I disagree. Ignoring SO's dubious classifications of what's on/off-topic. The last couple questions I've posted have been shit-shows of off-topic comments and answers.
If you ask a hard question, a lot of people seem to only want to uselessly question what you're doing then answer a different, easier question. Those are totally worthless, and force me to waste a lot of time writing really defensive questions that try to nip the derails in the bud.
Honestly, SO should have some feature where the asker can just delete responses as off-topic.
> A better change would have been to make any downvote COST 10 points. Or at least 1 point or 2.
IIRC, it already costs 1 karma point to downvote an answer. I would guess it's the same for a question.
I would urge you, or anyone who thinks that downvoting is toxic for this matter, to volunteer in moderating questions written by new users for a day or two.
Once you see a constant flood of "I made app, here my 1000 lines of code (screenshot), I get error, where bug?" that the moderators have to defend the site against - you might just change your opinion about the value of downvotes.
I'm not saying there should not be a mechanism for regulating that sort of stuff.
BUT on StackOverflow it is entirely normal, even expected, that you post a well thought out and reasoned question and its instantly downvoted because, well, who knows why? Drive-by downvoting is simply negative.
A different mechanism could be used for the sorts of issues you identify.
>A different mechanism could be used for the sorts of issues you identify.
Would the new proposed mechanism require more volunteer work or the same effort as now?
Btw, many people forget that back in 2011, Jeff Atwood was concerned that there was too much crap and the community was not downvoting enough. Therefore, he removed the points penalty to encourage the community to aggressively downvote more:
Is this really the common behavior? I feel like I live in a different world... Maybe I do; I don't ask a ton of questions, and when I do they are usually on pretty small niche tags, but the last several times I've asked something, I've wound up with a friendly soul suggest something within an hour or two that's gotten me pointed in the right direction.
Can you point to an example of a "well thought out and reasoned question" on SO that's at a negative score?
In my experience, that just doesn't happen. Most questions sit at a score of 0, and those which do receive votes are almost always upvoted. (The data supports this[1].) Questions rarely end up with a negative score, and when they do it's usually because there are obvious problems with their formatting or content. (Usually low effort, off topic, or an obvious duplicate.)
on StackOverflow it is entirely normal, even expected, that you post a well thought out and reasoned question and its instantly downvoted because, well, who knows why
Hmm, not my experience, at all. Though I heard this more than once, so I see a couple of possible reasons (or combinations thereof):
- depends on the community/tag; I've mainly asked questions tagged with the languages starting with a C, some git stuff and some more obscure things
- depends on the poster's reputation points i.e. maybe there's a bias to not downvoting/closing questions asked by people with 10k+ points?
- the questions were not actually well thought out/reasoned even though you think they are
Alternatively, maybe we could come up with a way to deal with those questions that is better than down voting.
Let time, and lack of up votes, sink them? Have moderators hide them with a FRIENDLY "Hey, we've hidden your post because it doesn't meet 'these requirements', once you've edited, we'll review it and show it again." message. If they don't improve, keep it hidden, tell them to do better. If they still don't improve, delete it, and require any new questions they post to be moderated first. If another question ends up deleted, don't let them post any more.
Basically, give feedback in a positive manner with the goal of getting that person to do better. It may seem like more work, but I think it would be much more rewarding than what I've seen on Stack Overflow and other support forums.
What is wrong with downvotes? Without downvotes upvotes loose their meaning. The number of upvotes would turn into a proxy for the number of people who looked at the question.
From my experience - big part of new users asking questions do not even care if they get any upvotes or downvotes, or if someone is friendly. They don't even bother to accept/upvote the answer when it's given. They simply want to get their question answered and then move on, forget their username and create a new account for the new question next time.
A better system is surely possible. But my main message would be - please spend some time doing it all: 1) asking questions; 2) answering questions; 3) volunteering for moderation; before starting to think about how to change the site.
They don't tell the person what they did wrong. Its incredibly frustrating to get downvoted and not have any idea why. For example, here on HN, I generally couldn't care less about what my score is, but when I write something based on my personal experience (ie something completely subjective from my life) and I get downvoted (it has happened!), I find find that incredibly frustrating because I don't understand why someone found my experience worthy of a downvote.
I imagine its similar on SO. If you're new and don't yet know the finer points of the rules, getting downvoted without an explanation is a surefire way of turning that person off the site.
> please spend some time doing it all: 1) asking questions; 2) answering questions; 3) volunteering for moderation
I have. At least, I've done the review queue a ton of times.
SO promts you to leave a comment about how a question could be improved. I always leave a comment unless the issues is already covered in another comment.
Sure, many people do and those really aren’t a problem at all. Its the drive-by downvotes that are a problem. Maybe forcing you to leave the comment would be enough to mitigate the problem.
>> I would urge you, or anyone who thinks that downvoting is toxic for this matter, to volunteer in moderating questions written by new users for a day or two.
>> Once you see a constant flood of "I made app, here my 1000 lines of code (screenshot), I get error, where bug?" that the moderators have to defend the site against - you might just change your opinion about the value of downvotes.
My favorite is always when they dont get a response and then get angry at the community -- "i asked three days ago, STILL NO ANSWER, ANSWER NOW"
Agreed. Down voting is toxic. Though, rather than costing points, I'd prefer to see the voter required to explain why they are down voting.
