It's often said on this site that meta means murder and that this acts as a deterrent. Our site, which is supposed to act as the central clearinghouse for network-wide matters, doesn't have a great rep despite the good efforts of many people here. Using a system designed for objective(ish) Q&A, "no chit-chat", for discussions about functionality, policy, seasonal fun, and strongly-held beliefs probably isn't helping.

Does it have to be this way? Meta.SE is already different in some ways from both the per-site metas and all the other main sites. Are there other things we should look at tuning? Can we reduce the bad experiences while maintaining the quality that keeps everyone here and ready to help?

I've seen some things here that make me a little sad inside, cases where it feels like frustration could have been avoided with better...something. Things like:

  • Support requests getting downvoted to oblivion, so they drop off the front page and don't get answered.
  • New users deleting their accounts after asking an unpopular (but not poorly-asked) question.
  • Questions that raise a reasonable point for discussion but don't follow some of our unwritten rules, so feedback focuses on language, formatting, and other flaws, instead of edits or constructive comments.

(I'm deliberately not linking to examples. I'm not going to draw even more attention to specific Meta trainwrecks. If you've read this far, you've probably seen them. If not, it probably won't be long before you see another one.)

I don't think these things happen because Meta is full of people who enjoy pulling others down. We've got a lot of great people here, people who tirelessly answer support requests and help figure out bugs and bring new ideas to discussions and patiently explain, again, why that thing the question asked about happens. I've been here a while and I've seen a lot of people helping people. And yet, we also have the less-positive things I've mentioned.

This makes me think there's something systemic that's contributing, something that works against good people trying to do good things.

This is a broad discussion, I know, but I think it's time we talk about what we could do to make Meta work better for its unique-on-the-network goals, without endangering the things that work well. Because many things do work well; Meta may mean murder sometimes, but it also means people working together to make our sites and the software that runs them better. Can we be less murderous without being less effective? What would we need to change?

  • 5
    Tim Post has made an excellent response here. May 14, 2019 at 3:47
  • 11
    Is this even just Meta? I know I actively try to avoid going to certain sites or chat room even when it would make sense to just because II feel uneasy to have to interact with some of them and their "culture". IMHO Meta.SE is not a problem, just one of the many effect of a broken community.
    – SPArcheon
    May 14, 2019 at 8:34
  • 6
    @SPArchaeologist other sites aren't the only way for you to get answers to your questions, though. If you don't want to post on SO, you can find programming answers elsewhere. (Maybe not as good, but...) If you want to participate in shaping how things are done at SE, Meta.SE is your singular option. (Yeah yeah Twitter, but not what I meant.) May 14, 2019 at 14:32
  • 4
    @MonicaCellio What I meant is that recently I feel discouraged in participating in many sites and communities over the network, not just on Meta.SE. My point is that IMHO we have a far larger problem than just a "murderous Meta" - the whole community cannot seems to just live in harmony - we are too busy blaming and attacking each other. But again, that probably just me. Didn't meant to derail your post, so I won't post an answer either.
    – SPArcheon
    May 14, 2019 at 15:42
  • 1
    Anyway, just a side notice.... I never stated that "getting answers" was my only purpose. Often I would go to chat just for the fun of it, no answer involved. Lately, I noticed that I can have fun only in an handful of the rooms I used to visit. Forget this nitpick, but your comment seemed to assume that I am not happy with the content or the great answers we can provide on the network, where my discontent is just about having to keep seeing people begging people to be more kind... and fail.
    – SPArcheon
    May 14, 2019 at 15:48
  • 8
    So true. I did a bug report once and got hammered for it. Lots of down votes and ridiculed for not knowing the underlying problem. It turned out to be a duplicate (I had done a search but used the wrong keywords) so fine, but I didn't need to be treated like an idiot. Makes me very unwilling to ever report another bug (or site request) again. And so far I haven't, even when encouraged to.
    – Cyn
    May 14, 2019 at 18:17
  • @Cyn downvotes are one thing, but were the ridiculing comments get flagged for abusive language? May 14, 2019 at 18:27
  • 4
    @SterlingArcher they didn't cross that line. I really don't want to go look at it again. I did count up the downvotes (from my rep cause I can't see them directly) and it was 10 I think. Certainly it could have been a lot worse but seriously, on regular sites if someone reports something and it turns out to be a dup, people just mark it as such and move on.
    – Cyn
    May 14, 2019 at 18:35
  • @Cyn personally, I just learned to stop caring about the votes. It's ok to be disagreed with -- but typically being ridiculed is a form of abuse. This is just how I do things though, doesn't work for everybody. May 14, 2019 at 18:37
  • 9
    @SterlingArcher Well I've been mostly off Meta.SE because of this and don't post questions anymore. Now I'm just talking about it to support Monica's post. I wouldn't care about the votes if it were 1) a real Meta site (where the votes don't change your rep) and 2) if they reflected a vote on my proposal or etc. Who downvotes a bug report? And even if there was an underlying reason for the problem, it is a bug that the user isn't getting correct feedback as to why the action failed. I honestly don't need to hear from anyone else that I "did something wrong."
    – Cyn
    May 14, 2019 at 18:45
  • 1
    "I don't think these things happen because Meta is full of people who enjoy pulling others down." I wonder, though, if you might think differently if it had ever happened to you. What do you suppose people who made suggestions that got dozens of downvotes think about this? I, for one, can certainly tell you.
    – user142148
    May 15, 2019 at 7:38
  • As a thought experiment, I wonder what Meta...heck, I wonder what all the Stack Exchange sites would be like if they had upvotes and no downvotes. A bad question just...wouldn't get any votes. Or what if there were downvotes, but a question wasn't allowed to get more downvotes than upvotes, so the worst score it could get was 0? Do we need to know the difference between a -1 question and a -50 question? Certainly it's a common theme here that people downvoting don't have to say why, but also that a downvote doesn't really explain anything. Maybe comments only would be more effective.
    – user142148
    May 15, 2019 at 7:45
  • 7
    @Kyralessa As for what all the SE sites would be like: I moderate one where awful questions (from a site quality perspective) do get upvotes, probably because of their 'entertainment value'. As for 'allow no more downvotes than upvotes': I don't want to have to write an userscript that casts my downvote on something as soon as it gets an upvote, just because I think it shouldn't have that upvote (in other words, I don't think that will work either, as people will then just wait with downvoting until they are 'allowed to'). So why scratching downvotes here might work, please not networkwide.
    – Tinkeringbell Mod
    May 15, 2019 at 7:52
  • 4
    I do understand where you're coming from (I saw it too). But there's a problem with 'a bad question just wouldn't get any votes'. Also, just dropping downvotes on their own is not that much of a solution, as long as reputation loss for a downvote doesn't cancel the reputation gain from an upvote, and posts can't go below 0, on main sites that means a 0 scoring post can still gain you considerable reputation, so you'd have to get rid of reputation too, and find some other way to indicate that people are writing good content.
    – Tinkeringbell Mod
    May 15, 2019 at 7:56
  • 1
    I agree that downvotes serve a purpose, and some downvoting functionality should remain, but @Tinkeringbell, I think piling on excessive downvotes on questions (-10, -16, -28?), which happens far more frequently on MSE than anywhere else on the network, goes overboard, and seems to be a "pile-up" effect of "i see others think it's bad, so me too!" (The "contagion effect".)
    – amWhy
    May 16, 2019 at 16:33

