Preamble: This is not yet a feature request. The point is to refine through discussion a potential future feature request.

First, what is a ban warning:
The ban warning makes some constructive suggestions, but it also threatens.

I recently got one on Meta, and I felt pretty bad, because I felt I was trying to contribute and I saw I could be automatically enter a ban that none could reverse manually. Then by asking this question I learnt that the system is now much more tolerant (on Meta). But still, I got the scary warning pretty early (after one question marked as duplicate and one not successful feature request). So my questions are the following.

Why do we threaten so quickly with an automatic ban (on Meta)?

I understand the banning formula cannot be shared, but I think I would like to understand, (in percentage?), how close I am to getting banned when I see the warning? And if I am far, why do you warn me so strongly and scarily?

Let's say you really cannot share at which point I am on my walk towards the ban:
Can't there be two levels of warning (on Meta)?
The first one only with the constructive suggestions not even mentioning the possible ban (not even a warning in fact), the second the real warning with the threatening part (maybe well highlighted to show the difference).

  • 9
    If a lot of your posts end up being pointed as duplicates, it's a warning to start using the site search
    – random
    Mar 29, 2015 at 22:16
  • @random Your comment is off topic. I am not asking what is such a warning.
    – Antonio
    Mar 29, 2015 at 22:18
  • 7
    I've re-read the warning several times since you asked this question and I can't see which part is threatening. It's honest about the dire straights the user has wandered into. If they don't consider their questions (both new and existing) more carefully, they will be blocked from asking more. It's designed to be like the signs warning of danger in dangerous situations. Would you consider this sign threatening? Mar 29, 2015 at 23:21
  • 1
    @JonEricson It's not the same thing. It's not my friend telling me not to got in that part of the city by night because it's dangerous, it is someone telling me don't do <something> or I will slap your face (the danger comes from the person warning me). The friend would tell me "stop asking bad questions otherwise people will start having a low consideration about you".
    – Antonio
    Mar 29, 2015 at 23:49
  • 10
    "I welcome downvoters, especially without comments: you make me stronger." Not a great way to attract votes. You'll probably get more than one downvote because of this.
    – ale
    Mar 30, 2015 at 0:11
  • 1
    I find it unfortunate that this question is getting downvoted so heavily. It seems like a fair question that deserves an answer. In other words, it seems like a reasonable question, and an understandable reaction to the warning message. A poorly-received feature request is something that could happen to almost anyone, and can't necessarily be predicted.
    – D.W.
    Mar 30, 2015 at 4:12
  • @Antonio: I see what you mean. But in that scenario, isn't the real problem that the other person is about to do something nasty to you and not that they warn you about it? Mar 30, 2015 at 10:47
  • @AlE. I tend to not care about "punishment" downvotes (in this specific case), I am confident about the quality of this question and I am very satisfied that it triggered discussion even within Stack Exchange staff. (Note: I wrote that comment after downvoting had already started)
    – Antonio
    Mar 30, 2015 at 11:02
  • @JonEricson What is the difference? The automatic system is telling me "Don't ask bad questions anymore or I will forbid you to make anymore questions". If there was a way to try putting the user on the right path before using the threat, my opinion is that it would be better. <funnyMode>BTW, regarding the picture you linked, I was thinking that we do not have that kind of verbose warning signs here in Europe (or at least, we have far less than my US trips allowed me to see :) )</funnyMode>
    – Antonio
    Mar 30, 2015 at 11:26
  • That's exactly what the system is trying to do. Any way you look at it, asking questions is a privilege. It is NOT a right. And an automated system is exactly like a sign. It's just enforcing the fact that it's a privilege. This sign is just trying to let you know before the system fulfills it's function.
    – fbueckert
    Mar 30, 2015 at 17:48
  • @fbueckert Asking questions is a right, which one might lose because of abuse. Actually asking questions and giving answers is only thing you can do at the beginning when you join the website. Seen with an example, being out of prison is not a privilege.
    – Antonio
    Mar 31, 2015 at 13:13
  • 1
    That's a staggeringly wrong analogy. Asking questions is not a right. If it were, there would be no need for the question ban in the first place. That enforces the fact that if you don't ask questions accepted by the community, you lose that privilege. The Internet is much, much larger than this network; nothing is constraining you from going elsewhere. Asking a question, while one of the only things you can do at first, does not entitle you to do it forever, without limit.
    – fbueckert
    Mar 31, 2015 at 13:27
  • @fbueckert You fail to explain why the analogy is so wrong. In the analogy, a question ban is a measure taken by the community/system to limit your freedom because you abused of it.
    – Antonio
    Mar 31, 2015 at 13:33
  • You fail to explain how it's even applicable; prison constrains your complete freedom. SE does nothing but prevent you from asking questions, if you have a proven record of your questions being low quality, or unpopular (in the case of Meta). It does not prevent you from accessing the rest of the internet, or, heck, walking outside and asking a random stranger your question. If you're going to use an analogy, try to use one that's at least somewhat accurate. Otherwise, it weakens your entire argument.
    – fbueckert
    Mar 31, 2015 at 13:39
  • If you're referring to the SE network as the whole of society...you need to gain some perspective, and it completely flaws your analogy. SE isn't the whole of society.
    – fbueckert
    Mar 31, 2015 at 13:50

