# How to prevent “no clue” questions?

A significant amount of questions asked in MSE begin or end with "I don't even know where to start", as a way to justify one's inability to provide own work on the problem and/or attract sympathy or rather pity.

Most of these questions end up [on hold] within half an hour, because of course this doesn't attract sympathy at all. This is a waste of time and energy for everyone, including the OP, the five reviewers who are going to close the question, and the good Samaritan who has answered in the meantime.

There are plenty of ways to get started on a problem when one has "no clue":

• Write down the definition of the keywords of the problem, thus making sure you understand them, using examples ($\star$) when applicable;
• If the problem involves formal computations, try with specific settings first ($\star$);
• If the problem involves large structures or numbers, try with lower numbers first ($\star$);
• Write down what you know that seems related to the problem: any relevant theorem not in that list will be spotted right away and people will point it out easily.

($\star$) you have to make them up yourself, and that very process is excellent to make progress in the way you think in general, what is a good, representative example in a given situation?

I suggest that the Ask Question form could suggest that, if one ever feels like including "I don't even know how to even start", one could consider trying one of the options above instead, to save time and energy for everyone.

A kind of "I'm not a robot" feature.

Edit: I wrote an answer to How to ask a good question based on this post, following the recommendation of @Jack D'Aurizio. It is community wiki, feel free to improve it.

• Sounds like a reasonable improvement of the policy outlined in How to ask a good question. – Jack D'Aurizio Feb 21 '18 at 19:14
• @JackD'Aurizio: and the guideline in Help Center as well: math.stackexchange.com/help/how-to-ask – Jack Feb 21 '18 at 19:56
• I think it's a bit unfair. A lot of students at one point or another are just not sure as to how to approach a problem. They don't know how to do that. I agree that this is not a great situation, and indeed the ideal response would be to take the "teach a man to fish" rather than "give a man a fish" and explain how to solve problems. That being said, I agree that some nontrivial percentage of people use this as a blanket term to avoid needing to work, or get some sympathy help. [...] – Asaf Karagila Feb 22 '18 at 17:17
• [...] All the more reason why an abstract answer about how to solve problems is better, since it would force them to work anyway. (And a bit of self promotion never hurt anybody... except when it does. ;)) – Asaf Karagila Feb 22 '18 at 17:17
• @AsafKaragila Clearly a number of those who ask such questions genuinely have no clue as to how to approach them; but that doesn't mean that they know nothing. If they really don't know a single definition of the words they are using, then they are asking the wrong question. – Arnaud Mortier Feb 22 '18 at 17:21
• I've had a lot of experience with students coming to my office hours claiming that they don't know how to solve some of the problems. And walking hand-in-hand, it was clear they know the definitions, and they know the theorems, but they don't have the confidence to apply them and see where this is leading. This is mainly as a result of terrible K12 indoctrination (at least in Israel) that math is to be solved via a concrete series of steps, and not as a free-form brainstorming. – Asaf Karagila Feb 22 '18 at 17:23
• @AsafKaragila: the terrible k12 indoctrination applies to India also and judging from many questions on MSE I think it's a global phenomenon. There has to be a formula for each and every kind of problem and one just needs to memorize all the formulas possible. – Paramanand Singh Feb 23 '18 at 3:13
• Can't we suspend such users for sometime who routinely ask such "no clue" questions? The decision can be taken on the basis on a threshold number of consecutive questions put on hold because of lack of context. – Paramanand Singh Feb 23 '18 at 3:24
• A Question ban is probably something you are looking for? @ParamanandSingh – user99914 Feb 23 '18 at 3:33
• @JohnMa : yes! Based on the history of questions one can decide the ban period and let the asker come back with some improvement. I don't know if that helps but sometimes one has to use stick also instead of carrot. – Paramanand Singh Feb 23 '18 at 3:56
• Great, Arnaud. Good Samaritans could now paste in a canned response as follows: "The following links to advice about asking questions, and about getting yourself unstuck. math.meta.stackexchange.com/a/27933/124085" – fredgoodman Feb 24 '18 at 2:32
• These older discussions are somewhat related: Suggested Guideline for “I Don't Know Where to Begin” Questions and Homework, reasonable to have no clue? – Martin Sleziak Mar 19 '18 at 1:19

In the meantime, here's a proposal:

1. Don't answer the question. Don't even provide a hint.
2. Leave a comment informing said clueless person that this site is a Q & A site for people who have specific questions, not a homework mill.

