Whenever I browse the QUESTIONS tab in math.stackexchange, I notice that all the questions, even the very basic ones, are already answered. And I also notice that even elementary questions are answered by high-level users.

So I am suggesting that we start a culture where high-level users will only answer the very difficult questions that really require ingenuity or a broad knowledge of advanced math, and leave relatively easy or basic (precalculus, calculus, logic, discrete math, number theory, combinatorics, graph theory, linear algebra, real or complex analysis, recreational math) questions to low-level users, so we can also do our part in our community.

Or maybe this can even be integrated as a feature in StackExchange in which users will be required to input an approximate level of difficulty for their questions (say 1 to 10, 10 being the hardest), and high-level users can only answer questions that are level 6 or above.

Definition: The level of a user is determined by the number of points he has acquired.

[For more info, see comments below.]

  • 5
    $\begingroup$ Out of curiosity, what do you mean by an "advanced" or "high-level" user? How do you judge whether or not a user is advanced (or high-level)? $\endgroup$ – user642796 Jan 26 '15 at 10:51
  • 2
    $\begingroup$ I was thinking that level is determined by the points a user has acquired. But now that you asked, I realized that it can also be very subjective. For instance, using your better judgement, if you think a question is too easy for your capabilities, then just leave it unanswered for other users to resolve. $\endgroup$ – Mark Lao Jan 26 '15 at 10:54
  • 8
    $\begingroup$ There seems to be roughly 1000 unanswered questions tagged (algebra-precalculus): math.stackexchange.com/questions/tagged/… $\endgroup$ – Did Jan 26 '15 at 11:24
  • $\begingroup$ "... questions that really require ingenuity or a broad knowledge of advanced math ..." These unanswered questions belong in the first category. $\endgroup$ – Mark Lao Jan 26 '15 at 13:59
  • 15
    $\begingroup$ You assume that a high reputation implies great strength in a particular subject. This is patently false; for instance, see my profile. Although I am in the top 3% of users, I don't know much abstract algebra and have never taken real or complex analysis. Topology is foreign to me, and there's a lot of linear algebra that I don't know. If I was required to only answer "hard" questions, I doubt I could answer many at all. $\endgroup$ – apnorton Jan 26 '15 at 18:14
  • 6
    $\begingroup$ What @anorton said. Although I could probably answer 85% of the set theory questions asked here, I estimate the number of questions regarding integration which I can answer at roughly 0.05%. And look at me, I'm No. 6 (I AM NOT A NUMBER! I AM A FREE MAN!) on the all time reputation list. $\endgroup$ – Asaf Karagila Jan 26 '15 at 18:39
  • 1
    $\begingroup$ @AsafKaragila: I look at your comment and suddenly find it odd that I speak of elephants in my post below. $\endgroup$ – Ron Gordon Jan 26 '15 at 18:44
  • 1
    $\begingroup$ @Ron: Would you feel better if you speak of goats and rams in your post instead? $\endgroup$ – Asaf Karagila Jan 26 '15 at 18:53
  • $\begingroup$ @AsafKaragila: Perhaps. moderatenerd.com/2011/12/02/hungry-goats $\endgroup$ – Ron Gordon Jan 26 '15 at 18:55
  • 1
    $\begingroup$ @Asaf I may have said this before. I think I read this somewhere: He says "Who is number 1?" and the voice ignores him, saying "You are number 6." The alternate theory, consistent with the end of the series, is "Who is number 1?" and the voice says "You are, number 6!" $\endgroup$ – Will Jagy Jan 10 '16 at 3:50
  • 3
    $\begingroup$ @Will: As supported by the finale. $\endgroup$ – Asaf Karagila Jan 10 '16 at 5:10

Most high reputation users have gotten that way by answering questions well, whether those questions are easy or hard, low-level or high-level. I think it is very important for the introductory questions to be answered well as those usually cover the bases for understanding the more advanced concepts.

Between two people who can provide acceptable answers, the difference in reputation comes from the amount of time they have given to the site. I would not want to ask some of the people who have given some of the best answers on this site to slow down their efforts. It is no easier for high-reputation users to find questions to answer than it is for lower-reputation users. The difficulty in finding unanswered questions is due to the large number of people answering questions.

Furthermore, there is usually more than one way to approach a question, so the fact that someone has already answered a question does not preclude someone else from answering the same question in a different way. However, if, for some reason, you feel uncomfortable answering questions that someone else has already answered, you can go to this page and narrow the list of topics by entering tags in which you are interested. Tags should be entered in square brackets; for example, [algebra-precalculus], [definite-integrals], and [geometry].