As small as it is, a down vote is a public action saying that the person posting what you are voting on has less value. At least that what they said has less value. That is never a fun experience, even if you have very thick skin.
Good. I'm glad you explain. That doesn't make it any less of a public reprimand.
In polite society, when you reprimand someone publicly, you only do so when private discussions have failed.
I'd also argue using down votes for outdated information is very inefficient. Wouldn't it make sense to just have a "this is out dated" flag, and list how many people have used it? Then, the author of the answer could update their post and ask that the flag be cleared.
For wrong answers, I'd say a comment "this is wrong because" first, then something other than down voting.
> I'd say a comment "this is wrong because" first, then something other than down voting.
I guess its just as much a signal to people looking for an answer as it is for the person who wrote the answer. At least with a downvoted answer, its easy to see "ok, maybe not this one" and you can check the comments to see why. Not that I'm a fan of downvoting, I think a flagging system would achieve the same thing while being more helpful to all involved than an opaque downvote.
I think downvoting answers is slightly less problematic than downvoting questions, although both suffer from the same issue if people don't explain why they did it. At least if an answer is downvited, its easy to assume "this is wrong", while if its a question its.. hard to know why unless explained.
Depends on if the question is easily googled or a general dupe.
I don't know about you, but I find when I'm on a poorly-moderated forum, it's just the same questions asked over and over again every day.
Do you really need to explain Search to every noob who you downvote?
(Note that I never look at new Stackoverflow questions. I always find the site through Google, because someone's always asked the question first. I do improve answers when needed.)
At the same time it's a great tool for assessing the quality of posts. It's the same reason HN allows downvotes.
However the dynamic of SO is different. Users are coming to the site to get help and solve problems, not to have open-ended discussions.
SO should have a grace period for new questions. No downvotes during the first 1-2 hours. Aside from being less discouraging for new users it would also mitigate cases of incorrect downvoting. A negative score is a signal to other users that a question is not worth consideration, and dissuades possible answerers from providing a solution.
The problem is that people have different ideas of what a downvote means.
To someone who has spent a ton of time answering questions, it means "this is not a well-written question and I do not have the time to explain to the poster why that is; I would rather spend my time on answering well-written questions, which is what I came to this site to do."
To a question asker, it means "you suck and I hate your question, and I'm not going to tell you why."
The problem is that there are a couple of mechanisms for getting rid of questions - downvoting, or flagging / voting to close. Some questions are clearly on the wrong site - just close them. Some questions are not good questions (screenshots of code, no explanation of desired behavior, no research done into error message) but they are not closeable according to the options provided. The downvote is just to send a message.
Answerers who downvote are people who genuinely want to help - they are spending their free time answering questions on the internet for no gain other than extremely localized imaginary internet points.
The solution is providing one method of feedback to the question poster, and couching it in a less negative way. Maybe all methods have the asker's feedback get hit, maybe none of them. Maybe the asker should be protected from feedback until they learn the site.
Asking technical questions properly is hard. I got downvoted and closed and duplicate questions on my first 5 attempts, and even my recent questions have 0-1 upvotes (my answers fare better, for whatever reason).
> Answerers who downvote are people who genuinely want to help - they are spending their free time answering questions on the internet for no gain other than extremely localized imaginary internet points.
I think it's more accurate to say that answerers who downvote want to give answers (vs genuinely want to help). Because explaining to someone why their question is problematic is part of helping.
This is more nuanced. And StackOverflow as a website themselves are part to blame. They cannot seem to make up their mind if they are there to help the people who ask questions, or if their task is to construct a database of problem-answer pairs for future reference, like a type of programming encyclopedia.
For the longest time everyone on the site believed it to be the "database" model. And helping individual users with their questions is not part of it. Under this framework answerers who downvote are simply doing what they are expected to do - weeding out poor entries from the encyclopedia. And the person who asks a question has no right to expect an answer, as stated in their 2011 blog entry:
> We believe asking questions on our site is a privilege, not a right. [1]
However recently they seem to be moving away from that model and only time will tell what it morphs in to.
I think they've made a lot of progress with the "be nice" thing. A lot of the complaints in this thread I would have heartily agreed with 4-5 years ago, but they seem dated to me now. Maybe it's my perspective that's changed, that's always a possibility.
The database model has never seemed right to me, but I do think SO is making progress moving away from it. Ironically, I believe they will be better able to achieve the database goal once they let it be an emergent property of Q&A rather than grudgingly allowing Q&A to get the database.
I find this is true on questions with the Go tag. They can't seem to let a question languish at zero. Or just downvote it once to -1, they have to hammer it. Upvotes are rare, so on balance the entire page of go questions frequently has a negative balance.
For a great many tags, it seems like there is a sport to show that a question doesn't apply to a specific tag, that it is a duplicate, or that the user should be doing something a different way than what they are asking for. There seems to be so much effort put into deliberately not answering the question that trying to participate is truly deflating.
As an experiment downvoting is disabled on some smaller reddit subs. The effect in the gaming /hobbiest communities is they get overrun with memes and content designed to get a reaction. Theres an overwhelming number of readers who never participate in discussions but will upvote the most mindless stuff. You can see consensus in a forum that few enjoy the content, yet it gets strongly upvoted if unchecked.
