MathOverflow is a nice high-level communitiy with a good collegial academic atmosphere.
I'd like to know about how this can be achieved; for example how does MathOverflow keep up the good qualitiy of the content of the site and what is important to successfully retain the good contributors and researchers?
What measures does MathOverflow take, or what do people here think, is most imporant to maintain a high-level site with such a good academic atmosphere?
The main things that are quite irritating now and can easily drive people away (IMHO) is the not always applicable assumption that everything is either a research level question or not, a very questionable assumption that the guys who vote to close always know the difference themselves, and a completely wrong assumption that if something is commonplace/uninteresting/easy/etc. for you, then it is so for everybody else, especially for the person who came to seek help.
It is ironic that it is I who is saying it because my answer to the first of two Paul's semi-rhetoric questions
Do you behave in this way towards your own advisees when you meet them face-to-face in your own universities?
is "Yes", though my answer to
Do you not have the imagination to turn naive questions into teaching opportunities?
is the same.
So, if I feel that something is wrong here, it means we are, probably, quite far overboard in this respect. I would prefer the practice of complete closing restricted to obvious spam and trolling only with the move to MSE as a standard for repelling the question from MO from now on. Compared to most of what is normally there on the front page, almost all the questions that were closed here with scolding or contemptuous remarks are pearls.
The other thing (which is the common problem of all fora with few separate streams) is that the questions flow into oblivion way too fast for more than a few persons who visit frequently and get interested enough in some of them to keep track of what's going on. The MSE is literally overflown by now and MO is coming close to justifying the second word in its name too. We need either to increase the number of streams, or to design some user friendly tools for keeping track of what they want to watch closely. So far, we have just favorites (which, as you can easily see if you bother to look at my profile, stretch for 24 pages for me as of the time of this writing, so I wouldn't call that heap "easily manageable" plus I always have some moral discomfort when removing the question from the favorites not because I don't like it any more but just because it is completely solved and I learned everything I wanted to learn from it) and notifications (and no, thanks, I do not want more e-mails than I have already). Of course, I'm literate enough in computer programming to write my own bookkeeping tools but first, I'm lazy and overwhelmed with other stuff, and second, I would prefer an integrated tool which would show when new answers appear, etc. rather than just allow me to have the links organized in a minimally decent way.
Other than that, we seem to be doing reasonably well :-).
I'd hoped other people would opine before me, but just to get the ball rolling...
Partly I think it has to do with mathematical culture generally: mathematicians are exceedingly fortunate to have a criterion of certainty that allows us in most cases to come to agreement quickly. I'd think that the situation might be quite different in other sciences, or in philosophy, etc., where coming to agreement on the right answer is inherently thornier and more contentious.
We're also lucky in that MO was so thoughtfully and carefully founded by some energetic and talented young people like Anton Geraschenko. (I hope no one minds my singling him out; I don't imagine anyone will.) Imagining Anton as a kind of hub for this activity, I think he and the people around him didn't just go at this casually, but developed a vision about what it would take to make this work. They must have talked and written each other back and forth a lot, and had the moxie to invest the requisite time and energy. I have no doubt they were aided by the vibrancy of the department at UC Berkeley (like their 20 questions seminar, now writ large by MO) and also by word-of-mouth and advance notice, e.g. through the influential blog Secret Blogging Seminar (founded mostly by grad students and past grad students at UC Berkeley). The fact that some really big names entered the arena (for example, blogging Fields Medalists!) also made a huge difference. Is this somehow peculiar to the world of mathematics? I don't know. But anyway, by now we really have a stellar array of mathematical power.
So preconditions for success were excellent from the beginning, and as we all know, MO made a big splash on its inauguration. A lot of people then got a sense of how MO could benefit them as professionals, and wanted to keep it going (aided by the SE model where greater "reputation" means a greater share in moderation). (Again, you know all that.)
The continued success of MO is of course another question, and cannot be taken for granted by anyone. One thing it takes is a small army of watchful eyes who take the time to tend to the garden; it requires a lot of careful pruning and nurturing. Just to name a few (and with regrets for not naming others also deserving of mention): Andres Caicedo and Ricardo Andrade and quid are three examples of people who really help keep tabs on the site. All the people who invest time and energy to help run a tight ship deserve thanks.
But feedback about how we can do even better is always welcome.
Let me chain together some weak syllogisms.
MO is such a success because it has many excellent mathematicians reading it.
They read it, because the ratio of interesting mathematics to uninteresting is high.
