I was trying to explain the current controversy how “welcoming” MO is to a non-mathematician friend. The current best hypothesis for why MO is either attractive or repellent in general is that it has to do with the reputation system. As I was offering reasons for why things are structured as they are, it occurred to me that there are different goals that could be separated in the point system.

Goal 1: Determining good questions and answers. Here, we vote on the questions and answers, and accept at most one answer. This is a crowd-sourced quality metric. Though one may argue about the details, it seems like basically a great idea.

Goal 2: Managing user privileges. When you achieve certain reputation, you get more powers on the site. This allows the editorial work to be distributed among many experts who earn their expert status through their dedication of time and the quality of their mathematical content.

Goal 3: Quantifying users’ credibility. This one speaks for itself.

In my opinion, the above goals are listed in decreasing order of importance; Goal 1>Goal 2>Goal 3. What occurs to me is:

(a) Goals 2 and 3 do not have to be so tightly linked to Goal 1. Goal 1 is an important local imperative, but good questions and answers can be generated by people who are do not have high rep or even participate much, and they frequently are.

(b) Goal 2 is quantified at a second level with a more coarse system of statuses. These could be awarded without the publicly viewable reputation points.

(c) It is debatable whether Goal 3 is worthy or not.

Proposal 1: Make reputation points a hidden feature that only administrators can see. Couple this with an appropriate adjustment to how user privileges are given and announced.

Proposal 2: Allow the possibility of granting user privileges based on verified external credentialing. For example, having an accredited degree in Mathematics (with tiers for BS,MS,PhD), holding an academic position, certain publishing achievements. This could turn out to be quite controversial and require significant work, but it could also be a way of making MO more inviting to more people.

  • 4
    $\begingroup$ It's definitely not the same, but somewhat similar suggestions have been discussed on Meta Stack Exchange. For example: Hide the rep and badge count on a user's info box on answers and other questions linked there. (One point which might be important is whether you suggest this only for MO or for the whole Stack Exchange network.) $\endgroup$ Jan 8, 2021 at 11:12
  • 4
    $\begingroup$ If you don't want criticism on your maths just write a blog or something $\endgroup$
    – trying
    Jan 8, 2021 at 11:27
  • 4
    $\begingroup$ @why In my opinion, your comment misses the point. I don't see MO as a way of showing my maths off to the world, but as a way of asking and answering questions to aid in research. $\endgroup$ Jan 8, 2021 at 11:30
  • 4
    $\begingroup$ @why Are you serious? You read the answer and judge it based on its content. Actually the fact that you ask that question indicates a problem on this site-- rep-based bias. $\endgroup$ Jan 8, 2021 at 11:38
  • 14
    $\begingroup$ We can discuss about this. However, in my opinion, the real issue is not the rep-based bias. Once one gets used to the site, it appears clearly that reputation is just a measure of the involvement in MO, not of the true mathematical value. Apparently, instead, some (potential) users (mostly graduate students, or at least this is the Twitter evidence) complain about lack of empathy, coldness and patronising attitude of some contributors, especially when some questions are downvoted or closed as "not research level". $\endgroup$ Jan 8, 2021 at 13:30
  • 8
    $\begingroup$ Disclaimer: this is some people perception. I am not saying that these claims are true or not, and to what extent. But, if they are true, we should improve our attitude, whereas, if they are false, we should understand why there is this perception after all. $\endgroup$ Jan 8, 2021 at 13:36
  • 7
    $\begingroup$ How does one exist as a math graduate student without experiencing and learning to live with some amount of condescension? Expertise is a real thing. $\endgroup$ Jan 8, 2021 at 13:58
  • 4
    $\begingroup$ I've seen people complain about answers that sounded too sophisticated wrt the question asked. However, in those cases, it turned out that the person complaining didn't actually understand the question. I saw a somewhat senior colleague harassed for giving a good but sophisticated answer to a mathoverflow question by people on twitter. If you walk around thinking people are condescending to you, everything will appear that way. $\endgroup$ Jan 8, 2021 at 14:15
  • 3
    $\begingroup$ Well, probably there is also some kind of "bubble effect": if you usually interact just with people agreeing with you, you will be convinced that you are absolutely right. $\endgroup$ Jan 8, 2021 at 14:37
  • 5
    $\begingroup$ Goals 1, 3 would never have occurred to me, and obviously, if they are indeed real goals, then the reputation system isn't working at all. I always thought it was in place almost exclusively for Goal 4: It's a toy that might make the site more fun and addictive (similar perhaps to a rating on an internet chess site). $\endgroup$ Jan 8, 2021 at 16:04
  • 7
    $\begingroup$ About Proposal 2: I am strongly against making a policy of judging over the user's resumé. Besides the technical and ethical difficulties to do so, I think it would increase an existing bias (a big name sounds much more intimidating than a big rep). Being both a frequent contributor and a frequent editor here, I take the rep into account much more when editing than contributing. On the other hand I'm certainly more intimidated in close-voting or downvoting a post by an identifiable user than one by an anonymous user. $\endgroup$
    – YCor
    Jan 8, 2021 at 16:07
  • 15
    $\begingroup$ My point about Proposal 2 is what I said before: reputation is a measure of involvement to MO, and it is involvement that grants privileges. Personally, I'm against granting privileges to "big names" that nevertheless do not actively contribute to the community. $\endgroup$ Jan 8, 2021 at 16:28
  • 3
    $\begingroup$ It is a bit unclear to me whether you want to hide also scores of the posts or only reputation of users. (As "goal 1" you mention rating system for questions/answers - I'd guess for that score of an answer is more a bit more visible data point than reputation of the poster.) $\endgroup$ Jan 8, 2021 at 18:05
  • 4
    $\begingroup$ These ideas are just thinking out loud, and it's not clear the extent to which they can actually be implemented (if I put my SE Inc-emulating hat on, absolutely none), but still worth discussing. Ideally we have input from people who actually have issue with the system. Else we are just a bunch of armchair sociologists. $\endgroup$
    – David Roberts Mod
    Jan 8, 2021 at 22:30
  • 4
    $\begingroup$ ... ability to interact on their own posts, which is indeed bad. Though this problem is not limited to experts. But let us say for the sake of argument that users with outside credentials should be given the ability to comment. To implement this, we would need a complicated system where someone (moderators?) would need to verify some user-supplied data confirming their real-life identity, and the fact that they have a degree from a reputable academic institution. I’m sorry, but the complexity of such a setup is quite out of proportion to what it can achieve, not to mention that this will ... $\endgroup$ Jan 9, 2021 at 10:03

