Bots Help Moderate Stack Overflow(stackoverflow.blog)
stackoverflow.blog
Bots Help Moderate Stack Overflow
https://stackoverflow.blog/2019/09/17/meet-the-bots-that-help-moderate-stack-overflow/
4 comments
I'd argue that these sorts of efforts do help move things towards a better community.
The site relies quite a lot on knowledgeable people to review and edit things, and the more grunt work that entails, the more you have to really care about being a part of the site to commit to it. The types who are interested in shouldering that burden will tend to be the ones who are extra hardcore, who are also more likely going to be impatient with mistakes and newbies.
The site relies quite a lot on knowledgeable people to review and edit things, and the more grunt work that entails, the more you have to really care about being a part of the site to commit to it. The types who are interested in shouldering that burden will tend to be the ones who are extra hardcore, who are also more likely going to be impatient with mistakes and newbies.
I think the sentence that got me was:
"Natty started to flag answers automatically once it had detected a particular answer as a non-answer with considerable accuracy."
It might be that I don't understand - does Natty flag something directly, or flag them for the moderator?
On the other hand, the Heat Detector bot sounded it was punitive too, but in a way that might strengthen the community. (interesting thought - could it flag moderators?)
"Natty started to flag answers automatically once it had detected a particular answer as a non-answer with considerable accuracy."
It might be that I don't understand - does Natty flag something directly, or flag them for the moderator?
On the other hand, the Heat Detector bot sounded it was punitive too, but in a way that might strengthen the community. (interesting thought - could it flag moderators?)
Flagging doesn't mean deleting. It brings those non answers to attention.
I spent an hour looking through them today and I've only seen one false positive and of course someone reported it instantly.
It's pretty useful. One of the questions I saw today was from few years ago and no one had answered it. Someone added a "me too" answer that wasn't helpful. I flagged that answer and provided a new answer to the question, and now it may help someone in the future.
I spent an hour looking through them today and I've only seen one false positive and of course someone reported it instantly.
It's pretty useful. One of the questions I saw today was from few years ago and no one had answered it. Someone added a "me too" answer that wasn't helpful. I flagged that answer and provided a new answer to the question, and now it may help someone in the future.
> could it flag moderators
Why shouldn't it? Moderators are just like normal users but with some extra privilege regarding editing/deletion.
There seems to be a bit of confusion on how SO moderation works every time SO is discussed here. Most of the moderation is done by other users. I'm very active in editing/flagging/closevoting, and I can assure you the instances I see an actual moderator (the ones with a diamond next to their name) taking matters into their hands in a given month can be counted in single digits.
Why shouldn't it? Moderators are just like normal users but with some extra privilege regarding editing/deletion.
There seems to be a bit of confusion on how SO moderation works every time SO is discussed here. Most of the moderation is done by other users. I'm very active in editing/flagging/closevoting, and I can assure you the instances I see an actual moderator (the ones with a diamond next to their name) taking matters into their hands in a given month can be counted in single digits.
Cool tools to leverage human effort in effective ways.
I’d love to see a “smells like homework” tag or review queue
to catch the questions that sound a lot like homework assignments from first courses in programming, so willing respondents can more readily draw out the querent’s understanding. (“print the squares of the first 10 integers”, “swap the first & last line of the input”,
and similar)
As much as people love to rip on StackOverflow for being unfriendly, it is in an entirely different plane of existence from all the little php forums that it replaced.
You haven't seen arbitrary, shitty moderation unless you were on the internet in the early 2000s.
The craziest thing is, those people are almost all still on the internet, but they've aged out of the worst of the terribleness.
You haven't seen arbitrary, shitty moderation unless you were on the internet in the early 2000s.
The craziest thing is, those people are almost all still on the internet, but they've aged out of the worst of the terribleness.
On the other hand you could switch communities if one style didn't suit your taste back then. It often seems like this was replaced by a choice between large communities like SO and something fringe.
Maybe stop banhammering newbies simply because you think their question about obscure arcane algorithm #44462 is too similar to an even more arcane question 10 years ago about algorithm #783429 and its immediately obvious to a regular with a pH.D in quantum statistics.
