Motivation 1

Currently the custom off-topic reasons for MathOverflow read:

  • This question does not appear to be about research level mathematics within the scope defined in the help center.

  • Questions about homework are generally off-topic. MathOverflow is for mathematicians to ask each other questions about their research.

There are some rather obvious overlaps between the two reasons, and I think we can perhaps optimize a bit.

Motivation 2

As a moderator on Math.StackExchange, recently we have had some discussions about the closure of migrated questions. As you may already know: if a question is migrated to another site, and if it is closed there for anything other than "Duplicate", the question gets bounced back to the original site, the migration canceled, the question now just put on-hold as generically "off topic", and the user who asked the question originally may be none-the-wiser.

You may be less likely to know that recently, due in part to the increased volume of questions, and in part to many members' preference to teach mathematics instead of handing out homework solutions, Math.StackExchange has implemented a "Lack of context" close reason, trying to encourage users to include the source/motivation of the mathematical problems that they are asking about, as well as any partial work that they have done toward resolving it.

Motivation 3

Already at times we close the question as off-topic, only to leave a comment stating that perhaps MSE is a better place for it. My proposal makes for less typing by the users.


We can keep the first off-topic reason (for questions that are not even mathematical), but change the second one to the following:

  • MathOverflow is for mathematicians to ask each other questions about their research; general questions in mathematics should be asked at Math.StackExchange.


In the spirit of "teach a man to fish", I would advocate for using this custom close reason instead of direct migration when dealing with all but the highest quality (but unsuitable for MO) questions. This way the user learns to register a new account on MSE, and ask his questions there. And if his question is not of a sufficient quality for MSE, he will be told why by members of MSE. Ideally this will keep users who do not care about research level mathematics from coming back to MO time and again.

Final remark

Math.SE is rather more tolerant of questions, overall, than many other sites. In the past three months, MSE tried to migrate 10 questions to Physics with a staggering 70% rejection rate [Edit: this I checked by hand, and the number is accurate. See comments below], whereas Physics dumped 40 questions on MSE with only 5% rejected.

MO is by far the largest migrator to MSE, with 185 questions in the past three months, of which 21% were rejected [Edit: this 21% may or may not be accurate. See comments below]. 11% were rejected (see comments below). The fact that there are still, on average, two questions per day (just shy of 10% of the questions asked per day) which are migrated from MO to MSE (and I am sure a few others that are just closed as off topic without migration), means that we are still not getting the message across about the difference between MO and MSE. Maybe this will help a tiny bit.

  • 3
    $\begingroup$ According to the data I get there are 188 question and only 11 percent rejected (not 21). The discrepancy of 3 in total count seems easily explained by the moving window, but 21 vs 11 is a bit much for fluctuation. Could you please double check this number. In fact I was extremely surprised by the over 20 percent rejection rate. Indeed I am even surprised by the 11 percent as last time I checked it was around 5, but the change in math.SE policy (lack of context closure) could well explain this. $\endgroup$
    – user9072
    Commented Nov 21, 2013 at 11:39
  • 1
    $\begingroup$ @quid: I double checked. It now says 20 percent. Where did you get your data? (Mine is from the Moderator Tools -> migrated -> stats on MSE). (Incidentally, $20 / 188 \approx 0.11$, I wonder if there is an extraneous or missing division somewhere.) $\endgroup$ Commented Nov 21, 2013 at 11:45
  • 1
    $\begingroup$ Thanks for the reply I got it on MO (10k-tools I assume), mathoverflow.net/tools/posts/migrated/stats and here one finds the questions that got rejected mathoverflow.net/tools/posts/migrated/rejected the total count of which seems a lot more in line with 11 percent than 21 percent; the missing division reasons seems feasible. $\endgroup$
    – user9072
    Commented Nov 21, 2013 at 11:50
  • $\begingroup$ @quid: I don't have access to 10K tools here. This discrepancy is slightly worrisome. But since the info is available to 10K, I can post publicly. Do you mind posting a screencap as an answer below? I'll grab it and ask about it on Meta.SO. (I would have asked for it to be e-mailed, but I know you value your privacy and all.) $\endgroup$ Commented Nov 21, 2013 at 11:56
  • $\begingroup$ At the moment this is difficult to do technically (although mainly as I am incompetent it needs to be admitted). Perhaps "double-moderator" Mariano would be a good person to check; or a more savvy 10kuser. But if nothing happens before I will follow up on this in half a day. (If you prefer, I can send you an email, no problem.) $\endgroup$
    – user9072
    Commented Nov 21, 2013 at 12:03
  • $\begingroup$ @quid: that would be appreciated. I'll also ping Mariano to see. I'll leave the method of communications up to you. $\endgroup$ Commented Nov 21, 2013 at 12:04
  • 4
    $\begingroup$ I'm looking into the discrepancy, I can't readily explain why it differs between sites. $\endgroup$
    – Tim Post
    Commented Nov 21, 2013 at 12:53
  • 2
    $\begingroup$ I like the Proposal for wording change. $\endgroup$ Commented Nov 21, 2013 at 13:50
  • $\begingroup$ Since Tim Post seems to be investigating the matter already I assume there is no more need for documenting it further. If there is still need for the screencap please let me know. $\endgroup$
    – user9072
    Commented Nov 21, 2013 at 13:58
  • $\begingroup$ @quid: that's correct (the screencap is no longer necessary). Thanks. $\endgroup$ Commented Nov 21, 2013 at 14:03
  • 8
    $\begingroup$ Okay. The "sending" site (in this case, MO) counts actual rejections, while the destination counts all questions closed or deleted - even duplicates, thus it's typically higher on the destination. The metric on the sending site is what's important, which is 11%. more here $\endgroup$
    – Tim Post
    Commented Nov 21, 2013 at 16:01
  • $\begingroup$ I like the proposal too, and am interested in hearing other opinions. $\endgroup$ Commented Nov 23, 2013 at 21:48
  • $\begingroup$ Actually, I don't like much the first off-topic reason. It sounds a lot like "off-topic question is off-topic"; it is not helpful at all. Maybe it should be removed, or maybe only kept as a last-resource (i.e., it should not be the first one in the proposed list). I'd rather like one that explicitly says "too easy" instead. (Incidentally, we discussed about a reason similar to the first one on academia.se; I had the opposite opinion initially, but the other users there managed to convince me.) $\endgroup$ Commented Nov 26, 2013 at 11:27
  • $\begingroup$ @FedericoPoloni: I disagree. We can only have up to three custom close reasons, so we cannot possibly be specific about every possible scenario. There should be a fallback option that says exactly "off-topic is off-topic". At the very least the first option links to (I suppressed it in my copy-and-past above) the FAQ telling the user what is and is not on topic here. $\endgroup$ Commented Nov 26, 2013 at 12:02
  • 1
    $\begingroup$ It seems that a question what is actually the difference between the two close reason comes up on this meta quite regularly: Custom closing link looks too much like it should be a migrate to MSE link, Redundant ways to close as off-topic/better on MSE and What's the difference between two off-topic flags? $\endgroup$ Commented Jun 13, 2018 at 3:53


You must log in to answer this question.

Browse other questions tagged .