For a great many of them, non-research level questions count as uninteresting.*
I think at this point the existence of MSE, and the strong impetus to push questions there if they would lower this ratio, is helping to keep around a lot of excellent mathematicians.
*And for a great many others, they don't. Some of those mathematicians read MSE also.
I would like to answer on the closing policy, which in my opinion is a crucial point. I truly feel we are too permissive to allow MO to truly fulfill what I believe is its main goal: have a place where as many mathematicians as possible are available to answer research-level question, more or less in the same way we are available to answer our lab neighbors questions.
Often, I feel that a borderline question deserves to be closed because it does not follow the guidelines of "how to ask". Sure, it happens that a question be closed as homework while it is in fact intrinsically legitimate, but it is up to the OP to clearly state the background and its motivation, not up to the people that might have the answer. Also, some questions are asked before the OP has thought them through; since asking a question is asking for busy people to help, it deserves a great deal of care.
Maybe we should seek a way to make closing less humiliating, if indeed it is felt that way. The goal of closing is only to keep the focus of MO, which is different from MSE; it can happen to everyone to ask a question that turns out to be considered inappropriate, there is no need for it to be such a big deal.
I know some colleagues that have stopped using MO because the front page contains too little relevant questions; and I might end up the same way. Getting good answers may become difficult, but more certainly this trend prevents MO to attract more people and develop significantly.
Overall, MO is doing well as a research level site and community and has become a very good research tool bringing together junior and senior mathematicians in different places of the world and different areas. The coverage in terms of areas of mathematics has improved substantially especially as very good mathematician from new areas entered the scene. (It is remarkable to see what a difference a single of a few mathematicians can make.)
There are still areas of mathematics which are not well covered, and it can be useful to make some effort in expanding the scope of MO. One concrete suggestion would be to devote some time to specifically encourage questions in a "featured" area. (of course, while maintaining the regular activity) Like "MO Applied Mathematics special months May-June 2015". (If this will become routine we can even devote special concentrated time to well-represented areas, as well.)
Is there a poll feature on this platform? Perhaps a version of this question could be taken to the main site. One could have a short main poll which would present two or three basic questions (is the level good for you? Are the right questions being closed?) along with a text response (what would you like to see to improve this site?). In addition, a longer poll with more specific issues could be offered, with the idea that the corporation MathOverflow Inc. is using the information to help determine the direction MathOverflow is going.
I imagine most lurkers would not participate in a poll on meta. (Indeed, I was surprised at the total vote count for the recent moderator election.) They may not even like to fill out the short poll, but they probably would if they thought it would determine the immediate future of this forum. One could use the longer poll data in relation to the short poll data to see the level of committment (size of core group vs size of participating group) of the community.
Gerhard "Ask Me About System Design" Paseman, 2013.11.05
I would like to thank Todd (whom I know to some extent) for telling us how MathOverflow came to be set up and also the people he mentions (whom I don't know) for doing it.
I see it as an extremely valuable site because it is the only place so far as I am aware where mathematicians of all disciplines congregate. However, it is precisely for that reason that I think it worth expressing a dissenting view.
I would like to see an end to the practice of closing questions (unless they are actually spam, offensive, etc).
Painfully frequently I see the high reputation users closing questions that plainly come from beginning graduate students, thereby humiliating them. Remember that these are your future colleagues and Fields Medalists. Do you behave in this way towards your own advisees when you meet them face-to-face in your own universities? Do you not have the imagination to turn naive questions into teaching opportunities?
Closing questions on the grounds that they are textbook ones is necessarily inconsistent. Quite difficult points of say number theory get closed when outsiders might have appreciated some explanation. Questions in logic that are comparatively easier (in my opinion) don't get closed; rather we have users like Joel David Hamkins who take the trouble to explain things, and it's a good thing they do.
Then there are the extra-mathematical questions, such as those about history and methodology.
Rhett Butler gave some excellent factual answers to numerous historical questions but somehow he no longer participates, probably on account of the (in my view hostile) reception he's gotten.
There was a question about Bourbaki and categories where I could have found people outside the site to give authoritative answers. But, had I done so, they would have arrived to find the question closed, which is insulting to them.
Amir Asghari asked a question about rigour that led to a flamestorm. However, it was only after that was over that I myself read his profile carefully and saw that he was trying to understand the cognitive processes behind doing and learning mathematics.
The value of this site is as I say that it encompasses many disciplines. Its participants should recognise that and value the role of people from outside the old core areas.