I predict that this will have zero effect on anything. The people who care strongly about accumulating SO e-points are the ones answering the questions, not the ones asking them.
SO has always been hostile to questioners by design. It's a natural consequence of requiring "high-quality" questions.
Since most questions will not be "high-quality" (for any reasonable definition of "high-quality"), most new participants' experience will be of asking a question and being told "that's a bad question" either explicitly (downvotes) or implicitly (editing their question to "improve" it).
It gets worse as time goes on because of the "no duplicates" policy; the more questions that are in the system, the harder it is to come up with a question that isn't marked (correctly or incorrectly) as a dupe.
Banning all politeness words (hello/goodbye/please/thank you/you're welcome) from questions/answers because they're "noise" is just the cherry on top that makes people who aren't already immersed in SO culture think that everybody there is rude/mean, no matter how many times they put "be nice" in the CoC.
Not saying these are _bad_ policies per se - SO has a stated goal of being a "curated resource of high quality questions and answers", and they need to separate the wheat from the chaff somehow, but the policies _will_ be perceived as hostile by most new users.
Giving the questioners an extra +5 fake internet points will not fix any of that.
You just summarised how I feel whenever I ask a stackoverflow question.
I wasn’t aware that politeness was banned but I avoid asking questions there even if it means joining free node channels and waiting hours in hopes that someone answers my question.
S/O has some very high quality responses but I actively avoid asking questions there because it feels over policed.
> It gets worse as time goes on because of the "no duplicates" policy; the more questions that are in the system, the harder it is to come up with a question that isn't marked (correctly or incorrectly) as a dupe.
I'm curious why they don't have a no duplicate answers policy.
I can't speak to their reasoning, but I value having multiple similar answers.
They often differ slightly in their assumptions or implementation, which can help in grokking the core concept of the answer and not getting distracted by implementation details, especially for languages and domains I'm not very familiar with.
Since most questions will not be "high-quality" (for any reasonable definition of "high-quality"), most new participants' experience will be of asking a question and being told "that's a bad question" either explicitly (downvotes) or implicitly (editing their question to "improve" it).
This describes the problem with SO remarkably well - how it constantly trends towards being even less nice than it was two months beore. I've gotten SO points and even been congratulated over some questions and some answer. But I'm still angry, seven years on, over having one particular question being deleted - it was my question and it was a question I wanted an answer to it.
> We know that people who contribute by asking have a harder time earning privileges than people who focus on answering. Independent research suggests this disproportionately affects women:
> > We also see that women contribute differently to building the community’s knowledge base: they are asking more questions. Stack Overflow’s current system strongly incentivize answering by rewarding upvotes on answers twice as much upvotes on questions.
This is a waste. Few people upvote questions because they are well-written; by and large it is that they had the same question and you asked it before they had to do so themselves. This heavily biases upvotes on extremely novice questions from 8 years ago (before they were all asked), and gives new question askers a bad example of what a good question is ("I just did the same thing this person did, and they got 200 upvotes, and my question was closed in 2 minutes!").
Question askers get the reward of having their question answered. I think votes on questions should be done away with entirely. Questions that should be exemplified can have some other metric (stars for "this is well written and generalizes easily for other users"), and low-quality questions should be flagged or closed, which at least gives some sort of reason. Downvotes aren't useful to askers because they have no idea what they're doing wrong when they receive them.
Keeping the participation friction as low as possible ensures the maximum number participate, which ensures questions have maximum visibility. Changing the simple vote up/down process with something requiring to specify a reason can reduce the number views a question gets.
Also nothing is stopping a downvoter from leaving a comment explaining why the question needs improvement, other than lack of reputation (you need a minimum reputation for your downvotes to be visible anyway). People who care to explain their downvotes can already use the existing process. Since comments can be upvoted (but not downvoted), there is already an incentive in place for that as well.
Many people mention they avoid commenting on downvoted questions for fear of the wrath of the asker. People seem to feel differently about close voting, perhaps since they are collected and presented as a group.
I don't really get the point of the vigilance toward bad but not intentionally spam questions. The question # has been used, derank it and call it a day until someone sees a reason to revive it.or the asker ASKS for help improving their question.
The specific change I would like to see is to really raise the bar on what is considered a duplicate.
I think a rule could be "if someone with a decent track record (for some definition of decent, maybe 100 points? 1000?) felt the need to ask it, it's not a duplicate." The bar could be raised in other ways as well, but this one I think would reward people who have done the work to be a positive part of the community.
We all have varying degrees of ability to generalize the principle of one answer to another. The other part of this is that you get better at asking questions through practice, which is hard to get early on. At this point, most problems I can't answer are good questions or too specific, but it took a while and I don't want SO to be the place where only those with 5-10 years in can get help.
It seems to me like someone with few points is least likely to understand a solution to a duplicate of their problem.. I don't really see the harm in linking it and still letting people answer or comment on the proxied solution.
SE spends half the time saying an overlapping diversity makes it easy to find a question and answer that fits and the other half eliminating that diversity (and bringing non-adept people to dead ends in their searches that land on SO.)