11 Answers 11


This is a topic near and dear to my heart...

This makes me think there's something systemic that's contributing, something that works against good people trying to do good things.

That thing is people. I'm not saying here that people are inherently evil and any non-trivial concentration of people turns into some sort of carefully-concentrated balsamic vinegar of badness*... But rather, that the process of meta-discussion invariably turns people away from their shared goals, distracts them from their original purpose, and invites them to fixate on trivialities. Then that old narcissism of small differences kicks in, and folks refuse to recognize the humanity of those who don't smoke the right cigarettes, er, repeat the same memes. And then nothing gets done.

Jeff observes the same in that old blog post you linked:

Meta-work becomes a reflex, a habit, an addiction, and ultimately a replacement for real productive work.

So you might well ask, if this is so well known - by the site's founders, by current staff... Then why have this site - why have 170-some meta sites - at all? Jeff hints at this in a follow-up blog post:

We now know that meta participation is the source of all meaningful leadership and governance in a community, so it is cultivated and monitored closely.

So, what then? Meta-work is evil and leads to pointless bickering, but is also the source of all meaningful leadership? This sounds like a recipe for painful cognitive dissonance, not productive, harmonious collaboration! WTF, Jeff?!

Sharpening your saw without keeping your nose to the grindstone

I found the answer to this in one of those stupid corporate fables:

"Excuse me Mr. Lumberjack, but I couldn't help noticing how hard you are working on that tree, but going nowhere." The lumberjack replies with sweat dripping off of his brow, "Yes... I know. This tree seems to be giving me some trouble." The bystander replies and says, "But Mr. Lumberjack, your saw is so dull that it couldn't possibly cut through anything." "I know", says the lumberjack, "but I am too busy sawing to take time to sharpen my saw."

I like this story in spite of its provenance because... It's true to life. If you've ever spent a day cutting anything, you know the importance of pausing now and then to sharpen your tools - you will literally never get done if you fail to do this, no matter how hard you work. But... You can't spend all of your time sharpening either - at some point, you have to put the stone away and start sawing.

Meta is a good place to sharpen saws. But spend too much time here, and nothing gets done. And eventually... We'll find we've ground the saw down to nothing.

The solution you seek then lies in finding a balance. Meta need not be murder... But the dose makes the poison.

Granted I am a babbler, a harmless vexatious babbler, like all of us.
-- Fyodor Dostoyevsky, from his book about Meta

To every thing there is a season

As we all know, this problem is hardly limited to a handful of sites on the Internet (or 19th-century Russian bureaucrats). Each of us must find a way to balance myriad meta matters in our daily lives, whether that be cataloging bugs vs fixing them or literally sharpening tools vs. cutting. This thing we struggle with here... Is in many ways a fundamental part of the human condition, as familiar and unavoidable as death, sore feet and insufferable government officials.

And we know that nearly everyone struggles to find this balance to some degree, with varying amounts of success, and with various points at which each of us deems ourselves sufficiently "balanced" and kinda just accepts our life for what it is - perhaps even taking delight in our own flaws, until such a time as they make life intolerable and we rage-quit.

If it doesn't spark joy, set it on fire

We know then, that just as some folks will cheerfully live in a messy house while others will clean and organize obsessively without stopping to do anything else, so too must there be a distribution of meta-users ranging from those who would participate here only under duress to those who can't seem to stop themselves even when they should be in bed asleep. This is normal, expected, and neither a problem nor a solution to a problem.

