-2
$\begingroup$

I wish to advocate changing the MathSE standards on posted questions. The change would be to allow opinion based questions, solely where the question is tagged book-recommendation, and the posted question is in fact limited to asking for such a recommendation.

On the one hand, it is plausible that a question poster can carefully construct such a question that is not opinion based. However, such a construction will often not approximate what the question poster is trying to ask.

I have noticed the following trends in opinion based book-recommendation questions:

  • Such questions are eventually closed as opinion based.

  • Prior to such closure, such questions typically receive extremely insightful responses. Such insights not only contrast one possible recommendation against another, but often attempt to relate a recommendation to the math background and motivation of the question poster.

    So, it is not uncommon for some MathSE reviewers to regard the posed question as worthy of a response.

  • My experience with internet searching is that (in general) the response of MathSE reviewers to such questions is irreplaceable, with (for example) Amazon book reviews coming in a poor second. In fact, I would take a (multi-) responded MathSE book-recommendation question over a private conversation with a single college math professor.

My bias is that (in effect) these questions + responses represent MathSE articles that are of very high quality, and make worthy additions to the MathSE book repository of questions.

My advocacy is not intended to affect the closing of such questions as duplicates. In such an event, the question poster can refer to the linked duplicate, and if needed, edit their posted question, explaining very carefully why the linked duplicate represents an insufficient response.

$\endgroup$
11
  • 2
    $\begingroup$ I agree that other websites don't even come up to the level of Math.SE, even on questions that aren't this site's core focus like book recommendations. This site is actually burdened with taking on too many jobs at once, discussion forum, repository, teaching people math etc. due to a lack of alternatives. I've written a lot here and on any other website it'd be irretrievable by search and lost underneath distracting comments. $\endgroup$ Commented Sep 28 at 16:37
  • 2
    $\begingroup$ Do we have any reason to suspect this is a big problem? For example, this post with the book-recommendation tag became a network hot-question last week, and appears not to have drawn any close votes. SEDE only reports 9 questions with the tag closed this year and 400 all-time (although I think that may be masking deleted ones). OTOH we apparently have 3000 nonclosed book-recommendation posts... $\endgroup$ Commented Sep 29 at 17:32
  • $\begingroup$ My impression was that most of the time, they were being closed as duplicates. I haven't been able to figure out how to query the exact reason for closure with my search... $\endgroup$ Commented Sep 29 at 20:46
  • 4
    $\begingroup$ I think the "opinion" that people object to in book-recommendation questions revolves around reasons given by D.W. below, where the question is trying to force the issue with "what's the best" or "what's your favorite" or especially "here's my bio ... what's the best for me?" A question that just asks "what are good quality books on <subject X>?" I think is generally accepted, assuming it is not covered by other posts. The problem is, though, that a great many subjects are already covered by previous posts (hence the tendency for duplicate closures.) $\endgroup$ Commented Sep 30 at 13:20
  • 2
    $\begingroup$ @rschwieb I think asking "what's best for me " given a specific ability, motivation and education is a very worthwhile question. I also think that a reviewer who provides an opinion based on such personal information is providing an extremely valuable service. Further, I think that the corresponding nuances (i.e. based on the provided personal information) in evaluating one book against another make for a high value article. ...see next comment. $\endgroup$ Commented Sep 30 at 19:20
  • $\begingroup$ @rschwieb Also, as I indicated, my proposal does not affect closing the article as a duplicate. It is then up to the question poster to explain why the linked (previous) article does not answer their question. $\endgroup$ Commented Sep 30 at 19:21
  • $\begingroup$ @user2661923 I recognize that (and welcome it) but I still wonder how often the types of questions you indicate are closed for reasons other than duplication. It looks a bit anecdotal at this point. $\endgroup$ Commented Sep 30 at 19:38
  • $\begingroup$ @rschwieb A MathSE search for [book-recommendation] (then webpage searching for "closed") gives this article, this one, and this one. ...see next comment $\endgroup$ Commented Sep 30 at 21:18
  • $\begingroup$ Strangely enough, the last two were both closed with the stated reason being "Questions about choosing a course, academic program, career path, etc. are off-topic". However, both of these two articles were in fact asking for a book recommendation, not a course. Personally, I distinguish between asking for a book recommendation and asking for a course. $\endgroup$ Commented Sep 30 at 21:18
  • $\begingroup$ OK, I went back to SEDE (which was flaking on me yesterday night and this morning) and did a hand count. There were 24 closed questions with book-recommendation in the tags. 9 duplicate, 7 opinion based, 3 off-topic/site-quality, and 5 clarity closures. $\endgroup$ Commented Oct 1 at 15:35
  • $\begingroup$ Another clarification: 24 such questions to date in 2025. $\endgroup$ Commented Oct 1 at 15:41