My proposal is intentionally moderate. If I had my way, I'd go further, but this is something I've thought about and think it may be palatable to those at SO who want to preserve the strictness and maintain respect for those who contribute. You still have to put in a lot of work to figure out how things work, but if you do the work you get credibility.
I also think being able to override votes to nuke an incorrect answer would be handy, I recently begun getting involved in HTML/CSS for a somewhat obscure project, some of the older answers around front end topics are extremely dated and incorrect by today’s standards (flexbox, for example, deprecates a bunch of hacks). Sure there might be the correct response 3-4 answers down, but I shouldn’t need to dig for that.
IMO the outdated info problem is the biggest one Stack Overflow has right now, much worse than posts getting flagged as dupes on a questionable basis. Theoretically, you're just supposed to add a new answer that's better — but you almost certainly won't get the checkmark, and you probably won't even get more votes than the old answer (since your answer will only get search traffic and not front-page traffic, and you're already below all the other answers to begin with). So Stack Overflow slowly transforms into w3schools.
Votes (both up and down) on questions are a way to prioritize questions that need answers. Upvotes should be welcomed from all (fellow users who want an answe), but downvotes should be welcomed only from reputable answerers who took time to read the question and decided it's poorly formed.
Answered questions would be fine having their vote button disabled.
Unfortunately, the moderators and I had a disagreement about the importance of asking questions without attempting to first solve the problem.
Well written (and well-defined) questions are the key StackOverflow. From the questions, you could extract a Rosetta Stone of programming. An experienced programmer could learn most languages on the job.
I sort of agree with you. The demand and supply for askers and answerers is imbalanced:
questioners: 24035
answerers: 10888
It's not that we are running out of people who have questions. The reward system needs to reflect this imbalance. Also, I'm afraid there will be increase in questions and corresponding increase in lack of answers demotivating people who genuinly wants to ask without regards to reputation.
To curb race to first asker, SO should implement exponentially discounting reward. This makes final commulative reward finite.
I've tried saying it on Meta, but that place is not great for discussion:
Newbies should help newbies.
There are lots of ways to incentivize this, but the basic gist is that users who are just growing out of the newbie phase are in the best position to be indoctrinating new arrivals. They'll have the most patience and understanding, and it's an easy job that shouldn't be done by super experienced users.
Most of this could easily be isolated to a newbie pre-stage too. Basically, a tutorial area in which new users work together to make their questions better so they can move on to the non-tutorial experience.
Personally, I'd filter these posts out by default, but on days where I was feeling nice I might go in and help out.
Worst case scenario: it takes some extra work from a newbie to get their first bits of rep.
Rewarding the question askers... like Monica Cellio, who asked a clarifying question about the not-yet-implemented pronoun policy, and you canned her for it? That sort of reward?
If they say that, in that way, then they are being misleading.
Months ago, well before the current Stack Exchange pronoun policy was a thing, she said (yes, explicitly) that she wasn't willing to use singular "they" because it feels so wrong and confusing to her. This wasn't a "plan to ignore the policy" because at that time there was no policy.
(It also doesn't appear to have been motivated by sexism, transphobia, etc., although people who say such things sometimes are; she is explicitly willing to use neopronouns -- "xe" and the like -- for people who find that "he" and "she" don't fit them.)
Then, much more recently, the upcoming change to the Code of Conduct was announced in the moderators' chatroom. M.C. asked some questions about that, very definitely did not state any plan to ignore the policy, and promptly got de-moderator-ated.
It's possible that she isn't telling the truth about what has happened in private, away from the view of other moderators. Maybe she sent an email to SE management saying "I'm not going to abide by this; fire me if you dare" or something. But as far as externally-visible things are concerned, her account of things matches up better with reality than SE's statements do.
There are, indeed, moderators who think M.C. should have been canned long ago, but my impression is that there's something close to unanimity that SE should not have canned her the way they did, and that the majority of those with any opinion on the matter think that they shouldn't have canned her at all.
(Source: am a moderator on a Stack Exchange site, have read all the relevant transcripts from said chatroom. Note that discussions there are confidential, so the above is about as specific as I am prepared to be about who said what.)
Because she at least offered a very detailed breakdown of events that can be fact-checked, critiqued, etc. StackOverflow has offered basically nothing more than vague PR BS.
I am sure they are quite happy. Users? Not so much.
Getting at great contributions is now much harder. I use Quora much less now.
Churn 'n burn for traffic dollars is a strategy.
Suppose it opens a door somewhere too. Do that enough, and there is clearly room for a less noise filled source.
IMHO, the difference here is Quora has a plausible, casual user case that makes sense. The people who helped to build Quora content won't give two shits and will move on, but Quora itself may do just fine. Meh.
Degrading the value to current users of SO to attract more in general, leads where for SO?
It might be better to allow the market to decide how the reward mechanism should work.
People who see a question as relevant and interesting to them, would add points or monetary incentives. Thereby increasing the speed and quality of the answers.
This mechanism has already been tested out and actually works only with crypto currency due to low transaction costs.
Here is an example:
What makes you nostalgic?
Answer with #Obyte Know-it-all and earn up to 410,172,272 Bytes : https://t.co/EaVFrscFC7 (enough Bytes to make > 700,000 transactions whenever you want in the future thanks to obytes' tiny predictable fees)
That's awesome and its a step in the right direction.