What is a problem is that, after 10 solid years of running these meta sites... We still don't do much to set expectations for how they should be used. There's, uh... A short help center article, a FAQ and a bunch of frequently-asked questions, but when it comes to the basics...

  • How do I start a constructive discussion?
  • How do I participate in a discussion constructively?
  • How do I report a bug?
  • How do I propose a new feature?
  • How do I chime in on a proposed feature from 10 years ago that still hasn't been implemented?
  • How do I get help using one of these sites?
  • When should I upvote / downvote / edit / flag / close / run in circles scream and shout?

...we kinda just expect folks to poke around and figure it out. And then get short with them when they don't. In other words, we have the same problems here that we have on normal Q&A sites - high expectations, but poor support. We're spending more time blaming people for using the site wrong than we are trying to help them use it right...

...and I include myself and my co-workers in that "we". Here at Stack Overflow Exchange Global Corporate Headquarters, we've been lamenting this problem for years... But lamenting it doesn't fix it. Some of the suggestions in that thread are just as good now as they were 4.5 years ago, and... also just as unimplemented. That's effectively several new generations of members who've learned to use the site - or left in disgust - with little if any additional help.

And that's on us to fix. Ultimately, that's why I'm taking the time to write this answer. As Jeff wrote all those years ago,

Half of community relationships isn't doing what the community thinks they want at any given time, but simply being there to listen and respond to the community.

Getting new features or changes built out into the software can be a challenge at times, but listening is something we - I - am paid to do here full-time... And I have to confess, I haven't always done the best job there. There's been entirely too much blame being thrown around, particularly over the past year... And far too little listening, acknowledgement, and assistance.

So in the spirit of changing that, here are a few things that you - and I, and anyone else reading this - can do personally to help make Meta a bit less, uh, murdery...

Four tips for kinder, gentler meta work

  1. Focus on goals and problems, not style and personality
    We give a lot of lip-service to this idea, but it's easy to forget in the moment when someone is proposing (or announcing) a change that might affect a tool we rely on daily... Particularly when the writing style comes off as arrogant, or the person writing seems to lack the necessary experience to adequately judge the effects of what they're proposing.

    But just because their suggestion is... Not good... Doesn't mean it wasn't inspired by a real problem or a worthwhile goal. If you're able to recognize that underlying motivation, you can address that instead of just shooting down their idea.

  2. Edit, edit, edit

    I've written about this before, but it's still a good idea: if you can jump on a question that's likely to do poorly before it's too far gone, you may be able to fix it. A heroic edit is not a trivial undertaking... But for those with the skill and experience necessary, it can be one of the kindest and most effective acts.

  3. Answer, don't comment

    This is definitely one that I need to get better about following myself, as do many of my colleagues. We know good and well that long comment threads under a question are rarely productive... So when we engage in them anyway, it just sets a bad example for everyone else. This doesn't apply to linking related posts, asking for clarification, etc... But as soon as you start to write an opinion or provide anything approaching an answer - consider taking the time to write it out as an actual answer.

  4. Focus on your own votes, not others'

    So much has been made of voting on meta over the years... Usually other people's meta voting. Because of course every single vote you and I cast is well-considered and justly deserved, right? It's those other people who are just voting all wrong.

    Except, we have a LOT more control over our own votes than we do others'. And none of us is using anywhere near as many votes as we could. Some of us are barely voting at all. So next time you see a post that you feel is under-rated (or over-rated!) take that as a reminder to vote on it... And on every other post you've read and neglected to vote on. And I'll try to do so as well.

Haven't you said all this before? What's the point of writing it again?

Yes, I've babbled here a lot - more than almost anyone else. So in a sense, anything I write here about meta participation is as much self-flagellation as it is anything else... But I never took the time to post a Dostoyevsky quote about this activity before, nor have I ever used that "spark joy" meme anywhere, so I felt it was high time I remedied both omissions. I'm frankly amazed you're still reading after that. As your reward for making it this far, here's a footnote that directly contradicts my first paragraph:

*they are though, and do.


I'm pretty sure comments are a key factor to how any post goes. We already know that for all other sites, it is no different here on Meta.

The few posts I participated on did go south due to a well placed comment, either by me or someone else. Having an up vote on a comment that is not in favor of whatever the post tries to discuss is bound to start the band wagon of votes. It needs some pretty good counter comments or an excellent answer to make that stop.

My personal running sore is users who contest down votes in comments. 9 out of 10 these comments, either direct or indirect, accuse the voters of doing something wrong. I vote often and early. I'm sane. You can trust me my vote means something. I engage in comments when I spot such comments. That often means I add more reasons why a post isn't that useful. If anything, I will not give up this behavior. It isn't helping nor is making down voters feel bad.

There are "tainted" topics. Anything on "be nice", "hot network questions", "(down) voting", "announcements" to name a few can go down the drain in minutes. I'm exhausted on these topics. They have no real solution, it will leave part of the population disgruntled, no matter what the outcome will be. So yes, these topics means murder. Maybe their tag wikis need extra guidance, so posters who care come prepared.