  • 11
    $\begingroup$ I am suspicious about any link between true quality and rep. I think that rep is more a measure of time spent on the site and question selection. $\endgroup$ – Carl Mummert Jan 26 '15 at 13:37
  • 6
    $\begingroup$ @CarlMummert I agree that reputation doesn't faithfully represent the quality of a user, but one can agree that in the first page of top reputation users most have contributed with quality posts and not only answered simple questions or "low hanging fruit." $\endgroup$ – Pedro Tamaroff Jan 26 '15 at 13:55
  • 1
    $\begingroup$ I agree that introductory questions be answered well. And I believe that, given the chance, low-level users can answer these questions with flying colors. It's just that high-level users have already done it for them. And providing an alternative solution is easier said than done because high-level user after high-level user usually post elegant solutions, leaving a very small room for low-level users to shine. $\endgroup$ – Mark Lao Jan 26 '15 at 14:07
  • 1
    $\begingroup$ "It is no easier for high-reputation users to find questions to answer than it is for lower-reputation users." - This is partially true, but not entirely. High-level users reached their level primarily because of two things - as @Carl Mummert said, the length of their stay in the community, and as Pedro Tamaroff said, their high quality answers because of their vast knowledge and experience in math. $\endgroup$ – Mark Lao Jan 26 '15 at 14:11
  • 2
    $\begingroup$ So we are not requesting that high-level users slow down their efforts. We are requesting that their efforts be spent on the harder questions, so low-level users will also have a chance to contribute to our community. I imagine it as a game. Let's say high-level users have 100 MP/ hour worth of energy to answer math questions while low-level users have 30 MP/ hour. Easy questions require 10 MP to answer while hard questions require 50 MP. High-level users can attack 10 easy questions in an hour, or they can attack 2 hard questions. $\endgroup$ – Mark Lao Jan 26 '15 at 14:14
  • 1
    $\begingroup$ In choosing to do the latter, they are not "slowing down" their efforts. They are just being more efficient and more effective, not just for themselves, but for the community as a whole. High-level users should be empowered to answer these harder questions since low-level users have a significantly lower chance of answering them. This gives the low-level user a chance to answer 3 easy questions, instead of spending $1 \frac23$ hours answering a hard one. $\endgroup$ – Mark Lao Jan 26 '15 at 14:21
  • $\begingroup$ So in the end, no one really slows down. The community just made the assignment of difficulty of problems to ability of users more efficient and more effective. This is of course a highly simplified model, but I hope it made my point clearer. :-) $\endgroup$ – Mark Lao Jan 26 '15 at 14:23
  • 3
    $\begingroup$ @MarkLao: "High Level Users", by your definition, are simply users with more reputation. That means they've been here longer or answered more questions, not that they know more than anyone else. Currently, I have the $10^\text{th}$ highest reputation on math.SE, however, I am quite sure that there are many, many users with less reputation that know more and can answer question far beyond what I can. Does it make sense that I should only attempt questions that are harder than those that they answer? The problem with your suggestion is that you seem to be equating reputation with ability. $\endgroup$ – robjohn Jan 26 '15 at 20:27

I have been advocating something like this for a long time (and trying to follow my own advice). I agree it is better, as a general practice, for users with significant reputation to wait a while before answering routine questions, so that users with less reputation can answer them.

Unfortunately, it is difficult as a matter of practice for high-reputation users to not answer questions to which they know the answer. The software of the website is not designed to encourage that sort of self-control. Some may even disagree about the need to allow newer users to develop reputation in this way.

But remember that even if someone else has written an answer, you can always write another. Add your own spin on the question, or explain the answer in a different way. You may also find particular tags where you can help, and where there are not so many people competing to write answers.

  • 7
    $\begingroup$ Not to mention that the high-rep users got there by answering lots and lots of question, so it's even less likely for them to reduce their activity. $\endgroup$ – Najib Idrissi Jan 26 '15 at 12:00
  • 2
    $\begingroup$ Carl, it's nice to know that I'm not alone in this. :) I agree with what you said. The thing though is that the solutions these high-level users write are so elegant that I cannot do any better. Likewise, questions that are left unanswered are usually so difficult that even experts can't answer them. $\endgroup$ – Mark Lao Jan 26 '15 at 12:37
  • 5
    $\begingroup$ Related: Letting less-reputable people answer questions. $\endgroup$ – Martin Sleziak Jan 26 '15 at 13:47

I see you're a new user. We were all once "new users" and I had the very same impression as you when I started to participate: these darn high reputation users seem to be everywhere and new users go unnoticed. This is partially true, but shouldn't be something alarming. Users with relatively high reputation are users that have spent a fair amount of time in the site, and hence are known in the community, and hopefully trusted. Hence, there is a tendency for their contributions not to go unnoticed and for people to upvote their contributions.