And I notice a lot of female employees or managers being mentioned which might be a coincidence but it also might be deliberate and somehow be related to the idea that females might have more empathy? Which could just be my own prejudice and not true but I dunno.
But anyway, if they are going for nicer/more empathetic employees and policies, that's great, but its not going to stop all of the (mostly male) super-critical narcissistic arrogant shaming-happy vote-whores who are eagerly waiting to get in another passive-aggressive veiled hint that I am an idiot while they downvote my question.
Not planning to ask another question on Stack Overflow anytime soon, thanks though.
I've never asked or answered a question on SO (I wouldn't dare!) because I can plainly see the horrorshow that is likely to come with doing that by reading other questions. Over time, I've come to simply ignoring SO for the most part.
I think I may check in from time to time to see if the place actually improves.
It is funny that we have the "techies" (programmers, engineers, etc) designing and managing these social media platforms when the techies are stereotyped as awkward and antisocial.
Who decided that software developers should have the final say over how people interact? I feel like there's probably a field or two of experts on that that should be being consulted.
When one considers that the median experience of asking a question on stackoverflow is "immediately closed as a duplicate of a vaguely similar question from six years ago", this seems like a pretty superfluous change.
A far more meaningful way to reward askers would be to simply change how dupes are handled. Instead of closing questions with a link to a related question, give users a mechanism to answer questions with a link to a related answer. Then if the link really does solve the problem the asker can accept that answer, and if not the original question can stay open.
I get why SO chose their current aggressive stance on dupes, but it's probably time to try something else.
Based on my past actions I guess I would be happy to donate a few dollars to anyone who makes a better q/a site and makes me aware of it.
By better I mean:
- important current database (allowed with current licensing)
- rewrites the meta game to make helping (answering) more valuable than administering (deleting).
- raises bar to asking questions (stack overflow seems to always be crying over all the stupid questions they get yet the thought of raising the bar for asking questions never seen to get any traction)
- have a karma tax or something on questions and contributors: just because someone was the most valuable contributor or something was the best question two years ago doesn't mean they still are.
- don't split into many subsites for every topic, if people want to ask it and other people want to answer it and it isn't illegal or scary or anything, bust leave it alone. Karma tax will take care of it in the long run.
It took them 7 years to change their minds on it. Sigh.
---
Side rant: This whole piss people off/break things and expect people to stick around is a dumb idea. Company does stupid things, I'm going to find an alternative company and not return.
79 comments
[ 3.0 ms ] story [ 129 ms ] threadDownvoting is toxic.
A better change would have been to make any downvote COST 10 points. Or at least 1 point or 2.
> Downvoting is toxic.
I disagree. Ignoring SO's dubious classifications of what's on/off-topic. The last couple questions I've posted have been shit-shows of off-topic comments and answers.
If you ask a hard question, a lot of people seem to only want to uselessly question what you're doing then answer a different, easier question. Those are totally worthless, and force me to waste a lot of time writing really defensive questions that try to nip the derails in the bud.
Honestly, SO should have some feature where the asker can just delete responses as off-topic.
> A better change would have been to make any downvote COST 10 points. Or at least 1 point or 2.
IIRC, it already costs 1 karma point to downvote an answer. I would guess it's the same for a question.
- https://stackoverflow.com/help/privileges/vote-down - https://stackoverflow.blog/2011/06/13/optimizing-for-pearls-...
Once you see a constant flood of "I made app, here my 1000 lines of code (screenshot), I get error, where bug?" that the moderators have to defend the site against - you might just change your opinion about the value of downvotes.
BUT on StackOverflow it is entirely normal, even expected, that you post a well thought out and reasoned question and its instantly downvoted because, well, who knows why? Drive-by downvoting is simply negative.
A different mechanism could be used for the sorts of issues you identify.
Would the new proposed mechanism require more volunteer work or the same effort as now?
Btw, many people forget that back in 2011, Jeff Atwood was concerned that there was too much crap and the community was not downvoting enough. Therefore, he removed the points penalty to encourage the community to aggressively downvote more:
https://stackoverflow.blog/2011/06/13/optimizing-for-pearls-...
In my experience, that just doesn't happen. Most questions sit at a score of 0, and those which do receive votes are almost always upvoted. (The data supports this[1].) Questions rarely end up with a negative score, and when they do it's usually because there are obvious problems with their formatting or content. (Usually low effort, off topic, or an obvious duplicate.)
[1]: https://data.stackexchange.com/stackoverflow/query/1096887/q...
Hmm, not my experience, at all. Though I heard this more than once, so I see a couple of possible reasons (or combinations thereof):
- depends on the community/tag; I've mainly asked questions tagged with the languages starting with a C, some git stuff and some more obscure things
- depends on the poster's reputation points i.e. maybe there's a bias to not downvoting/closing questions asked by people with 10k+ points?
- the questions were not actually well thought out/reasoned even though you think they are
Any ideas?
Let time, and lack of up votes, sink them? Have moderators hide them with a FRIENDLY "Hey, we've hidden your post because it doesn't meet 'these requirements', once you've edited, we'll review it and show it again." message. If they don't improve, keep it hidden, tell them to do better. If they still don't improve, delete it, and require any new questions they post to be moderated first. If another question ends up deleted, don't let them post any more.