One of the issues on Meta is the wrong expectation users have when they reach out on Meta. For a support or bug report you might find something useful but anything tagged discussion or feature request is at best getting a handful of voices heard, maybe some more on really controversial topics (I think licensing, CoC, twitter drama). There is no visible or measurable progress on any of these topics. I personally have no problem with that. Network wide policy is complex and needs to be slow so you don't break anything. On top of that Stack Exchange is a privately held company and we can't know their objectives nor do we need to. But enough users posting here either expect or even demand an official stance from a CM or SE Staff. On Meta you're not communicating with Stack Exchange. Your post is not going to be the talk of the day at the coffee machine at the New York head quarters. That is fine. Be happy if your suggestion makes it into a policy or if a feature gets implemented. Don't go nuts if you get no response. Send an e-mail or tweet.

Can Meta be less murderous?
I doubt it. I'll watch myself leaving comments. That will be my first step. And I'll post an answer instead of a comment if I feel my position on a topic might meet both support and opposition to prevent only the supporters have a vote on my comments.

  • 2
    here's a follow-up suggestion "show all the comments collapsed by default..."
    – gnat
    May 14, 2019 at 12:25
  • 2
    "Your post is not going to be the talk of the day at the coffee machine at the New York head quarters" Or it might be, just as the subject of ridicule instead of being discussed constructively, according to comments overheard by Tim Post (or it might have been Jon Ericson, I can't recall).
    – TylerH
    May 14, 2019 at 21:11
  • 3
    "The few posts I participated on did go south due to a well placed comment, either by me or someone else." I think you're inferring causation from correlation. By this point, I can predict whether a meta question will "go south" based entirely on the question itself. I have only rarely seen a question attract more downvotes because of bad behavior in the comment section. It typically starts with bad behavior in the question. Or at least bad implications in the question. May 15, 2019 at 2:59
  • 3
    @NicolBolas that is a fair point. I still think that a bit too often comments aren't used to ask for clarification about the post but instead voice the prediction the post will go. Those comments can act as an accelerator, no?
    – rene
    May 15, 2019 at 5:16

I wasn't going to go out on a limb but this answer really clicked with me...

I'm pretty sure comments are a key factor to how any post goes... The few posts I participated on did go south due to a well placed comment...

...and I would like to follow up on that.

Above observation rings true to me, I also have an impression that most of the voting that felt a bit off (especially voting down) was somehow triggered / correlated with comments. I am not fully certain if this is indeed so or I am mistaken, however I would be very interested in running an experiment that could probably help us estimate impact of the comments on voting patterns here at meta.

Specifically, I suggest to show all the comments collapsed by default (yes, even when there is only one comment). I would like to learn if this would have a noticeable / substantial impact on voting, editing and answering over here.

Note to those who may not know, turning such a feature on and off doesn't seem to take any dev effort and is already well tested.

(For the sake of completeness, when writing this I was also considering a "partial" version of above experiment - one that would be limited to only posts tagged support and bug - but in the end decided to drop it in favor of more straightforward proposal. Though we probably can reconsider it later if it turns out that total collapsing of comments breaks things too much.)

  • 2
    one caveat, I expect that suggested experiment (if it lasts long enough) will substantially change meta culture and it is going to lose something that I am used to and something that I learned to like. That makes me a bit unhappy proposing it but I am too much curious to give it a try and see how it would go
    – gnat
    May 14, 2019 at 12:24
  • Similar feature suggestion (view the Stack Snippet in fullscreen mode) - Full disclosure, this is my own suggestion.
    – TylerH
    May 14, 2019 at 21:15
  • 4
    FWIW, I think the experiment would be much better applied to discussion posts rather than feature requests or bug reports. Discussion posts tend to go down the comment drain much more often and much faster than the other two, though feature-request may benefit from it some, too.
    – TylerH
    May 14, 2019 at 21:21
  • 1
    @TylerH you have a point. And... that's why i picked the simplest approach. I first tried to figure how to justify preferences to limit experimenting to particular tags and it turned out complicated and I decided it would be simpler to just start with all of them and only drop collapsing in particular tag if we discover experimentally that it suffers too much
    – gnat
    May 14, 2019 at 22:17
  • 1
    I agree this sounds like a great experiment. SE already tells us that comments are second class citizens - maybe if this was made yet more obvious in the UI, people will be more inclined to write their dissent as a full answer instead..
    – Em C
    May 16, 2019 at 4:10

Does it have to be this way?

I'm going to go against the grain and say... yeah, it does. There are several reasons for this, and your post only brushes up against many of them.

First off: well over half of the people who come to Meta (particularly MSO) to ask about some aspect of a question or post they made (why was it closed/downvoted/etc) will not listen to reason. Many of them don't want answers; they don't want correction. They want to complain and have their complaints validated. The only useful way to deal with them is to downvote and move on.

To those who say that answering is better than commenting, no, it's not. To the unreasonable user, an answer is just a bigger target than a comment. It almost never helps. Whatever they would say in response to your comment would be said in response to an answer.

Second, questions that suggest ideas/changes to the working of the site. I'd say that over half of them on MSE are garbage (and well over 90% on MSO). Typically, they're off-the-cuff suggestions that the user didn't think about at all. They're frequently antithetical to how the site is supposed to work, non-functional, or have obvious and pretty horrible side effects that are far worse than the "disease" they're trying to cure. And then there are the ones that have been asked many times already (MSO gets the "force comment with downvote" question at least once a month).

And similar to the first group, they usually don't want discussion. Most of them want validation and agreement; anyone who dares to criticize, whether in a comment or an answer, is opening themselves up to being call all kinds of things. Elitist is a popular insult, but there are too many to list. Though a few days ago, I was call "evil" on this very site, which was kind of a new one (the user was sanctioned, so I've got no grudge).