It is great you're looking to participate and be noticed (note however that your answer to a post never goes unnoticed by the person who asked the question: they will get pinged and informed you have answered.) To this end, I recommend you simply answer questions in your best ability and be patient. Eventually, people will start noticing you, and you will get feedback for your posts and participation. It just takes time, like anyone joining a new community anywhere.


I see a lot of advice here and think I can add something. One elephant in the room here is that while, yes, we were all once new users with reps of "1" or "101", we were not all new users now or even recently. I have been answering questions in Math.SE since Dec 2012. In that time, the 36th-best score (i.e., the minimum score on the front page) has jumped from about 20K to almost 60K today. That is, the number of high-rep users has exploded by a factor of about 4 in that time. Thus, new users today have a far more daunting challenge ahead of them than many older, high-rep users in being the first to answer new questions in the most heavily-trafficked tags (e.g., , , , and so on).

(BTW I recognize that the OP has been here over a year despite the rep. Still, the frustration is real.)

So what can a new-ish user do? Decide if it is recognition or the joy of solving problems that (s)he wants.

If the former, then you must understand the main trick that some (not all!) high-rep users employ to be the "fastest gun" : posting incomplete answers and editing them along the way until they are complete. Sometimes this is known as providing "hints" that, ten minutes later, become full-blown answers. I admit to having done this a few times myself, even though you might tell that I am mostly hostile to the practice. If you are here to gain a name and some quick rep, this is the way to go, although I warn you that you had better be on the right track all of the time, otherwise, you will just get people quite angry and develop a lousy reputation.

If you just want to solve problems, then...what's the problem? Don't look at anyone's answer - just solve the problem! Maybe when you are done, your answer will be more thorough and complete than anyone else's. I have found this to be the case many many times I have taken this approach - the majority of my most-upvoted posts (and many of my least-upvoted posts, sadly) were the product of my not caring if anyone saw the solution, ever. (Then again, I have a blog where I post my own solutions to problems and it doesn't matter what anyone else has posted.) When I finished, I found that either someone posted something, but nothing close to what I had, or, occasionally, I had the only answer.

One other observation: as Carl Mummert pointed out, high rep is not the same thing as high quality. If you are really interested in someone's answer quality, look at the number of answers that person has provided. People who take the former approach tend to have a low rep/answer ratio, while the latter tend to have a high rep/answer ratio. I do not believe this is a coincidence.

  • 4
    $\begingroup$ "Just solve the problem." Yes, please. I think that closing one's eyes for other answers is the only way to get beyond a certain reputation and experience threshold. I also have had to learn that taking your time and writing a proper answer is often much more rewarding than the quick-and-dirty rep grab. $\endgroup$ – Lord_Farin Jan 26 '15 at 18:37
  • 8
    $\begingroup$ There are many (other) ways to achieve a low rep/answer ratio. One can answer questions that have fallen off the front page --- those who read the website by "newest" instead of by "active" will never see your answers, so won't vote them up. Another is to answer questions in more advanced or specialized topics, which again reduces the potential voting population. $\endgroup$ – Gerry Myerson Jan 27 '15 at 1:26
  • 2
    $\begingroup$ @GerryMyerson: this is true. $\endgroup$ – Ron Gordon Jan 27 '15 at 2:04
  • $\begingroup$ I have a 4.3 average score for my answers. Does that mean that my answers are high quality or low quality? $\endgroup$ – Asaf Karagila Jan 27 '15 at 18:56
  • $\begingroup$ @AsafKaragila: low. Extremely low. Lower than mine, which is 3.6...whoops, I mean high. Wicked high. $\endgroup$ – Ron Gordon Jan 27 '15 at 18:58
  • $\begingroup$ A lot of truth in this post (but as Gerry pointed out - not the whole truth). The real reason I'm commenting here is about the blog. Initially (=about a year ago) you were at least somewhat interested in contributing to our site blog. I fully realize that by now you have settled to your own style, rhythm and level of posting in your own blog. Nevertheless, I would like you to also consider posting at our blog as well. If you look at the material posted so far, you will find out that many levels and styles are ok. $\endgroup$ – Jyrki Lahtonen Feb 15 '15 at 13:37
  • $\begingroup$ @JyrkiLahtonen: I did myself and the M.SE blog no favors by dropping my proposed contribution last year. I simply realized that the scope of my proposal was simply overwhelming to me. As I began, I was compiling a book's worth of material and I used up my limited free time. I will certainly reconsider and my interest is high, but I cannot commit to posting something there for a few weeks. $\endgroup$ – Ron Gordon Feb 15 '15 at 13:51
  • $\begingroup$ You are more than welcome to take your time, @Ron! I have run into the exact same difficulty many times - it is not easy to find a satisfactory way of narrowing down the scope. At this time I am in the process of locating contributors, and I was delighted to hear that you are still interested. $\endgroup$ – Jyrki Lahtonen Feb 15 '15 at 14:45
  • $\begingroup$ @JyrkiLahtonen: I had a thought about the blog this morning. Instead of posting about such a broad topic that may end up being the subject of a book that I may someday write (heh heh), I was thinking about complex techniques in the evaluation of inverse Laplace transforms. There seems to be some demand for this, and I have solved a few nice problems lately in that realm. Anyway, don't give me the keys yet, I'll still be a few weeks yet before I can start writing. I'll let you know when I am ready. (But I am excited to start.) $\endgroup$ – Ron Gordon Mar 4 '15 at 12:40
  • $\begingroup$ Thanks for the update, @Ron. Just ask in the blog chatroom (or here), and we give you a contributor status. $\endgroup$ – Jyrki Lahtonen Mar 4 '15 at 13:35

