I think that there should be a list of frequently asked math questions. Let a question be called a sub-duplicate of another question if it is that other question or it is closed as a duplicate of a sub-duplicate of that question (This definition is inductive.) We should have a list of questions together with their number of sub-duplicates with questions with high numbers of sub-duplicates at the top, so that new users know what questions are frequently asked. I think that this might reduce the number of duplicate questions on this site.
I'll at least mention what already is available if you look only at the direct duplicates (rather than recursive duplicates). And let me mention various other lists that exist.
- There already are some manually created lists of questions. Some of them which are available on the site are mentioned in the part about "Lists of questions" in the FAQ post: How to search on this site? (And some of such lists have already been mentioned in the comments.)
- There exists a tag called faq on main. (And there is a tag faq on meta, too - but I suppose we're discussing mainly the main site.) And you can combine this tag with another tag or with any other search.
- On the main site you have the frequent tab - which is based on the number of links. And you can look at the frequent tab for any tag or combinations of tags.
- You can create various queries in SEDE - there is a table PostLinks where
LinkTypeId=3means duplicate. (And
LinkTypeId=1means a link.) See also: Add a PostLinkTypes table to SEDE and Add a PostLinkTypes table to SEDE.
Here are some examples of such queries - they only count "direct" duplicates not "recursive duplicates". They are from this answer on Meta Stack Overflow: Most frequent questions for a tag combination - but the linked queries run on MSE rather than SO. You can use parameters of those queries to change whether you look at duplicates or links.
- Most frequently linked questions
- Most frequently linked questions in the given tag
- Most frequently linked questions with tag1 and tag2
Creating a similar query with looking at all "recursive duplicates" in SEDE might be difficult. It goes beyond my knowledge of SQL. (Moreover, a query trying to do so for all posts at the same time would likely time out.)
Here you can find a query which looks at a single question and finds the linked posts: Get all interlinked questions. I have modified the query to look only at duplicates - and not all links. You can change the parameter Level to increase the depth of the search.