It is not difficult to find soft questions on the website which do not clearly strike as research level. To give examples of some: 1, 2, 3, 4, 5. Yet some of them are upvoted so highly that even many research level questions do not share the status.

I could deduce that some of these questions actually would help researchers. But it is clearly not true for all of them

  • 8
    $\begingroup$ Conjecture: soft questions are more comprehensible to most participants, active and passive, on this site. A substantial research problem often requires a certain amount of specialised knowledge to appreciate its qualities. Hence the former have a much greater potential audience (and so group of potential upvoters). One can observe a similar phenomenon between branches of mathematics on the site--some are more accessible to a transparent formulation of (often quite deep) problems and this can be reflected in the comparative number of upvotes. $\endgroup$
    – memorial
    Jun 8, 2022 at 10:02

1 Answer 1


This is more of a comment or an observation than an answer, but perhaps it might give you some insights into your question.

All questions you have listed as examples are tagged or . I am not sure whether this was your intention to ask about such questions, but let me put here some stats (and a few comments) anyway. (Strictly speaking, this is not exactly what you asked - since you have formulated this as a question about "research-level questions". There are certainly some questions on MO that could be considered freshman exercises, but still gained enough upvotes. But for now let us look only at tags similar to these two - and some related statistics.)

  • It is perhaps not surprising that such tags can get more upvotes. They are more likely to have a bit "clickbaity" titles. And if somebody wants to say whether or not some question is interesting, they need basically no mathematical knowledge to say that a question asking for "examples where more a general approach made a proof easier" is interesting (and perhaps worth upvoting). Judging whether a question is interesting or not becomes more difficult for a question which needs working knowledge of some advanced area of math to be at least understood.
  • You can check the top tags sorted by average question score. The tags with the highest numbers are not the tags related to some mathematical discipline, but the meta tags such as these two - which can be used for questions from any area. (On purpose I have looked only at big enough tags - if you restrict the list even more to the tags with at least 500 questions, the tags and become the top two tags.) One could check the average answer score for tags with at least 100 or at least 500 questions - the situation is similar.
  • It is quite plausible that at least some of the upvotes came from the users outside MO, who simply noticed the question in the list of hot network questions. Here is the list of tags with the largest percentage of HNQs. These two tags are above 9%. (They are among the top tags if you look at the tags with at least 500 questions.)
  • If you look at the time when such questions were posted, it seems that there are less such questions now than in the past. You can have a look at the graph for the tag big-list and the tag soft-questions. You can see that since 2012 there were almost always less than 10 questions tagged and 20 questions tagged per month. When this site was relatively new, such questions were more frequent.
  • Closures could serve as another datapoint. These tags have a relatively high percentage of closed questions - even if we do not count duplicates. (Again, here are all closures and stats after omitting duplicate closures for the tags above 500 questions.)

You must log in to answer this question.

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