Basically, give feedback in a positive manner with the goal of getting that person to do better. It may seem like more work, but I think it would be much more rewarding than what I've seen on Stack Overflow and other support forums.
From my experience - big part of new users asking questions do not even care if they get any upvotes or downvotes, or if someone is friendly. They don't even bother to accept/upvote the answer when it's given. They simply want to get their question answered and then move on, forget their username and create a new account for the new question next time.
A better system is surely possible. But my main message would be - please spend some time doing it all: 1) asking questions; 2) answering questions; 3) volunteering for moderation; before starting to think about how to change the site.
They don't tell the person what they did wrong. Its incredibly frustrating to get downvoted and not have any idea why. For example, here on HN, I generally couldn't care less about what my score is, but when I write something based on my personal experience (ie something completely subjective from my life) and I get downvoted (it has happened!), I find find that incredibly frustrating because I don't understand why someone found my experience worthy of a downvote.
I imagine its similar on SO. If you're new and don't yet know the finer points of the rules, getting downvoted without an explanation is a surefire way of turning that person off the site.
> please spend some time doing it all: 1) asking questions; 2) answering questions; 3) volunteering for moderation
I have. At least, I've done the review queue a ton of times.
Nothing is inherently wrong with them, but downvotes without an explanation are without value.
My favorite is always when they dont get a response and then get angry at the community -- "i asked three days ago, STILL NO ANSWER, ANSWER NOW"
Then you ask them a question to get more specific info and they just reword the same info from the question again.
Cherry on the top is when someone new comes in the comments and they yet again ask the same thing, I loved that.
Make accounts cost something:
- limited number of invites per year for existing valuable members
- small payment
- hassle (send us a post card or something like that, a "proof of work for humans")
Stack Overflow seems to want:
1. asking to be frictionless even for complete strangers
2. yet to be taken seriously when they continue to complain that this is going on.
if it's really that bad of a question, why are there 20 answers for it?
of far greater concern are the outdated and incorrect answers with the coveted green check mark next to them.
one sees that green check mark and believes all manner of drivel written next to it.
As small as it is, a down vote is a public action saying that the person posting what you are voting on has less value. At least that what they said has less value. That is never a fun experience, even if you have very thick skin.
Why do I downvote? I downvote when an answer is wrong or outdated.
That's the point of downvotes, to push incorrect or outdated answers to the bottom.
In polite society, when you reprimand someone publicly, you only do so when private discussions have failed.
I'd also argue using down votes for outdated information is very inefficient. Wouldn't it make sense to just have a "this is out dated" flag, and list how many people have used it? Then, the author of the answer could update their post and ask that the flag be cleared.
For wrong answers, I'd say a comment "this is wrong because" first, then something other than down voting.
I guess its just as much a signal to people looking for an answer as it is for the person who wrote the answer. At least with a downvoted answer, its easy to see "ok, maybe not this one" and you can check the comments to see why. Not that I'm a fan of downvoting, I think a flagging system would achieve the same thing while being more helpful to all involved than an opaque downvote.
I don't know about you, but I find when I'm on a poorly-moderated forum, it's just the same questions asked over and over again every day.
Do you really need to explain Search to every noob who you downvote?
(Note that I never look at new Stackoverflow questions. I always find the site through Google, because someone's always asked the question first. I do improve answers when needed.)
SO heavily encourages leaving constructive feedback rather than simply downvoting.
At the same time it's a great tool for assessing the quality of posts. It's the same reason HN allows downvotes.
However the dynamic of SO is different. Users are coming to the site to get help and solve problems, not to have open-ended discussions.
SO should have a grace period for new questions. No downvotes during the first 1-2 hours. Aside from being less discouraging for new users it would also mitigate cases of incorrect downvoting. A negative score is a signal to other users that a question is not worth consideration, and dissuades possible answerers from providing a solution.
To someone who has spent a ton of time answering questions, it means "this is not a well-written question and I do not have the time to explain to the poster why that is; I would rather spend my time on answering well-written questions, which is what I came to this site to do."
To a question asker, it means "you suck and I hate your question, and I'm not going to tell you why."
The problem is that there are a couple of mechanisms for getting rid of questions - downvoting, or flagging / voting to close. Some questions are clearly on the wrong site - just close them. Some questions are not good questions (screenshots of code, no explanation of desired behavior, no research done into error message) but they are not closeable according to the options provided. The downvote is just to send a message.
Answerers who downvote are people who genuinely want to help - they are spending their free time answering questions on the internet for no gain other than extremely localized imaginary internet points.
The solution is providing one method of feedback to the question poster, and couching it in a less negative way. Maybe all methods have the asker's feedback get hit, maybe none of them. Maybe the asker should be protected from feedback until they learn the site.
Asking technical questions properly is hard. I got downvoted and closed and duplicate questions on my first 5 attempts, and even my recent questions have 0-1 upvotes (my answers fare better, for whatever reason).
I think it's more accurate to say that answerers who downvote want to give answers (vs genuinely want to help). Because explaining to someone why their question is problematic is part of helping.