The only major support site I'm aware of with a 'stupid questions' punishment policy. Which isn't a bad thing in and of itself except its implemented in such a draconian and nontransparent way. Demerits never seem to expire, even after years of no 'stupid' posts and theres no explanation where they come from.
The bizarre thing is even with this and newbies banned left and right it seems the admin is on another planet running some other website as all they want to talk about is social justice and how evil heteronormative SO users somehow can psychically tell who is a woman or minority and supposedly harassing them to epidemic proportions but nobody somehow ever sees this.
Maybe get a handle on a few real basic problems before you start crowing about your geewhiz new bot.
The only major support site I'm aware of with a 'stupid questions' punishment policy. Which isn't a bad thing in and of itself except its implemented in such a draconian and nontransparent way. Demerits never seem to expire, even after years of no 'stupid' posts and theres no explanation where they come from.
The bizarre thing is even with this and newbies banned left and right it seems the admin is on another planet running some other website as all they want to talk about is social justice and how evil heteronormative SO users somehow can psychically tell who is a woman or minority and supposedly harassing them to epidemic proportions but nobody somehow ever sees this.
Maybe get a handle on a few real basic problems before you start crowing about your geewhiz new bot.
> Maybe stop banhammering newbies simply because you think their question about obscure arcane algorithm #44462 is too similar to an even more arcane question 10 years ago about algorithm #783429 and its immediately obvious to a regular with a pH.D in quantum statistics.
You don't get banned for questions closed as duplicate, what are you talking about?
You don't get banned for questions closed as duplicate, what are you talking about?
Yes you do. Among other things. Its one of the things that counts toward your 'demerit' score. Or at least it certainly appears to. Theres people who haven't done anything but ask a couple 'stupid questions' that were answered before that were modded. Of course their moderation policy is so opaque only the admin could tell you exactly what happened.Or maybe not since they seen keen to adapt the 'Google approach' of having their bots take over sight unseen.
I stand corrected on the fact that there's no ban procedure[0]. But there's also a provision that this ban lasts 6 months, not indefinitely.
And it's not about questions getting closed, it's about downvotes.
And from that to saying that the procedure is opaque, I'd say you'll have to provide some sources about that.
[0] https://stackoverflow.com/help/question-bans
And it's not about questions getting closed, it's about downvotes.
And from that to saying that the procedure is opaque, I'd say you'll have to provide some sources about that.
[0] https://stackoverflow.com/help/question-bans
The procedure is opaque because:
He posts questions that are duplicates but are only obvious duplicates if you're have a PhD in quantum statistics. And he attracts downvotes and gets demerits. Most SO users don't have PhDs, therefore most of the the downvotes and demerits must be due to something mystifying.
The procedure is not opaque because:
Well, maybe his questions may be regarded differently by other people, and then it makes sense.
There's a rule on meta.stackoverflow.com that each such complaint must point to an example question. A good rule.
He posts questions that are duplicates but are only obvious duplicates if you're have a PhD in quantum statistics. And he attracts downvotes and gets demerits. Most SO users don't have PhDs, therefore most of the the downvotes and demerits must be due to something mystifying.
The procedure is not opaque because:
Well, maybe his questions may be regarded differently by other people, and then it makes sense.
There's a rule on meta.stackoverflow.com that each such complaint must point to an example question. A good rule.
Oh they definitely downvote you simply for closed questions.Which you'd know if you regularly used the site or didn't assume that I was unfamiliar with it. Also the bans might be temporary but the demerits last for years at least. I know because I have ones from 2016. Of course most of what you and I are might say about actual policy isn't certain since its all hidden. The mark of bad moderation policy is hidden/vague policy but the overwhelming practice in most places including here.
They added all these automated ways to schedule service for your car through the app or website, and after they got comfortable with it, they made it the ONLY way to schedule service. [good for tesla, not good for community]
The problem I see is that Stack Overflow tries to be an exclusive community.
I think maybe instead of automated ways to flag bad things, maybe they can figure out automated ways to praise people, be inclusive especially with new users and redirect mistakes into constructive avenues.