There is a new posting on the Stackoverflow blog, entitled Stack Overflow Isn’t Very Welcoming. It’s Time for That to Change. A quote:

Too often, someone comes here to ask a question, only to be told that they did it wrong. They get snarky or condescending comments for not explaining what they’ve tried (that didn’t work). They get an answer… but the answerer gets scolded for “encouraging ‘low-quality’ questions.” They get downvoted, but don’t know why, or called lazy for not speaking English fluently. Or sometimes, everything actually goes well, and they get an answer! So they thank the poster… only to be told that on Stack Overflow, “please” and “thank you” are considered noise.

The blog ends with "This post focuses on Stack Overflow, but most of it applies to the broader Stack Exchange network as well." Are there lessons for us here at MO?

One simple modification of our "best practices" could be this: votes to close questions that are not research level will not be accompanied by a down vote. A down vote feels like slap in the face. A vote to close should be sufficient to drive the message home that this is the wrong site.

  • 7
    $\begingroup$ Isn't the main question here how many mathematicians asked a question on MathOverflow some time, and did not become regular users due to the feedback they got? -- And if so, how would you get useful information on this by asking on this site? $\endgroup$
    – Stefan Kohl Mod
    Commented Apr 27, 2018 at 9:48
  • 10
    $\begingroup$ @StefanKohl Word of mouth I suppose. For example, during a department tea someone may mention using MathOverflow, and others chime in with why they don't use MO because of treatment they got. Of course this may be anecdotal and hard to quantify. But there could be lessons nonetheless. $\endgroup$
    – Todd Trimble Mod
    Commented Apr 27, 2018 at 10:38
  • 3
    $\begingroup$ This blog post by Jon Skeet, particularly the section on "Differences in goals and expectations" seems relevant to me. $\endgroup$
    – j.c.
    Commented Apr 27, 2018 at 15:06
  • 18
    $\begingroup$ Well, seeing this post it is difficult not to mention relatively recent fedja's question: Should we exercise a bit more tolerance towards newcomers? $\endgroup$ Commented Apr 28, 2018 at 7:50
  • 3
    $\begingroup$ Partly before Carlo Beenakker's question seems to be asking also what experience users have on MO and partly because some of the answers in the linked question seem to address also some related issues, I will also add link to this Why all the negativity?: $\endgroup$ Commented Apr 28, 2018 at 9:24
  • 2
    $\begingroup$ @MartinSleziak, we might also consider: Should we exercise a bit less tolerance towards old-timers? Newcomers might feel more welcomed by more consistency. $\endgroup$
    – user44143
    Commented Apr 29, 2018 at 0:20
  • 8
    $\begingroup$ There is a technical problem with the "no downvotes" suggestion that you have added to your post. When a question is closed as off-topic, it should eventually be removed from the site. (Probably after the OP was told where to ask and why the post is off-topic, ideally in a manner which is compatible with the "Be nice" policy.) Without any downvotes, the question is less likely to be deleted by roomba or at least it takes much longer. And judging by the past discussion, MO users are not too keen on manual deletions. $\endgroup$ Commented Apr 29, 2018 at 12:01
  • 4
    $\begingroup$ @MartinSleziak: On the other hand, a question might be closed in order to get edited and improved by its author, only to be reopened later. Alas, in this case the downvotes will stay, even if the edited version is better - how many of you have returned to a post that you have downvoted to see if it had been edited, and possibly retract your vote? I say close, but do not downvote unless the question is really bad. $\endgroup$
    – Alex M.
    Commented Apr 29, 2018 at 17:10
  • 2
    $\begingroup$ @AlexM. I should probably have that made clearer, but my response to the last part of the (current revision of the) question was made under the impression that the OP refers to the posts that are clearly off-topic. For example, stuff such as introductory calculus or linear-algebra exercise has virtually no chance of getting reopened, even if it is edited. (And it also depends of number of such posts. Having 10% off-topic questions on the front page would probably still be tolerable. If they were 90% of new posts, that would certainly be a problem.) $\endgroup$ Commented May 1, 2018 at 7:04
  • $\begingroup$ While looking for something else I stumbled upon a thread called MO unwelcoming? posted on tea (the old meta) back in 2010. $\endgroup$ Commented Feb 16, 2021 at 13:11
  • $\begingroup$ @AlexM. After a question is edited, you can reverse your downvote. $\endgroup$ Commented Jan 31 at 1:20
  • $\begingroup$ @DavidWhite: Yes, but how do I find out whether it has been edited? Especially if the edit comes a long time after the downvote, when the question has become inactive and has slipped out of my mind. $\endgroup$
    – Alex M.
    Commented Jan 31 at 8:24
  • $\begingroup$ I guess the edit bumps the thread to the main page, and my browser colors the link differently so I know this is one I already clicked. Then, I remember that it's one I downvoted and seeing that it's been edited, I click it again to see if it's better. The ones I miss are the ones that get retagged and end up hidden from me on the front page. $\endgroup$ Commented Jan 31 at 11:59