Echoing and emphasizing several good points made by others:

First, don't think about "reputation points", even though that can be an amusing game.

Second, don't allow "higher status" of others as (necessarily) giving their responses any greater weight than yours. For that matter, often beginners are more enlightened by ideas from just-slightly-more-senior beginners, than those from ineffective-communicator "experts".

Yes, at the same time, one would hope that more-expert people could give better answers and explanations. Unfortunately, this is not quite true, for the obvious slightly-complicated reasons, AND is not so well correlated with "high rep points". The most obvious is that "high rep" arises from many other factors, only weakly correlated with actual expertise, and can be "gamed" by those who desire to.

Another issue is that there is no restriction of "expert judgement" on the voting on answers, etc., hence, no restriction on acquisition of points. It's a sort of popularity contest! ... which we hope is somewhat correlated with correctness, but, really, it's not very strongly correlated (in my observation, on Math SE. On MathOverflow it's a bit better.)

But, yes, I myself do not jump in to every elementary question (unless it has been neglected for quite a while).

Yes, I have for years similarly advocated that very-senior mathematicians (math is my milieu) not compete in refereed journals with junior people whose very survival depends on high-status publications. (But this advice has been called "dangerously radical"... which, at first, confused me, but by now I'm proud of it, I think.)

That is, while paternalism/maternalism can be oppressive, it can also be beneficent/wise (I imagine, in some ideal world), in the sense that "old" people, a.k.a. "people who've already got a secure spot", can be generous with younger, less secure people... rather than endlessly bloodying them in "competition".


I think this is a good thing to bring up. I have read many questions where there appears to be a mathematical knowledge gap between people who post and people answer.

I really don't agree with what you write here:

High-level users should be empowered to answer these harder questions

The level of a user is determined by the number of points he has acquired

I am a member of Cryptography SE and have gained my rep through only asking questions. But according to SE, I still have enough 'experience' to be able to down-vote answers (which I have no idea if they're correct or not...). As other people have stated, reputation does not necessarily indicate the mathematical ability of the user, rather the amount of time dedicated to the site.

Instead of focusing on the amount of reputation that the user has (both the person who is asking and the person who is answering) I think it is better to focus on the answer that is wanted. Even is the user has 15k rep, they may not have knowledge past pre-calculus maths: why should they be encouraged to answer questions which they cannot? (I am deliberately interpreting your question this way). I fully support your idea of encouraging new users to participate in Math SE, so that this does not become an exclusive community for the experienced users.

I had some sort of idea regarding a 'levels' check-box. I was thinking something along the lines of when people write a question, there should be a check box for the 'level' of OP such as High School, Uni Freshman, PhD candidate etc. etc. If a HS student asks then answers will be expected to be given in a format so that most HS students would understand the answer. Additionally, if the person asking is a PhD candidate in Mathematical Analysis then of course we can expect OP to be familiar with undergrad mathematics. The way I see it the check-box system would complement the tags system, and hopefully it will be easier for people asking to get answers and people answering to find questions to answer.

The only problem I see with my suggestion is that the mathematics taught in e.g. High School will vary from country to country (and I don't think we should demand that people state where they live just to get a good answer!)

Please let me know what you think!


You must log in to answer this question.

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