Large bold text could be taken the wrong way by some of the audience that we care the most about, namely mathematicians. I could imagine it being seen as condescending or unnecessarily aggressive. In the end, these are the users we actually want to attract and keep. Inevitably that means we aren't going to adopt policies that are optimally suited to turning away other users, and as long as we want to maintain the high quality of posts here, we'll need to keep turning away other users.

I think generally the comments people leave when closing inappropriate questions are not too snobby or elitist. If someone gets offended by "this is not appropriate for this site, please read the FAQ" then, well... would they get offended if they walked into a medical convention and started asking questions about high school biology, then were shown the door?

I'm simply suggesting that for the first time someone tries to post a question a more prominent question of the likes of "Are you sure this is appropriate" might not be a bad idea.

If this somehow dissuades a professional mathematician from posting, then I would be a little surprised.

I agree that first-time askers would do well to show that they had put some thought into their question, but honestly, how do they know that unless they have already read the FAQ? Since it's clear that on pretty much any community website, the FAQ is the last document that anyone reads, this is sort of an unreasonable request unless this particular social more is made more visible.

This sort of view seems reminiscent of the stereotype of the BOFH. I know that these questions are very quickly re-directed to other sites, but why should we even need that intermediary step?

I think Todd and Qiaochu's responses are pretty much on the mark. The amount of questions that are closed isn't surprising considering how relatively prominent MO is. So I think we're in a situation where "if it isn't broken, don't fix it" applies.

Todd Trimble, regarding 'most cursory glance at the neighboring questions': yes and no. Rarely I visit other SE sites and look around a bit. Typically I rather do so by looking at the questions with most votes. Looking at this list of questions, one could get a different impression of MO: common false beliefs, department closure, refereeing, polynomial rep. all pos. integers, thinking and explaining, colorful language ... not that clear what impression one gets.

And, it is my impression, that there are some users (that seem like reasonable people from what they write and how they react, just not mathematicians) that are honestly surprised when they are told there question is too simple. An easily visible hint regarding the essentially purely professional nature of the site might help. It is true that its says it is a site 'for mathematicians' but perhaps this is not explicit enough. True, writing something like: "if you do not consider yourself as a professional mathematician you are most likely wrong here" might also not be the way to go, but perhaps there is something in between.

Here's a new example. Anybody write any languages used in Serbia? I don't think any FAQ would have helped.

http://mathoverflow.net/questions/67624/volume-of-the-circular-segments-figure

simoncfr, I made various suggestions of the type you have in mind, sometimes for the comfort of new users, sometimes for the convenience of moderators and others with the power to vote to close questions. The response was fairly consistent, which is that the most important thing is that some shy giant of mathematics be comfortable posting an anonymous question or answer, without discouragement or impediment.

I am a bit torn. On the one hand, I do agree with Ryan Budney and some others that it would not change that much for a problem that does not seem to be a major one. And, I think that what is actually more important is to try to further reduced, in line with Todd Trimble, occassional but still regular s(n)arky comments towards newly arrived or somewhat occassional user (that thus have a low reputation). For general reasons, but also since they, still more rarely but I can remember more than one instance, 'hit' people plainly in the key target group of MO, so working research mathematicians, even when they post under their real name. Actually, this seems to me by far the most critical scenario for loosing valuable participants: You are a research mathematician, come to this site (under your real name) somebody told you about, ask a reasonable question, and then...somebody perhaps somehow misunderstanding the question or missing the key point starts talking down to you as if you were a not overly bright freshman (and even, or also particularly if you like, towards them one should be more respectful; but in that case at least the mathematical level would be about right). [I am not complaining about personal experience; but I have at least two/three precise instances in mind, though I'd prefer not to link to them.]