2 Answers 2


Update: After talking with Anna, I've relaxed the warning trigger several notches on Meta Stack Exchange. People will still be warned, but it will be much closer to the point where they are in serious danger on this site.

First, voting on meta can make post bans a little strange. Asking a controversial question might trigger the warning even if you do a good job of posing the question. For that reason, the threshold is set a little looser on Meta Stack Exchange than on other sites.

Second, the idea of two levels of warning sounds good in theory, but it would be a bit difficult to calibrate. With the new rate limits, people get blocked a lot more quickly than they used to. Since we'd like to warn people first, there's not a lot of room to fit in a softer warning. The sooner a user knows they are in hot water, the better.

Third, we are certainly open to suggestions about how to improve the wording of the message. The purpose isn't to make you feel bad, but to help you do better in the future. Please feel free to answer your question with any improvements you'd like to suggest.

Finally, the purpose of the warning isn't to indicate how much closer you can fly to the sun, but to encourage you to be more cautious with your questions. If you got a warning, then you should assume you are flirting with being blocked from asking more questions. Ideally, we'd like you to consider how to edit existing questions and pose future questions as carefully as you are able.

  • Wait a sec: are the limits lower or higher? Here it says it's higher meta.stackexchange.com/a/197335/225842... If it is higher only in Meta, can't the two level warning be applied only for Meta? Because of feature requests that might be not liked by majority, one can get warnings also while being contributing. Finally: As long the warning contains a threaten, there's no wording that can help improving the situation.
    – Antonio
    Mar 29, 2015 at 22:48
  • 1
    Higher if you look at it from one side, and lower if you look at it from the other side... In other words: Higher tolerance = Lower settings Mar 29, 2015 at 22:51
  • In this case looser means the same as higher.
    – PolyGeo
    Mar 29, 2015 at 22:51
  • @MeNoTalk I said higher, I will say "more tolerant" to avoid ambiguity. While Jon says with the new rate limits, people get blocked a lot more quickly than they used to. So is it more tolerant or more strict? Is it strict, but made more tolerant only for Meta? Does this leave space for 2 thresholds at least in Meta?
    – Antonio
    Mar 29, 2015 at 23:06
  • 2
    @Antonio: The rolling rate limits kick in more quickly, but are less permanent than the legacy question ban. I'm not sure if that is more or less "tolerant". The settings on meta generally are designed to take the unusually voting patterns on meta into account. I guess you can think of that as more tolerant, but that's not how I'd prefer to phrase it. Mar 29, 2015 at 23:13
  • @JonEricson If I posted a feature request to put 2 levels of warnings but only in Meta, assuming there is enough margin to do that (only the people managing the website knows), would you vote against? If yes, what would be your motivations?
    – Antonio
    Mar 29, 2015 at 23:17
  • 6
    @JonEricson The warning thresholds are configurable. Since the question ban thresholds are relaxed on Meta SE compared to other sites, we should consider adjusting the warning thresholds to be more reasonable in context.
    – Adam Lear StaffMod
    Mar 29, 2015 at 23:27
  • 1
    Jon, you call for feedback on the current warning. Antonio has already provided some feedback in the question and in a comment underneath the question. As he says, the current language is threatening. I think there's a valid insight there, and I think it would make sense to acknowledge that validity and consider whether there is anything we can do to address it or not -- and whether we should address it or not.
    – D.W.
    Mar 30, 2015 at 4:02
  • Also: On Meta.SE, sometimes, questions can be poorly received even if you put a lot of effort and thought into them, and this can't always be predicted or anticipated, no matter how hard you try -- sometimes it's just out of your control. When you combine this with a warning that warns such a situation could lead to you being banned and prevented from having any voice in the future of the site, I can sympathize if it feels off-putting and makes users feel threatened and powerless. (Nobody likes to feel powerless.) We might decide this is unavoidable, but I'm sympathetic.
    – D.W.
    Mar 30, 2015 at 4:15
  • @AnnaLear: I might be misreading the settings, but I believe the warning is already triggered later on MSE than on other sites. Mar 30, 2015 at 4:59
  • 1
    @JonEricson There are multiple knobs to tweak here and not all are modified on MSE. I'm leaving you more details in chat. :)
    – Adam Lear StaffMod
    Mar 30, 2015 at 5:10
  • @AnnaLear I would be curious to know if there was any follow up regarding this topic.
    – Antonio
    Apr 7, 2015 at 16:28
  • @JonEricson That's great!! Would it make sense to mark this question status-completed/status-declined, (I have seen this is possible). I understand that this question was not posted exactly as a feature request, but since it generated some discussion and changes I think it could be documented.
    – Antonio
    Apr 14, 2015 at 11:30