Meta is where decisions get made by the community. And the only way for that to function is if the community can discuss things without all of the crap. If you're operating in a system where the vast majority of interactions are trying to tear the entire system down, you play defense. Anything that is not obviously and clearly good is treated as an attack.

Just look at lower volume Meta sites and you can see the difference. There, well-written questions that are rejected by the community still get upvotes. They haven't had to deal with people asking the same question 30 times a month. They haven't had to deal with trying to debate the same issue over and over and over again. So they're a lot more tolerant of this stuff.

Meta is "murder" because it's the only way for meta to function. If you want to fix the system, then you have to get rid of the crap. Just like with Stack Overflow, the best way to make people friendlier is to keep bad content from showing up in the first place. The constant avalanche of bad questions and bad behavior on Meta habitualizes people to downvote. So it's no surprise that downvotes flow very freely.

When reasonable posts become the norm rather than the exception, Meta can change. Until then...

  • 2
    On this site, MSE, I don't see this happening very often, and in the few cases where it does, I'm often able to deal with it reasonably. Save blatantly off-topic posts, most of the posts here are reasonable. May 15, 2019 at 3:07
  • 4
    Don't get me wrong, though...I do see what you're saying happening fairly often on MSO, and I have a few gripes about it...but this question is specific to MSE, where it doesn't really happen that often (for instance, it's not so that "98% of feature requests are garbage"). May 15, 2019 at 3:09
  • 2
    @SonictheInclusiveHedgehog: For MSE, 98% is too high. But it's still an exceedingly high percentage compared to what it ought to be. And MSE doesn't exhibit the problem the OP discusses to nearly the degree that MSO does. May 15, 2019 at 3:14
  • 5
    My question is about MSE, not MSO. May 15, 2019 at 3:45
  • 4
    @MonicaCellio the reasoning here looks relevant for MSE to me
    – gnat
    May 15, 2019 at 6:12

TLDR: Meta is scary, and we should consider disagreeing by posting answers.

There was an extremely long conversation (part 1 and Part 2 for context). I'd like to say - I'm posting this as a regular meta user over being a mod.

There's a lot of it, but in short the idea folks seemed to have was that meta was ... kind of broken in many ways. We have too many secret handshakes, topical landmines that new users (or even experienced ones) shouldn't step on and so on.

Considering its origins and how Jeff felt about it, I guess it did... ok. Its probably scaled terribly over time though.

Meta's somewhat overloaded. It worked ok with a smaller site and a BDFL, but from a moderator's perspective, stuff like FRs and such don't really work that well. As a point of contact between users and the company, there's currently a lot of issues, small and big in terms of tone, how people react to each other, and various disconnects. Its a work in progress but that's kind of the social debt we have.

Its a meh way to post feature requests (and there's no great way to check if a FR was accepted in some form, and/or fixed), its got a learning curve that feels like a cliff, and its nearly impossible for new users to get involved without learning a lot of unwritten rules. I personally took ages, becoming a mod on a major site, and some minor drama I absolutely had to navel gaze to get me engaged enough to aim for 10k.

While fundamentally reimagining what main meta could be would be nice, I do feel that SE's not had great success in re-inventing non core products. While it would nice to be able to bolt on better ways to handle FRs and communication, its not likely to happen without a push from the top.

Fundamentally though, meta feels broken in a few ways.

Some of it is semi technical - we have had some fundamental features that other metas have dumped (reputation), that we both can't live with, and can't live without. A new user coming in with a post that's reasonable but unpopular is not going to think of "eh, I'll take a hit for what I feel is the greater good". They'll go "THEY HATE ME THEY REALLY HATE ME" over disagreement. Finding the balance between downvotes as disagreement, and downvotes as pillory is pretty important. While reputation is sacred and any attempts to change the reputation system's unpopular, I would probably debate either taking away the penalty for question downvotes (We do have other quality management tools - a closevote is as good, and downvotes would still allow for deletion), or conditionally considering using the community wiki on controversial posts of community importance. Neither option seems perfect but would probably reduce pain for new users.

Some of it is social. We've had breakdowns in communication, and sometimes with good intentioned changes not quite working out. I got the idea of there being a lot of frustration - and talk us vs them, and meta being scary for people who work here and occasional failures in communication making things worse.

There's also a feeling with a lot of folks about a disconnect between the company and the users, and meta isn't that valued. That's something that's not really fixable at our level, but least from my perspective, is a little better than it was.

The FAQ used to say....

"Be tolerant of others who may not know everything you know."

I've asked myself a few times, both as a user and as a moderator if we've failed there.

Disagreement is an essential part of discourse. A fundamental problem with downvotes as disagreement on meta is that we've added a cost to it. I personally don't care - I've got broader goals for being in meta. On the short term, if I take a small reputation hit, its not much in terms of my overall reputation. Even a -30 on a post is something I can make back in a good month or so. For a new user this can be devastating.

Meta isn't just documentation (though its useful for that) or curation. Its people.

One of the things that I've wanted to fix as a mod was the more toxic aspects of meta, and meta chat. People get frustrated, and angry, or worse scared. Its a hard squishy problem to fix.

I'd love to go "hey, lets turn off reputation loss for question downvotes! Tadah! Fixed! Scooby snack plz"

But I suspect its a little tougher than that. Effective site meta was folks posting answers -whether or not they were a moderator. We need folks to actually both step up and help as well as folks showing a little more empathy for well, folks who don't know everything we know. I think part of the trick to breaking the cycle is to stop letting us affect it as well.