If you are nice, you may provide some tips to the clueless user that will help him/her ask a proper question, i.e., something that shows an investment in the question and material.

**I do not endorse the idea that a person willing to "help" such a questioner is doing anything good. See, e.g., the reason Tahani Al-Jamil is in the Bad Place.

• Isn't voting to close the most urgent action? – user99914 Feb 23 '18 at 2:07
• Yes and perfectly valid. But I am also expressing distaste so that it reduces the likelihood of this happening so much. I mean, we will always get lazy people who think they have discovered El Dorado or something here. I have learned though that the ones who simply didn't understand what was expected of them will change their ways and ask better questions (good) and the ones that were hoping to have someone do their homework for them will simply disappear, never to return (also good). – Ron Gordon Feb 23 '18 at 2:10
• I disagree with (3). I don't see a good reason why the answerers should share the burden of a not-so-good-(yet) question (unless an answer itself is problematic). For instance, these two old questions: math.stackexchange.com/q/331404/9464, math.stackexchange.com/q/406200/9464 more or less would be closed very soon using today's "criteria". But I would be very reluctant to downvote your answers because they are really good! And I don't think they are detrimental to the site at all. – Jack Feb 23 '18 at 2:57
• @Jack: downvoting is not a reflection of the rightness or wrongness of the answer necessarily but a reflection of the usefulness of the answer to the community. Perhaps that sounds like splitting hairs but it is not. In this case, I feel that such answers are a detriment to the community because they encourage more of these "no clue" questions. Enough answers to these questions and Math.SE gets a reputation as a homework mill. No thanks. So I downvote to discourage people from answering these questions. – Ron Gordon Feb 23 '18 at 3:00
• For routine problem your 3rd point seems to make sense, but for questions which are not run-of-the-mill variety I think one should not downvote a mathematically correct answer (mostly this happens eg for questions with challenging integrals/series and no context). – Paramanand Singh Feb 23 '18 at 3:19
• @Jack: re my answers. The questions in these cases are clearly not run-of-the-mill homework questions but very difficult questions that occasionally get posted here. These too are controversial and nobody has really come up with a great criterion for deciding what is OK and what is not. (Which is why after 5 years I am still having these conversations.) I try my hardest to avoid answering clear HW questions and I apply my advice above when I deem appropriate. But difficult questions that have little context I think are OK if they are clearly not homework. – Ron Gordon Feb 23 '18 at 3:29
• I'm a simple man, and I'll be honest -- that allusion in the sidenote totally tipped the balance from $\pm 0$ to $+1$ for me. – pjs36 Feb 23 '18 at 4:26
• @RonGordon: Thanks for your explanation. I would wholeheartedly agree with you that "(answering) difficult questions that have little context I think are OK if they are clearly not homework", although I think the level of "difficulty" of a given problem may very much depend on the experience of the asker. – Jack Feb 23 '18 at 14:09
• I understand why one shouldn't answer the question, as stated in your first bullet point. But why not give a hint? This seems to assume that the person is "malicious" when he states that he has "no clue". It can be that his only "clue" is using the definitions and he feels that this would not add anything to the question, it can be that he has never faced a similar problem before, or a myriad of other reasons. If the person has the only intention of getting an answer for homework, I understand why not even a hint may be proper. But that is not only assuming the worst, but also unfalsifiable. – Aloizio Macedo Feb 23 '18 at 22:46
• Let me elaborate: If the sole purpose of the person is getting an answer asap, then a hint will most likely not give her that. If the person has legitimately "no clue", a hint can be very fruitful. It seems like a win/win situation, differently from answering straightly in such a situation. – Aloizio Macedo Feb 23 '18 at 22:48