For the longest time everyone on the site believed it to be the "database" model. And helping individual users with their questions is not part of it. Under this framework answerers who downvote are simply doing what they are expected to do - weeding out poor entries from the encyclopedia. And the person who asks a question has no right to expect an answer, as stated in their 2011 blog entry:
> We believe asking questions on our site is a privilege, not a right. [1]
However recently they seem to be moving away from that model and only time will tell what it morphs in to.
[1] https://stackoverflow.blog/2011/06/13/optimizing-for-pearls-...
The database model has never seemed right to me, but I do think SO is making progress moving away from it. Ironically, I believe they will be better able to achieve the database goal once they let it be an emergent property of Q&A rather than grudgingly allowing Q&A to get the database.
For a great many tags, it seems like there is a sport to show that a question doesn't apply to a specific tag, that it is a duplicate, or that the user should be doing something a different way than what they are asking for. There seems to be so much effort put into deliberately not answering the question that trying to participate is truly deflating.
Can you elaborate? It's difficult to agree or disagree since I can't tell what you have in mind. What does 'Toxic' mean to you in this context?
Since most questions will not be "high-quality" (for any reasonable definition of "high-quality"), most new participants' experience will be of asking a question and being told "that's a bad question" either explicitly (downvotes) or implicitly (editing their question to "improve" it).
It gets worse as time goes on because of the "no duplicates" policy; the more questions that are in the system, the harder it is to come up with a question that isn't marked (correctly or incorrectly) as a dupe.
Banning all politeness words (hello/goodbye/please/thank you/you're welcome) from questions/answers because they're "noise" is just the cherry on top that makes people who aren't already immersed in SO culture think that everybody there is rude/mean, no matter how many times they put "be nice" in the CoC.
Not saying these are _bad_ policies per se - SO has a stated goal of being a "curated resource of high quality questions and answers", and they need to separate the wheat from the chaff somehow, but the policies _will_ be perceived as hostile by most new users.
Giving the questioners an extra +5 fake internet points will not fix any of that.
I wasn’t aware that politeness was banned but I avoid asking questions there even if it means joining free node channels and waiting hours in hopes that someone answers my question.
S/O has some very high quality responses but I actively avoid asking questions there because it feels over policed.
I'm curious why they don't have a no duplicate answers policy.
They often differ slightly in their assumptions or implementation, which can help in grokking the core concept of the answer and not getting distracted by implementation details, especially for languages and domains I'm not very familiar with.
This describes the problem with SO remarkably well - how it constantly trends towards being even less nice than it was two months beore. I've gotten SO points and even been congratulated over some questions and some answer. But I'm still angry, seven years on, over having one particular question being deleted - it was my question and it was a question I wanted an answer to it.
> We know that people who contribute by asking have a harder time earning privileges than people who focus on answering. Independent research suggests this disproportionately affects women:
> > We also see that women contribute differently to building the community’s knowledge base: they are asking more questions. Stack Overflow’s current system strongly incentivize answering by rewarding upvotes on answers twice as much upvotes on questions.
Question askers get the reward of having their question answered. I think votes on questions should be done away with entirely. Questions that should be exemplified can have some other metric (stars for "this is well written and generalizes easily for other users"), and low-quality questions should be flagged or closed, which at least gives some sort of reason. Downvotes aren't useful to askers because they have no idea what they're doing wrong when they receive them.
Also nothing is stopping a downvoter from leaving a comment explaining why the question needs improvement, other than lack of reputation (you need a minimum reputation for your downvotes to be visible anyway). People who care to explain their downvotes can already use the existing process. Since comments can be upvoted (but not downvoted), there is already an incentive in place for that as well.
I don't really get the point of the vigilance toward bad but not intentionally spam questions. The question # has been used, derank it and call it a day until someone sees a reason to revive it.or the asker ASKS for help improving their question.
I think a rule could be "if someone with a decent track record (for some definition of decent, maybe 100 points? 1000?) felt the need to ask it, it's not a duplicate." The bar could be raised in other ways as well, but this one I think would reward people who have done the work to be a positive part of the community.
We all have varying degrees of ability to generalize the principle of one answer to another. The other part of this is that you get better at asking questions through practice, which is hard to get early on. At this point, most problems I can't answer are good questions or too specific, but it took a while and I don't want SO to be the place where only those with 5-10 years in can get help.
SE spends half the time saying an overlapping diversity makes it easy to find a question and answer that fits and the other half eliminating that diversity (and bringing non-adept people to dead ends in their searches that land on SO.)
I see it as a way to meet in the middle.
Answered questions would be fine having their vote button disabled.
https://stackoverflow.com/questions/2170528/writing-hello-wo...
Unfortunately, the moderators and I had a disagreement about the importance of asking questions without attempting to first solve the problem.
Well written (and well-defined) questions are the key StackOverflow. From the questions, you could extract a Rosetta Stone of programming. An experienced programmer could learn most languages on the job.
To curb race to first asker, SO should implement exponentially discounting reward. This makes final commulative reward finite.
Newbies should help newbies.
There are lots of ways to incentivize this, but the basic gist is that users who are just growing out of the newbie phase are in the best position to be indoctrinating new arrivals. They'll have the most patience and understanding, and it's an easy job that shouldn't be done by super experienced users.
Most of this could easily be isolated to a newbie pre-stage too. Basically, a tutorial area in which new users work together to make their questions better so they can move on to the non-tutorial experience.