I'd actually suggest we consider how we react to things we disagree with. Lets make a case for why we disagree as answers. If an idea is terrible, lets talk about why, and how it could be better. Upvoting those (and not piling on the downvotes on the questions) probably means the questions don't sink into the mire and folks can debate them more constructively. Some stuff is just awful, and downvoting that is fine.

So yeah, I don't really think downvoting as disagreement on questions works very well any more.

  • 1
    The first sentence is good, use your effort to post a better answer than make criticism. Behaviour isn't modified by penalty for some people, and lessons not learn by teaching. Your answer is shorter than Shog's, that's also good.
    – Rob
    May 14, 2019 at 11:16
  • 1
    "Lets make a case for why we disagree as answers. If an idea is terrible, lets talk about why, and how it could be better." Um, why? I've posted many such answers, to no avail. Garbage ideas are garbage, and there's no "better" to things like this. Posting answers to such questions is just a waste of time. I mean, it's a great way to get Reversal badges, but other than that? It accomplishes nothing productive. If you want to improve Meta, stop the terrible questions from happening. May 15, 2019 at 2:29
  • 2
    I do have a bunch of those. But primarily it's also a way to teach and inform in terms of meta May 15, 2019 at 3:13
  • 2
    @NicolBolas There is a nuance between approving that an idea is bad, than to murder it. I do agree with you that posting an answer in such question don't help, as the user can't remove the question after if the answer is upvoted (or for the community to get the question deleted). The OP just feel like it get crucified in public to stay there. (and in your case you linked, it's exactly the problem IMO, he know he will be downvoted to hell, but his feedback would stay there forever)
    – yagmoth555
    May 15, 2019 at 13:03

Motivation & Enervation

My sense is a lot of the regulars on Meta come here and hang out here because there’s not a lot on offer on the main sites for them anymore. I know it’s true for me.

Like a cop on a beat, when you’ve hung around on Main enough, you start to get a sense of endless drudgery. That you’ve seen it all, that each new question is a microscopically small take on dozens of conceptually identical questions that came before it. That your Main life is tied up in endless, grindy curation, which you do only because there’s nothing else to do.

There are not enough interesting, meaty fresh questions which are challenging and interesting to answer. Which is the main purpose of most of the Meta-type people, and ostensibly (to your point and Jeff’s before you) what our main purpose should be. It should be 80% content creation, 15% curation, and 5% meta discussion.

Iteration & Desperation

But that 80% isn’t there. So the next step in the traditional path for long time users is to spend that (now) 95% time curating. But that’s dull and soul-sucking, and so you come to Meta to see if there’s some solution to it, and low and behold at least there’s something new here. For a while. And then you spend about 100% of your time on Meta.

But, just like Main, once you do that long enough, the same beat-cop pattern emerges: it’s new for a while, but then you start seeing the patterns, and every new question is the same as the last, and you came here to get away from that, and now you’re sick of that too, and you start curating on Meta... and Meta becomes hostile.

Orchestration & Regeneration

The elephant in the room here is once you’ve been here long enough, there’s nothing left to do.

The cure is to find a way to broker the deal between the people who desperately want to create content, and the people who want that content. We are embarrassingly rich with both.

We need to somehow harness the drive of askers and channel them into productive, interesting questions, even if a big percentage of them fall out, everyone wins. We’ll go back to the business of making the internet a better place.

  • 2
    Wow, I'm sorry to hear you've gotten to the point where you no longer enjoy the rest of the network. I'm active on several sites and I know there are a bunch more that I'd be more active on if I had the time, so if one of my "regulars" loses its shine, there are others waiting. If you haven't looked around the network in a while, you might want to give that a try. Nobody should feel that coming here is the only alternative when your once-favorite site gets boring. May 14, 2019 at 22:21
  • 2
    @MonicaCellio Thanks! I realize in retrospect that my post has a downer tone. That wasn’t the message I wanted to convey. Rather, I wanted to point out what I think the common path for Metaphiles is, and how that path has led us to where we are.
    – Dan Bron
    May 14, 2019 at 22:59

Maybe we should apply a little friction or just-in-time help to questions from new users.

Here's something that happens a lot: a new user posts a that boils down to "I think this would be useful", without any supporting evidence or reasoning. Naturally, that FR gets downvoted to oblivion. People respond in comments (maybe helpfully, maybe not), and maybe a comment argument ensues. Meanwhile the score is dropping and the OP is getting more and more frustrated. Sometimes there are edit wars and attempts to delete the question, and then there's the frustration from that -- having posted a bad question, you can't easily back out and stop the bleeding.

Wouldn't it be better if we could intercept a user's first feature request and say: here are some things you need to know to succeed with that?

Any first question from somebody without the association bonus would probably benefit from some "hey, meta is different and here's what you should know" notice before posting. It looks like feature requests are particularly prone to this because "I disagree" voting is strongest there, so even if we don't try to solve the problem generally, let's consider what we can do for FRs.

Everyone would be happier if we could intercept those questions before they get posted and downvoted to oblivion. Can we do anything to guide new users before they blunder?