1 Answer 1

Reset to default
5
$\begingroup$

I don't think we should adopt a policy that all book recommendation questions are accepted.

I do not think the site should accept questions that are purely subjective, e.g., "What's your favorite calculus textbook?", "What's the best calculus textbook?", "Which calculus textbook would you recommend?", etc. See our help center, which says:

To prevent your question from being flagged and possibly removed, avoid asking subjective questions where …

  • every answer is equally valid: “What’s your favorite ______?”

https://math.stackexchange.com/help/dont-ask

Such questions tend devolve to an opinion poll. In an ideal world, a opinion poll of experts can be quite helpful, as the ability to aggregate expert judgement in a particular domain can be powerful. However, we have long experience with opinion polls like this: they might start off well, but over time they often end up badly. Often the vote counts reflect factors other than quality (e.g., answers posted earlier tend to have more votes). Often such questions become out of date, or accumulate many answers. Allowing such questions tends to attract many low-effort low-quality questions that are of dubious value to others. As a result, many have concluded that the site is better off without such questions.

I think it could work to accept high-quality questions about what to read on a topic, if the questions meet a set of strict criteria. But before trying to propose such a thing, I suggest looking at the criteria that other sites have set for such questions, and their experience. See, e.g., https://literature.meta.stackexchange.com/q/2, https://literature.meta.stackexchange.com/q/534, https://softwareengineering.meta.stackexchange.com/q/2003/34181, https://softwareengineering.meta.stackexchange.com/q/2665/34181, How can we discourage book recommendation questions?.

If anyone is considering asking for a book recommendation, and stumbles upon this page: I recommend you identify the requirements and criteria you will use to evaluate an answer, and communicate those requirements and criteria (e.g., what background you want the book to be at, what topics it must cover); and please do some research before asking and show why the books you've found don't meet your needs. A good way to find candidate textbooks is to look at courses on the topic that are taught at reputable universities and see what textbooks they use.

$\endgroup$
4
  • 1
    $\begingroup$ "What’s your favorite...?" : This isn't the typical question. Typically, the question poster is self-studying, is about to tackle a new subject, and wants to know which book on this subject is best for their motivation, education, and ability. This will very often result in MathSE reviewers sharing insights that are very valuable to the question poster. My perception/bias is that this results in an article of high quality. ... see next comment $\endgroup$ Commented Sep 30 at 5:30
  • $\begingroup$ The usual quality standards (except for opinion-based) are still in play. For example, part of the MathSE standards is that the question itself should be of general interest (i.e. the question poster is required to show context - why is this question being posed, what is the value of posing this question). Granted, this particular standard is often relaxed when the question poster shows significant effort. However, a question such as "what is your favorite..." would still be closed as being a question of no general interest. $\endgroup$ Commented Sep 30 at 5:33
  • 1
    $\begingroup$ @user2661923, I get it. Unfortunately "Which book is best?" is a matter of opinion and subjective (it's similar to "Which book is your favorite?"), and unfortunately "Which book is best, given constraints X?" is too. The level of education, ability, etc., can be used to narrow down the candidates, but choosing within the resulting set of books remains an opinion poll / aggregating opinions. $\endgroup$ Commented Sep 30 at 6:42
  • 2
    $\begingroup$ I do get what you're trying for, and if we could make these questions work well in the long run, it would be valuable. Setting specific quality standards seems like a plausible angle to try to push on. But it's just challenging to avoid the problems of these types of questions, even with quality standards. Hence why I suggested studying what other sites have tried, and what their experience has been. $\endgroup$ Commented Sep 30 at 6:42

You must log in to answer this question.

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.