1 Answer 1


This is a meta-answer, rather than an answer, but I post it here since there is no meta-meta-mathoverflow. I am not hostile to the proposition per se, but the problem is, why change mathoverflow, if we don't like how it works? Why not create a new, independent, math forum, with a completely different form, spirit, and set of features? The web is still mostly free, and there is no difficulty in creating a new site.

Users cannot leave, in any meaningful way, mathoverflow : one cannot remove one's contributions. Changing mathoverflow means forcing people there to live with the changes, whether they like them or not. Their contribution will take a new sense in the new context, which violates in spirit at least what they agreed to when they posted them.

Isn't it better to be a little less repressive, and a little more creative ?

Cours, Camarade, le vieux monde est derrière toi !

  • 4
    $\begingroup$ Obviously, I am not again small technical changes, but big change (analogous to change of constitution, in the political realm) are to be avoided. Each one (like joining the stack exchange network) has been felt as a treason by many, who stopped contributing. In my field, 10 years ago, there were dozens of regular contributors. Now there is just one, David Loeffler. $\endgroup$
    – Joël
    Jan 8, 2021 at 19:46
  • 7
    $\begingroup$ "meta" is an idempotent operator. $\endgroup$ Jan 8, 2021 at 22:11
  • 4
    $\begingroup$ " there is no difficulty in creating a new site." [citation needed] $\endgroup$
    – David Roberts Mod
    Jan 8, 2021 at 22:31
  • 3
    $\begingroup$ Technically creating the site, and opening it on the internet, requires a lot of work, but is not, nowadays, difficult. I suppose you mean that attracting people to it to make it live is difficult. In that case, if the site has some feature that distinguish it clearly from MO, I am sure that a letter published in Notices of the AMS and a few other similar publications, promoting the new site and explaining its interest, will attract a lot of users. Speaking for myself, I would certainly (if I am allowed to) go on the site out curiosity, and make it my new home if I like it. $\endgroup$
    – Joël
    Jan 8, 2021 at 22:57
  • 1
    $\begingroup$ @Joël I'm willing to bet that any kind of change to the mechanics of the site, no matter how hard technically, is easier than rebuilding the community. Of course one might want to jump ship for unrelated reasons (e.g. for disagreements with SE), but that's a whole other kettle of fish. $\endgroup$ Jan 17, 2021 at 21:19

You must log in to answer this question.

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