answer & we both downvote it. It's at -2. Then 3 people come in with nothing but the association bonus and up-vote it. It's at +1. There are 147 cases like that

enderland

10:16 PM

I think it identifies that not 100% of the "lots of low quality crap" upvoted is from non-site regulars

Shog9

Now imagine @enderland and @RhysW both downvote the same answer, so it's at -1 again. There are 277 cases like that.

enderland

which means there are plenty of people HERE who upvote that crap too (I think that's a safe conclusion from this data)

Shog9

@enderland yes

enderland

which I guess should be a logical inference, from some of the low quality answers which still pick up upvotes even without hitting the hot questions list

hrmmm

Monica Cellio

@enderland sure looks that way to me. And while outsiders are contributing to crap-votes, they are a small part of the problem it appears.

enderland

10:19 PM

it'd be interesting to see something like, for users whose average post is below say 200 words (this is arbitrary) how often users like that upvote similar content, vs people who's average post is say 400 words (again, arbitrary, I don't think you can only do this)

Shog9

Oh, shoot... I forgot a really critical one:

Total insider votes during the past 180 days: 29,479

Total outsider votes during the past 180 days: 12,837

enderland

wow, that's impressive, actually

so a very good percentage of voting here IS from non-"regulars"

Monica Cellio

@Shog9 am I reading this right -- there were ~700 posts that only outsiders voted on (insiders didn't)?

Shog9

@MonicaCellio something like that

enderland

@MonicaCellio that's fascinating

Shog9

10:22 PM

I'm not breaking this down any further right now, but a really common case for that involves answers to questions being upvoted by the asker

Monica Cellio

@Shog9 ah.

enderland

thanks for this data @Shog9. I'm not sure how hard it is to generate but it's definitely interesting and very appreciated

Shog9

I doubt that's all of it, but at least on other sites you tend to get this really long tail where the only two people interested in voting are the ones asking and answering

Monica Cellio

Ok, so if I'm reading all this right, we have significant "outside" participation, but only in 5-10% of cases does that participation disagree with the community.

RhysW

oooh im a variable, serious note, thanks for taking the time to do this

Monica Cellio

10:25 PM

@RhysW I think you're the value of a variable, technically. :-)

I've been wondering how much of the issue comes from within versus without, and it's great to have some data to look at instead of everybody guessing about causes and then moving straight to remedies. Thanks again @Shog9. This is very helpful.

RhysW

@MonicaCellio potato potato (this phrase does not translate well to the internet)

Monica Cellio

@RhysW I believe I get your reference.

Shog9

@MonicaCellio Right. And if you consider that even most of those 5-10% are "controversial" in the sense that they're getting both up-votes and down-votes from insiders, you can see there's a lot to be gained from community education.

RhysW

@MonicaCellio phew, that would have been difficult to explain otherwise

Monica Cellio

@RhysW tomato tomato, yes.

RhysW

10:28 PM

Side note, im happy that disagreement in and out of the site is as low as that

Monica Cellio

@Shog9 it sure looks that way, yes.

RhysW

i think this user would have been valued over on relationships.se

-1

A: How to best deal with a flirting co-worker?

I think you should play along with the flirt and if that doesnt suit start avoiding them. What i did with a serial flirt I work with who I did fancy by the way but didnt want to meet outside work was play hot and cold. So you need to take a week and disappear altogether and stay well out of their...

enderland

data.stackexchange.com/workplace/query/178983/…

this is interesting

RhysW

a user with 100% upvotes

enderland

@MonicaCellio you need to downvote more :P

RhysW

10:35 PM

ahh thats why i dont exist, you put a 5k rep limit on it

enderland

yeah, whatever :P

RhysW

:P expanding to my rep range only adds like 3 users to it!

enderland

it gets more terribly revealing if you expand it down to 1k rep

can you do multisite queries in the data explorer? be interesting to look at total vote counts (up/down) for all sites

RhysW

me chad and jmac have almost identical percentages

though chad and jmac have consideraly more volume

@enderland dont think so

as you pick a site to run it against

side note, remove rep boundary, marvel at our 18 100% upvote users

enderland

wow. that's a lot of upvoting...

RhysW

10:43 PM

@enderland remove rep boundary, remove vote boundary, within 200 rows we hit 90% of votes as upvotes

thats 5300 users who mostly only vote up

enderland

yeah....

totalUpvotes: 112,552
totalDownvotes: 16266

RhysW

just a for fun one

data.stackexchange.com/workplace/query/179026/…

enderland

way to leave me out, jerk :)

RhysW

...oops thought i was missing someone!

data.stackexchange.com/workplace/query/179029/…

:P

enderland

it's nice because that includes total votes I think, whereas posts only includes existing votes (I think?)

Monica Cellio

10:52 PM

@enderland I agree. I've been working on that lately. (Early on I had the misguided idea that comments + waiting would help users learn and be friendlier.)

RhysW

@MonicaCellio I did that, now if im first there i DV to set an example, otherwise im a little more lenient, only throwing on a dv if its been around a while unedited or is really bad

enderland

so Workplace has 101,277 total score on posts, with a total of 96,286 in the "up minus downvotes" total, so what, that means about a total of -5k score has been deleted?

Monica Cellio

I thought I had more DVs than that actually. Does this include votes on deleted posts too? Hmm.

enderland

no

wait, yes it does

the users table shows ALL votes

the posts table show only votes on existing content

sadly you can't select the up/downvotes on posts individually

Monica Cellio

@RhysW if it's bad and old I DV. Before we graduated, if it was already negative I didn't bother voting and went straight to a delete vote. That may be why I have the impression I've downvoted more; I was probably counting VtDs.

@enderland ah, yes. Users table should have all of them and you have no need to go to posts (which doesn't). Thanks.

RhysW

10:56 PM

@MonicaCellio Yeah that seems similar to what i do, but if i VtD i put a DV on there too usually

enderland

@MonicaCellio yeah. so you've got 99 downvotes, I guess, hard to say on deleted/public content though

ffffffffffffff I love data analysis

why am I doing programming

data analysis is WAY more fun

Monica Cellio

@enderland user profile says 107, but that might just be caching. I am sad to learn that the "delete votes" tab on my profile doesn't include historic votes, only ones on posts that haven't been deleted yet. So I don't have a way to find out how many VtDs I've cast.

RhysW

damn i meant to leave like an hour ago, thanks for getting me caught up in this! :P

enderland

@RhysW lol. I'm going to be leaving shortly

RhysW

this happened last time i went to the data explorer too :/

Monica Cellio

10:59 PM

@enderland so you say. But ooh, look at the shiny data!

enderland

no I have a committment in 45 min I'm already going to be late for

peace out folks, don't have too much fun

RhysW

@enderland great, get me hooked and leave me, thanks!

right! i must away, cya all laters

jmac

11:44 PM

@enderland Wow, gnat has nearly triple the votes of the next highest voter (Chad). That's incredible!

I give two answers yesterday, and both hit the hot list. To not anger @gnat, I should stop answering. Clearly correlation = causation in this case.

@jmac After all the edits, there isn't much left of my original wording. But the end result does convey what I was trying to ask better than my first draft. — Fireflight 6 hours ago

Aggressive edit success?

Transcript for