On the other hand, I think that simoncfr has a point. Todd Trimble says 'analogous professional site dedicated to something like biochemistry' (emphasize mine). What I am a bit dubious about is precisely if everybody understands this professional nature of MO (without reading the FAQs). On the front page it does say 'a site for mathematicians'. But, if I'd see a Q&A site which says 'a site for photographers/gardeners/golfers/fishermen/whatever' (these sites are made up, though some might even exist, no clue) then I actually doubt my first thought would be this is a site almost exclusively for professional photographers/gardeners/golfers/fishermen/whatever, but rather for somehow motivated amateurs in the respective activity where some pros occassional drop by to help with tricky questions. And, it is my impression though I am as said rarely on other Q&A sites, that it is quite exceptional that a site of such a level exists for a subject; but Andrew Stacey has some essay on his website, written for the tex-se site, which makes among others such a point. So, it is not so surprsing that some people are mislead. And, now it said 'biochemistry' and not one of the activities I gave that are more likely to be a hobby. But, here I think we also face the 'problem' (though most of the time it is not such a problem)that we (or at least I, but then there was a thread on 'responses to bad at math' so I guess it is not such a unique experience) face frequently in real life; namely, that by no means all people even know that something like professional mathmatical research exists at all, and among those that do know still most have a completely flawed idea that is, only slightly exagerrating, somehow along the lines 'convoluted calculus problems, hard logic/number puzzles, programming a computer to compute something really long and complicated'. And, more or less these are the question that we are 'complaining' about here. So, what rests for me is somehow the argument that people ought to understand form the other questions that they are wrong here. But, also this is not so clear (even leaving aside the point on this I made before and focusing on the more random sample of newest or active questions). Perhaps it is helpful to think about how easy or not it is for a 1st year undergrad to distinguish what is an assignment for a 3rd year undergrad form what is an ansignment for a beginning grad student from what is research level mathematics.

And of course I do not want to claim it is in anyway hard to find out about the professional nature (well, it suffices to read the FAQs of course), but still I share simoncfr's view that it is at least not obvious, and perhaps could be stressed a bit more. But, how exactly I am not sure, as to overdo it certainly could back-fire. Maybe just a line above the ask box: Please note this site is mainly intended for professional mathematicians, for details please see the FAQs.

simoncfr: As someone that has to sift through a lot of verbose nonsense from funding organizations and universities and such, I'm inclined to think the *less* volume of information we put on the "ask a question" page, the more likely it is that people will read the information that's present.

So here is a counter-proposal. Instead of putting *more* information on what is appropriate or inappropriate, how about we reduce the amount of page space devoted to typesetting and subject tags? That would make the links to the FAQ and "how to ask a question" page more prominent.

Perhaps we could organize the links in one spot, in order of relevence. FAQ first. How to ask a question next. Then information relevant to actually asking questions.

I'm only a recent user of MO, and here are the steps I followed before asking my first question:

0. I looked at the newest questions. Several were closed. Probably not a good template.

1. To get a sense of what were good questions for the site, I looked at the 'voted' list. There was a lot of non-research, 'sociology of mathematics' questions which got a high number of votes.

2. I looked at the 'hot' and 'active' lists. The distinction was not immediately apparent, and again there were a lot of 'sociology of mathematics' questions.

3. I looked at the questions tagged 'numerical analysis', and found these to be a mixed bag.
These observations constitute 1 data point, but may help explain why the casual new user needs a clear description of what is a good question for the site, since the evidence (highly voted questions, questions which draw a lot of responses) are NOT typical of what MO is supposed to be for. I actually signed up, and then remained inactive, because I concluded MO was basically a site where people discussed soft questions.

I've always found those "hot" and "active" filtering criteria to be useless.

Perhaps it's time to close some of the more populist non-research oriented threads, like:

http://mathoverflow.net/questions/63221/ideas-on-how-to-prevent-a-department-from-being-shut-down

and

http://mathoverflow.net/questions/32566/careers-advice-for-ph-d-s-without-current-postdocs-or-university-jobs

@Nilima: several of your suggestions would likely require changes to the software. At present those aren't feasible. But creating a non-automated list of "great questions" is something that's perfectly managable, linked to from the FAQ or the "ask a question" page.

I am in the midst of discussions with Scott Morrison (and Alexander Woo, and who ever else cares) about reimplementing the API, and it looks like other features like this may be possible to incorporate in making a new front end as well. I am also incrementally adding to a design document for a platform which suggests an alternative solution, involving sponsored questions, or a two tier system. I intend to have a first draft available for critical review soon on the new platform, but I consider this facet interesting enough to present here.

Consider having two categories of questions: research and other. There are several ways to work a forum using these; I suggest the following: users above a certain activity/reputation threshold (e.g. 10 answers or 100 reputation) can post to either tier. A certain amount of work is required (either moderator work, or something similar to votes to close/advance) to move questions between the tiers. Users below the threshold can post questions to the other tier and ask to have them sponsored (moved) to the research tier. Anyone can choose one or more of three basic views (research, other, both) to see the questions. There are slight variations on this idea which I hope the community will discover, discuss, and determine how good/bad an idea this is. I also hope there is discussion on why something like this should be done at all, so that there is thinking going on about MathOverflow as a whole, and not just an grand object to be decorated, tweaked, or adjusted at the whim of the community.

If the basic idea is intriguing, I can elaborate on the problems this idea could potentially solve and some problems it would create. However, I would like to see some consensus on interest level as well as what problems there are that need solving.

Gerhard "Ask Me About System Design" Paseman, 2011.08.10