Personally, I'd filter these posts out by default, but on days where I was feeling nice I might go in and help out.
Worst case scenario: it takes some extra work from a newbie to get their first bits of rep.
I have. They say her behavior from start to finish was incredibly rude and explicit in stating her plans to ignore the policy.
Months ago, well before the current Stack Exchange pronoun policy was a thing, she said (yes, explicitly) that she wasn't willing to use singular "they" because it feels so wrong and confusing to her. This wasn't a "plan to ignore the policy" because at that time there was no policy.
(It also doesn't appear to have been motivated by sexism, transphobia, etc., although people who say such things sometimes are; she is explicitly willing to use neopronouns -- "xe" and the like -- for people who find that "he" and "she" don't fit them.)
Then, much more recently, the upcoming change to the Code of Conduct was announced in the moderators' chatroom. M.C. asked some questions about that, very definitely did not state any plan to ignore the policy, and promptly got de-moderator-ated.
It's possible that she isn't telling the truth about what has happened in private, away from the view of other moderators. Maybe she sent an email to SE management saying "I'm not going to abide by this; fire me if you dare" or something. But as far as externally-visible things are concerned, her account of things matches up better with reality than SE's statements do.
There are, indeed, moderators who think M.C. should have been canned long ago, but my impression is that there's something close to unanimity that SE should not have canned her the way they did, and that the majority of those with any opinion on the matter think that they shouldn't have canned her at all.
(Source: am a moderator on a Stack Exchange site, have read all the relevant transcripts from said chatroom. Note that discussions there are confidential, so the above is about as specific as I am prepared to be about who said what.)
Because she at least offered a very detailed breakdown of events that can be fact-checked, critiqued, etc. StackOverflow has offered basically nothing more than vague PR BS.
https://cellio.dreamwidth.org/2019/10/05/stack-overflow-fias...
And a ton of them.
Net value loss.
I might be pessimists but I think they knew exactly what they were doing and are happy with results.
And I am guessing new SO leadership wants the same.
Getting at great contributions is now much harder. I use Quora much less now.
Churn 'n burn for traffic dollars is a strategy.
Suppose it opens a door somewhere too. Do that enough, and there is clearly room for a less noise filled source.
IMHO, the difference here is Quora has a plausible, casual user case that makes sense. The people who helped to build Quora content won't give two shits and will move on, but Quora itself may do just fine. Meh.
Degrading the value to current users of SO to attract more in general, leads where for SO?
People who see a question as relevant and interesting to them, would add points or monetary incentives. Thereby increasing the speed and quality of the answers.
This mechanism has already been tested out and actually works only with crypto currency due to low transaction costs.
Better explained 2 mins video: https://youtu.be/lcsgx_5c7D0
Here is an example: What makes you nostalgic? Answer with #Obyte Know-it-all and earn up to 410,172,272 Bytes : https://t.co/EaVFrscFC7 (enough Bytes to make > 700,000 transactions whenever you want in the future thanks to obytes' tiny predictable fees)
And I notice a lot of female employees or managers being mentioned which might be a coincidence but it also might be deliberate and somehow be related to the idea that females might have more empathy? Which could just be my own prejudice and not true but I dunno.
But anyway, if they are going for nicer/more empathetic employees and policies, that's great, but its not going to stop all of the (mostly male) super-critical narcissistic arrogant shaming-happy vote-whores who are eagerly waiting to get in another passive-aggressive veiled hint that I am an idiot while they downvote my question.
Not planning to ask another question on Stack Overflow anytime soon, thanks though.
But also, less than 10% of Stack Overflow users are women.
I've never asked or answered a question on SO (I wouldn't dare!) because I can plainly see the horrorshow that is likely to come with doing that by reading other questions. Over time, I've come to simply ignoring SO for the most part.
I think I may check in from time to time to see if the place actually improves.
Who decided that software developers should have the final say over how people interact? I feel like there's probably a field or two of experts on that that should be being consulted.
A far more meaningful way to reward askers would be to simply change how dupes are handled. Instead of closing questions with a link to a related question, give users a mechanism to answer questions with a link to a related answer. Then if the link really does solve the problem the asker can accept that answer, and if not the original question can stay open.
I get why SO chose their current aggressive stance on dupes, but it's probably time to try something else.
Work is under way here and elsewhere: https://forum.codidact.org/
Based on my past actions I guess I would be happy to donate a few dollars to anyone who makes a better q/a site and makes me aware of it.
By better I mean:
- important current database (allowed with current licensing)
- rewrites the meta game to make helping (answering) more valuable than administering (deleting).
- raises bar to asking questions (stack overflow seems to always be crying over all the stupid questions they get yet the thought of raising the bar for asking questions never seen to get any traction)
- have a karma tax or something on questions and contributors: just because someone was the most valuable contributor or something was the best question two years ago doesn't mean they still are.
- don't split into many subsites for every topic, if people want to ask it and other people want to answer it and it isn't illegal or scary or anything, bust leave it alone. Karma tax will take care of it in the long run.
It took them 7 years to change their minds on it. Sigh.
---
Side rant: This whole piss people off/break things and expect people to stick around is a dumb idea. Company does stupid things, I'm going to find an alternative company and not return.
It is beyond me why they would want to award spammers for creating more spam.