  • 4
    Maybe we can incorporate this into the proposal to enable the Ask a Question Wizard here? Jul 8, 2019 at 18:11
  • @SonictheAnonymousWizHog I think I have already included it in the proposal. If a user chooses „I have an idea ...“ (template for feature requests), there would be a field asking, why the change would be beneficial. Jul 9, 2019 at 6:16
  • I don't really agree. From what I see, most of the times feature requests are downvoted because people simply do not agree with the request, no matter how much "supporting evidence or reasoning" there is, or isn't. And that's exactly how meta should work. I do agree for the mentorship-like suggestion, I was part of the experiment on SO and it was really nice. Having this, we can warn the people in advance "Look, this request isn't very popular, it might get some downvotes as result. Sure you want to post it anyway?" Jul 9, 2019 at 14:05
  • @ShaWiz we get FRs that are well-argued but bad ideas (and thus downvoted for that reason), yes. But we also get a lot of the kind I described -- also correctly downvoted, but disasters we might have been able to avoid entirely. I suggest poking around in the massively-downvoted ones to see examples you might have missed on the front page -- and keep in mind that those are just the ones that haven't been deleted. Jul 9, 2019 at 14:29
  • 1
    @ShaWiz I think the point made here is that we want to prevent FR's being posted. I do believe that if some of those OP's would have had an heads-up warning they would have dropped their FR. That will marginally help I think but I agree that at least it could prevent confusion due to the expectation that any FR will be well received. It is not watertight but it might prevent one or two major mishaps per week and any collateral damage ....
    – rene
    Jul 9, 2019 at 14:49
  • @rene it's the confusion and resulting negative effects for, specifically, new users that I'm trying to mitigate. New users are still welcome to post FRs, but we should provide better guidance where they'll see it so they're less likely to post the ones that are going to fail immediately. Jul 9, 2019 at 16:09

This makes me think there's something systemic that's contributing, something that works against good people trying to do good things.

Yeah, there is.

From my observation, this is primarily because how meta is explained and cultured to deal with posts. We keep saying "downvotes are different here" but that is such a vague statement that they end up being used for distinctly different purposes. This directly leads to the situations you outlined above.

On main sites, downvotes are somewhat rare (much rarer then upvotes, for one), and they sting when they happen. If I ask a question on SO and get it downvoted to -5, I might seriously consider what I did wrong. That's a good impulse that the main sites spend a lot of time tutorializing.

Now you come onto meta. The problem is, here, downvotes can mean different things:

  • I don't want to talk about this topic right now
  • This post is poorly formatted
  • I disagree with the conclusion of your argument
  • This is a duplicate
  • I disagree with your suggestion
  • You have done some other things on meta that make me dislike you

This is shockingly different from main. People that come here aren't prepared for that and don't know how to properly parse the feedback they're getting. This leads to massive friction. And what does that result in? Rude comments, harsh rebuttals, flags and other things. This is where the feedback system fails in communicating the expectations of meta. I complained about this on Meta Stack Overflow a while ago

So what can be done?

I think meta ought to evolve and detach from the normal site engine and treat different kinds of topics differently.

  1. Bug reports and support requests should not be able to be downvoted.

The Closing system already suffices to give feedback on how to actualize posts better. Unclear reproduction steps? Close as unclear. Can't reproduce? Close for that. Already reported? Duplicate. There is no real reason to attach a further value judgement to bug reports.

  1. We need to shift from using downvotes to express disagreement to expressing actual disagreement via upvoted answers.

This is going to take more work and is probably more controversial. This is not mainly for the benefit of the OP. Discussions around features are a form of stakeholder dialogue. It's important and necessary that the different positions to consider are outlined outside of a simple number. A number doesn't provide much in the way of actionable feedback outside saying "Don't touch this" in very negative cases.

I think we need to stop pretending the exact same engine that works for main is equipped to deal with the unique challenges and pitfalls of meta.

  • 6
    Whatever your post is about, just tag it bug or support (or both), so no one can downvote it. Roll back any edit removing these tags until a moderator comes along and locks the post. Also, how to deal with retaggings, e. g. a downvoted discussion gets tagged support etc.? May 14, 2019 at 7:06
  • 7
    @AnneDaunted These are kinks that can be worked out. That sort of abuse is no harder to deal with then people deleting & reposting their question to get rid of downvotes.
    – Magisch
    May 14, 2019 at 7:08
  • 6
    The system certainly doesn't work well for actual discussion of issues; it wasn't meant to be used for that purpose. So rather than try to mangle a broken system more in an attempt to squish it into the round hole, an entirely different system should take it's place. One that actually enables and encourages discussion.
    – fbueckert
    May 14, 2019 at 14:38

If I may make a probably unpopular suggestion? Stop taking rep away after ten downvotes or so

A) At that point, it's probably occurred to the user their questions or answer isn't well received

B) At that point you also can't delete it to get back your rep

C) Voting on Meta SE seems to follow "mob mentality"; that is, if users see a downvoted question they feel they must downvote it too. That means that my maybe duplicate from 2013 or my poorly worded question could just sit there and suck away rep

As to why this occurs, I refer back to "mob mentality": what one thinks or feels all do. This is by nature a more passionate and all-encompassing site: we air our complaints, feature request, or just plain rant here because it covers the whole SE network that we know and love/hate. But if a few users disagree strongly with a question or answer, they'll hit that vote button. Then someone else sees it, and seeing the downvoted question they've already formed the opinion it must be terrible before even reading it

That leads to my second suggestion: make the voting button not so prominent or hide votes until you vote (maybe not the latter). In it's current layout, I see the votes before everything else. That means if a couple people dislike the question or it's subject matter, I see their opinion before I can form my own