There are lots of ways to participate on Meta. You can ask how things work, you can answer questions about how things work, you can react to other people's suggestions of changing how things work, or you can suggest that things should work differently.

Of these four ways of interacting, the last is the most dangerous. If you propose something many people disagree with (and especially if it's been proposed many many times before, and your question implies that it is a new idea) you will get a TON of downvotes, and these could lead to a post ban. The threshold is different on Meta, but it still exists.

And it should, in my opinion. Being told "you're doing it wrong" by someone who isn't actually that clear on how anyone is doing it is a tiresome experience. A downvoted question on nonMeta sites means "can't be bothered to search, write well, or include what is needed". Sometimes it means that on Meta too. The number of well written carefully researched feature requests that were heavily downvoted only because others didn't want that feature, and were written by people with very few concurrent well-received posts? I think it's zero.

I would suggest that anyone receiving that warning on Meta back off the feature requests for a while and look at other ways of participating. You'll come up with better feature requests if you do, and you're less likely to get banned.

  • Look at this, was it a badly searched feature request? Later on, I even digged out a reference in another old question, that was not at all a feature request. And it's not that I am offended by the downvotes, because it's part of the game. But if now you tell me I got downvotes over there because of the poor way I acted, beh, ok, in that case I am a little bit offended :)
    – Antonio
    Mar 30, 2015 at 13:09
  • 1
    I wasn't speaking about you specifically, and I didn't go and check out your questions. I was simply addressing the case of someone who has highly downvoted questions - chances are they are either rants disguised as questions, or premature or duplicate feature requests. And in both cases, a warning to behave differently is appropriate. In my opinion. Mar 30, 2015 at 13:48
  • You said The number of well written carefully researched feature requests that were heavily downvoted only because others didn't want that feature, and were written by people with very few concurrent well-received posts? I think it's zero. I am trying to say, check that question, the number is not zero. Now, unless you want to tell me that my question has been judged badly because I didn't have enough reputation and feature proposal history.
    – Antonio
    Mar 30, 2015 at 14:08
  • I didn't downvote your other question, but apparently you want me to tell you whether I think it is well written and carefully researched. OK, I don't. It seems like you want an automatic banner telling people not to downvote certain questions. In my opinion there is no kind of question (not self answered, not from new user, none) that deserves such a banner. Much of meta feels the same way which is why you got comments saying self answered questions don't get a free pass. If you don't know what they're referring to then you don't have the background to make that proposal. Mar 30, 2015 at 15:10
  • So I should stay enough here in Meta to learn the common trend and understand if a feature request can be accepted before posting it. Got it.
    – Antonio
    Mar 30, 2015 at 15:19
  • 1
    Consider what background there is on the topic, whether it is part of a larger category of recurring requests, and how people feel in general about that larger category. It might not be accepted - I have requested several things that have not been accepted - but without research (in this case into the sometimes irrational way Meta voters can react to certain kinds of feature requests) you are taking a chance. That is my only point. I'm not saying "sit down and shut up" and don't appreciate being paraphrased as so doing. Mar 30, 2015 at 15:21

You must log in to answer this question.

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