4 Answers 4


It depends. I think we don't want to be welcoming with a new user posting "plz solve teh linear algebra homework <see photo>". These questions deserve to be closed and downvoted into oblivion; they should not have been posted here in the first place, the user did not bother to read any instructions, and we don't have any interest in keeping them here. But we should be welcoming with people writing meaningful research-level questions, even if they are not well formulated (because of language difficulties, or inexperience with Markdown or with this site). Often, this means silently editing their questions or asking them for clarifications, and I think that in most cases we already do it reasonably well.

But I think that the true problem is a different one.

In many cases, your "being welcoming" means covering up for a hole in our instructions to new users. I tried opening MO in an anonymous browser window to ask a question, and I think that the instructions that were presented to me as a to new poster are a bit underwhelming:

What's your research level mathematics question? Be specific.

How to Ask

Is your question about research level mathematics?

We prefer questions that can be answered, not just discussed.

Provide details. Share your research.

If your question is about this website, ask it on meta instead.

visit the help center »

asking help »

We never state "this site is not about undergraduate homework", "there's another site where you can ask general question about mathematics", or "explain what you tried (that didn't work)". There could be a link pointing to more explanation in the first lines. I think that the first thing to do is fixing and improving these instructions.

Clear and to-the-point instructions should be our main way to "be welcoming".

  • 5
    $\begingroup$ This suggestion seems related to some of the points you raised in your post: Should users be shown some basic information before posting the first question? $\endgroup$ Commented May 2, 2018 at 14:21
  • 1
    $\begingroup$ If the moderators do not make this change to the instructions then I am very confused with their intent. We should prefer clarity over shortness when informing newcomers. $\endgroup$ Commented May 3, 2018 at 20:45
  • 2
    $\begingroup$ I agree that improving the instructions should be done. I disagree about this tactic substantially improving the reputation of MathOverflow, or improving the feel or experience of most new users because most new users (especially those of the type Federico wants to discourage) do not read the instructions. There is a hope they read the answers and comments, and if those are phrased well, the answers and comments may be more effective than the instructions. Gerhard "Thank You For Reading This" Paseman, 2018.05.03. $\endgroup$ Commented May 4, 2018 at 4:00
  • 1
    $\begingroup$ The feature request linked in my first comment is now moot, since Stack Exchange changed what is shown to the first-time askers. Now a similar feature request can be found here: Customize the modal window for the first-time askers. $\endgroup$ Commented Jan 11, 2021 at 9:45

I think there are contrasting goals within the group of long term members (the established MathOverflow community, say those who have been posting five years or more to pick an arbitrary division) and thus the lessons (if any) to be learned are different for different members.

I believe that the forum needs to grow (possibly with some cultural change, although I don't want much change) and that what often happens in handling of new questions is very uneven, which contributes to the bad part of the reputation MathOverflow garners. Although I sympathize with the professionals who want questions to be well written, interesting, and new, (and prefer to discourage things that are otherwise), I think it should be recognized that many newcomers do not understand and will not understand what is appropriate to the forum without making mistakes. I think that having a large volunteer team address such questions quickly and politely would incur the greatest benefit to the reputation of the forum. This is why I spend roughly 10% of my responses in telling some that this is the wrong forum for their question, and for most of those a hint or two as to where to go and what to think regarding their question. I sometimes lose my patience, and (hopefully rarely) this shows in some of my responses. (There are also issues not involving newcomers and first questions, but I think solving this issue well will pay many dividends.)