It's just my two cents and I'll be happy to discuss in the comments

  • 1
  • 1
    You can delete a downvoted answer unless it's been accepted, so I don't see a reason to limit the effects for answers. For questions, though, where one FGITW answer (or snarky comment disguised as an answer) can prevent you from deleting the question you now realize was ill-advised, it seems reasonable to stop the bleeding at some point. If there's nothing valuable in answers then mods might even be willing to delete, but fairly often a bad (or on meta, unpopular) question had a good answer that we don't want to lose, so the drain continues way past when you were ready to bail. Jun 27, 2019 at 17:11
  • 1
    @MonicaCellio generally answers don't see the sort of downvoting questions do. But I agree with what you're talking about
    – unixandria
    Jun 27, 2019 at 19:04

I don't understand why there is reputation lost there on MSE.

On per-site meta it's less problematic as people don't loose reputation and feel less attacked.

That really change the dynamic. I even see post on per-site meta from users that answer a yes, and a no (differents answers), for like a tag burninathing request to have fun.

To display a reputation on MSE they could simpy sum up the network wide reputation.

  • 1
    MSE has its own separate rep rather than a network-wide sum because the latter model would unfairly benefit Stack Overflow users over those of other sites and it would un-value users like me who contribute mostly on Meta rather than anywhere else. May 15, 2019 at 3:20
  • @SonictheInclusiveHedgehog Someday even MSE would no longer be there, Tim stated it in the past, MSE is not the correct tool for keeping on tracks feature request or to be a bugtracker. The question is, where MSE would be standing in the futur ? for support, site-recommandation ? but it got to change, as I can understand people that feel harrassed when they get murdered on MSE.. I erased my MSE account, took a 2years break to cooldown before coming back on MSE. dont be afraid to tell the word harasment, a -30 question with positive answer, you then loose control and the bad feedback just enter
    – yagmoth555
    May 15, 2019 at 4:21
  • 5
    Somebody who has a lot of rep from other sites doesn't necessarily understand meta issues and shouldn't come in with rep and privileges that indicate otherwise. By the same token, somebody with high rep on SO even has to start over on Software Engineering, let alone Literature. "The SE network" is a topic that one can have expertise in. May 15, 2019 at 14:26
  • @MonicaCellio I stated to have the reputation displayed, I didn't stated the priviledge associated with it. I try to brainstorm some ideas, as currently MSE is kinda dysfunctional to say at the least.
    – yagmoth555
    May 15, 2019 at 14:39
  • @yagmoth555 oh, thanks for clarifying that. So maybe you mean something like the multi-site flair that shows your aggregate rep and your top several sites? That could be handy to have visible here on meta. May 15, 2019 at 14:45
  • 1
    @MonicaCellio yes, and maybe just have the rep calculated for the privilege when users answer (aka help on meta), but maybe when people do ask question(s), I would act like on other meta for question, aka no reputation gain or loss there for the OP.
    – yagmoth555
    May 15, 2019 at 14:50
  • @MonicaCellio that sounds like a known feature, combined-flair. I show mine in my profile: stackexchange.com/users/flair/446514.png
    – gnat
    May 16, 2019 at 9:18
  • @gnat I think yagmoth555 is suggesting showing that on meta instead of the usual site-specific usercard. May 16, 2019 at 13:53
  • 1
    "I even see post on per-site meta from users that answer a yes, and a no ... to have fun." That's not about having fun: it's an attempt to work around the problem that people are confused as to whether up/downvotes on the question mean "Agree/Disagree" or "Well written question/Poorly written question". However, polling on meta has its own problems which I'm not going to get into in a comment thread. Jul 31, 2019 at 11:09
  • @PeterTaylor I agree with your point, but you miss my argument. We can do work without having to downvote to hell someone, by dooing so it's almost harassment at some point, it's counter-productive
    – yagmoth555
    Jul 31, 2019 at 13:46

Despite having mostly upvotes and some reputation on the site, when I visit some sites I still get a banner encouraging me to upvote.

Vote it Up!

I don't think the solution is that simple. Some people are resistant to change and don't accept what's provided or offer better self answers. It's common enough to see a comment explaining that they don't see the answer and multiple replies explaining the answer.

Running the SEDE query: "How many anonymous upvotes and downvotes are there?" returns this result:

Name Name Anonymous feedback
DownMod Answer 14248
UpMod Answer 57390
DownMod Question 12848
UpMod Question 49257

(4 rows returned)

Name Name Votes
DownMod Answer 67433
UpMod Answer 1046543
DownMod Question 180550
UpMod Question 833460

(4 rows returned)

Perhaps the solution is that once a question or answer on this site hits -5 it should be greyed, permitting further votes from people with more than (for example) 1000 reputation, and that lower reputation users wouldn't see the post. That avoids users whom might be unfamiliar with (the proposed) voting practice preferences (don't pile on) being included after a general consensus is reached about the post.

We could add greying for higher level users whom have a significantly greater than average number of comments and downvotes and a low number of highly upvoted answers.

@Monica, can you find out if lower reputation users are the downvoters on the Q&As to which you refer (but offer no URL), or are higher reputation users also pilling on? Since I don't have access to that it can't be part of my answer. Also of interest is if it's the same small group of people, no names are necessary and the solution can be automated. Statistics to support your theory would improve your question.


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .