# Mathematics StackExchange vs Mathoverflow

I was curious about the difference between Maths StackExchange (MSE) and Mathoverflow (MO), because it was never obvious to me. After a couple of minutes it became clear that MSE is aimed at anybody with questions in maths, while MO is dedicated to research level questions. However, while MO claims "it is for professionals", by the nature of the website anybody can ask questions.

This directly lead to two questions:

1. Wouldn't MSE be superior even for higher level mathematics? If someone has a very specific/difficult question, I feel that asking the website with more people is a better bet, as it is more likely to be seen by someone familiar with the ideas being discussed.

2. How does one decide whether or not a question is "high level enough" for MO? For example, I'm still a student, but I'm relatively knowledgable in a field we call "Generalised Inverse Limits" (an approach to continuum theory, in topology). I'm expecting to publish sometime this year. However, I'm most definitely not a "professional" or an "academic", while some of my questions may be "research level". In this sort of scenario, what determines whether a question should be posted here or on MO?

• Try, you'll see if your question is or is not downvoted in less than 30 min – reuns Apr 2 '17 at 5:05
• The strangest coincidence is that one of the "Hot Network Questions" featured right as I was reading this was this one from MO: mathoverflow.net/questions/266089/… The question was related to algebraic topology, and asked by a high-schooler. It was quickly answered by the author himself. The asker had cross-posted to MO from MSE, but the question hasn't been answered on MSE (and now might as well not be, as the author has already weighed in). – Deepak Apr 2 '17 at 6:19
• This is very subjective. See this, for instance. – Renan Maneli Mezabarba Apr 3 '17 at 14:34
• as an idea of how to use both: I usually post here as a default, if something goes un answered i'll move to math overflow, sometimes if i'm very very lucky someone with the expertise can crack it, but usually math.se does a pretty good job at indicating if i'm working on something intractable/beyond what current machinery supports – frogeyedpeas Apr 4 '17 at 6:05

1. No, since the noise to signal ratio is going to be terrible. You have a much better chance on a site whose users are all experts and which gets less traffic.

2. A rule of thumb is that if you're not sure, post here. Worst case scenario, you can always go on meta.MO and ask if that question is suitable, or something like that. A slightly more specific rule of thumb is that if your question originates from actual mathematical research (as opposed to a hobbyist trying to prove the Riemann hypothesis), it's probably going to fare reasonably on MO, granted you formulate it properly.

• I think the noise-to-signal problem may be solvable by tagging. The level of the problems could be tagged like Knuth did it in the TAOCP. Although it would require a lot of work (maybe doing it retroactively would be even unpractical). Typically, the questions representing the "noise" would be tagged by Knuth between $0<hardness<=9$. – peterh Apr 8 '17 at 23:38
• Trust me, it cannot be solved by tagging. If anything else, I am getting a bit exhausted of misusing the tags on the site. If the solution to the noise requires you to actively filter out the noise, then it's not an actual solution. – Asaf Karagila Apr 8 '17 at 23:50
• The PSE tries to make the "noise" simply offtopic, also this requires active - manual - filtering, meanwhile they lose also a lot of good content and aren't very nice guys. The noise problem is a big disadvantage also on the StackOverflow, currently they seem to follow the MathSE way, but it wasn't always so. They have also a special review queue for quality improvement, it exists only on the SO, but the MathSE is the second largest SE site, maybe could it work also here? – peterh Apr 8 '17 at 23:57
• What do you mean by "PSE"? – Asaf Karagila Apr 8 '17 at 23:58
• physics.stackexchange.com - I was always surprised how could be these sites so different, while they are very similar from the SE viewpoint. – peterh Apr 8 '17 at 23:59
• I don't know if they are so similar from the SE viewpoint. – Asaf Karagila Apr 9 '17 at 0:00
• Both are from hard science, their communities have a very broad spectrum from novice elementary school stundents until researchers. The problems what the sites are facing are to me practically the same. They are also much smaller (around 10% of size of the MathSE), which may be the result of their active noise-filtering customs. Note: the stats of both sites have strongly increasing tendency. – peterh Apr 9 '17 at 0:03
• Do you mean Physics and MathOverflow? In that case, I agree they are sort of similar. I thought you compared Physics with this website. – Asaf Karagila Apr 9 '17 at 0:04
• No, I compared the Physics.SE to the MathSE. – peterh Apr 9 '17 at 0:05
• What I tried to ask (sorry for the inconvenience), that this noise problem exists also on other sites, too, and the SE has many ways to solve them. One of them is to change the site topic into a more narrow direction, this has disadvantages and may be not a reachable direction. A second is to tag the questions by their level. A third is what the SO does: they have 2 additional review queues for the quality improvement. Are you sure that all of them are impractical? – peterh Apr 9 '17 at 0:08

• MathOverflow is being done by a non-profit organization which pays for the SE for their site.
• MathSE is an SE site about mathematics, started by the community with the ordinary Area51 site construction procedure.

The MO was started in the era as the rules of the SE weren't so strongly fixed. As far I know, currently this option is not available any more (more exactly, being the SE part of the free market, probably everything has its price, but to develop an SE engine clone may be a cheaper option). But practically no U.S. company would ever expel a paying customer, so they can remain.

Thus, the MO is partially out of the common rules binding the CMs, while the MathSE is not.

In most cases, the SE doesn't really like to maintain similar sites, for example this is why the ancient theoretical physics were integrated into the PSE. Regularly occuring initiatives to create a similar dual-site solution also in the physics world were rejected by the SE (although a PhysicsOverflow exists, it is a site out of the SE network). The only alternative dual-site solution exists in a quite different topics, the English Language Learners and English Language and Usage.

Here you can ask any level questions, even research-level ones and you will have a good chance to get an answer. You can also ask your homework problem. It is because mathematicians are good people liking enthusiast-level things.

If you are not a research-level Mathematician, I think it is better to honor them from a distance. If you are, I think the best is if you hold the front here, and improve them there. :-)

• It would do a lot of good for people to stop thinking in terms of "topics" and start thinking in terms of "communities" when it comes to websites on the SE network. From a sufficiently distant viewpoint, Superuser should include Unix, Ubuntu, and Server Fault; whereas SO should incorporate a bunch of other websites also. Sure, the topics of MSE and MO are similar, but they cater to entirely different communities (even if these have nontrivial intersection). – Asaf Karagila Apr 9 '17 at 0:08
• @AsafKaragila Well, logic may dictate quite different things as we experience here. In my opinion, the database of all sites should be melt into a single one and tagging, or meta-tagging should decide, which questions will exist on which sites. Of course also the reputation structure should be adapted to it. The current behavior of the SE shows that any significant change has only a near-zero chance, but important improvements may be reached also by not significant changes, too. For example, I think the quality improvement queues of the SO could maybe significantly reduce the noise problem. – peterh Apr 9 '17 at 0:23