I think the biggest lesson to learn is how united and formal we want to operate as a community. I can see arguments on the side of being less welcoming as well as on the side of being more welcoming. Over the years I have suggested polls, codifying standards of behaviour, getting a consensus on various issues, and doing activities that would require more energy and thought on behalf of the long term (and other) community members. As a group, I think we are not wanting such formality, and instead prefer to muddle along somewhat independently. So I am unsure that there will be any lesson for the group.

Gerhard "What Lessons Do You Want?" Paseman, 2018.04.27.


Interesting topic. Good thread. I have been here only a month, and so far I find these forums to be a terrific resource.

I'm not seeing how a forum such as MO (or to a lesser extent MSE) even could be able to stay the way it is without some harsh screening.

THAT SAID, no one can know everything about math, although sometimes we all feel that we need to. This is what I mean: I am a research mathematician who has used this site to ask math questions that happened to come up in my own research, and that happen to be outside my area of expertise. So chances are the questions are fairly basic to an "insider". For example, I am in graph theory who has needed to be able to "get" a theoretical construction that uses a lot of high-powered number theory so I've asked questions about number theory that probably aren't near research-level to the insider. Maybe several others on here have done likewise. Anyway, I have gotten good answers, BUT (a) I do my best to explain my predicament, and (b) I make a point about being as clear and specific with the question as possible.

  • 2
    $\begingroup$ Chuck Norris knows everything about mathematics. He can count backwards from infinity. $\endgroup$
    – Asaf Karagila Mod
    Commented May 11, 2018 at 6:56
  • 1
    $\begingroup$ Only if he does it in tens though. Otherwise it'd take him too long $\endgroup$
    – Mike
    Commented May 11, 2018 at 15:31
  • 3
    $\begingroup$ No, he counts all the way through the real numbers. Each number and its predecessor. Yeah, Chuck Norris can tell you what is the predecessor of $\pi$ on the real line. $\endgroup$
    – Asaf Karagila Mod
    Commented May 11, 2018 at 15:33
  • 3
    $\begingroup$ Well then. His talents truly are uncountable indeed! $\endgroup$
    – Mike
    Commented May 11, 2018 at 16:45

I think MO has a smaller problem compared to SO. After reading the blog post regarding the highlighted difference in goal of the asker and answerer at SO, I believe that Mathematics is of a different culture.

We have (I hope) a culture where asking questions is just as important as answering them, and a question without an answer can be just as valuable as an answer if it is interesting enough.

However, I think the last point is also a reason why so many professional mathematicians avoid MO. One does not wish to "give away" a valuable question, and risk having the ongoing project being solved in an afternoon by a quite brilliant community.

  • 5
    $\begingroup$ is there any evidence for the stated "reason why so many professional mathematicians avoid MO"? I am not a mathematician, but as a physicist I would be thrilled to have a forum of brilliant scientists that would solve my research problems, and would be delighted to have these as co-authors. $\endgroup$ Commented May 1, 2018 at 15:44
  • 1
    $\begingroup$ @CarloBeenakker: Well, my former PhD supervisor was a bit reluctant, or, well, expressed that one needs to keep this in mind. And he is not active on MO (to my knowledge). I guess it depends a bit if the goal of the research is solving the problems, or building a career - The most efficient way of doing research might ruin your career (if one does not have a tenured position yet) in some sense. $\endgroup$ Commented May 1, 2018 at 17:32
  • 21
    $\begingroup$ A problem that can be solved in an afternoon and whose answer fits into a MO answer is not nearly substantial enough to form an important part of one's research program. I don't think people generally post the main problems that they are working on, but rather questions concerning technical issues or side problems that arise while working on bigger things. $\endgroup$ Commented May 2, 2018 at 16:00
  • $\begingroup$ Well, there is also too the fact that if you truly have a specific question related to your own area of research, as brilliant as this community is, the probability that one of the 4 or 5 people who are experts on that particular area will be hanging out here (e.g, any one particular place), within a short time interval, are pretty rare. If you are that well-known yourself chances are you know who can answer your question and you'd just email them directly, expecting a timely response $\endgroup$
    – Mike
    Commented May 10, 2018 at 20:22

You must log in to answer this question.

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