[Disclaimer. I'm making this feature request not because I want it nor because I think it's a good idea, but because I heard it so often that there is need to discuss it. Please don't shoot the messenger!]

Let users opt out of reputation

Indeed, some users have resorted to various schemes to avoid accruing reputation on MO such as answering in comments, answering in community wiki mode, discarding points through bounties, etc.

There are a few obvious caveats that pop to mind immediately:

  • Since user reputation is tied to their ability to participate in the site, this option mostly makes sense after a certain minimum threshold is reached.
  • This should not stop others from voting on posts from users who have opted out since that serves other purposes.
  • Since it is undesirable to have users opt out and back in, the choice to opt out would probably have to be irreversible.

Could such an opt-out work on MO?

  • 2
    $\begingroup$ I would rather have BCnrd giving visible answers, should he be interested. But there are enough people for whom this would probably not be enough, i.e. R. Borcherds. Which is too bad for me. That is, it may not be fresh in his mind anymore, but he knows those integral lattices. $\endgroup$ – Will Jagy Jun 28 '13 at 23:57
  • 5
    $\begingroup$ If there is some conservation law about Stack Exchange reputation that would be violated by this proposal, I volunteer to receive all of the leftover rep! $\endgroup$ – Manishearth Jun 29 '13 at 0:01
  • 6
    $\begingroup$ Was the point of BCnrd really to opt out of reputation? I heard it said (on the old meta) the point was to self-limit the length/level of detail of contributions as a "protection" to spend too much time with MO. $\endgroup$ – user9072 Jun 29 '13 at 0:14
  • 1
    $\begingroup$ Yes, you're right, @quid, that was one of the main reasons that BCnrd reported for answering in cmts. $\endgroup$ – François G. Dorais Jun 29 '13 at 0:16
  • 2
    $\begingroup$ Brian should develop more self control. quid can give him some lessons. :) $\endgroup$ – Bill Johnson Jun 29 '13 at 0:39
  • 1
    $\begingroup$ I removed the two examples since they were more distracting than anything else. $\endgroup$ – François G. Dorais Jun 29 '13 at 1:05
  • 14
    $\begingroup$ Although I see some sense in requiring reputation for certain site privileges, I don't understand what purpose is served by making reputation public information. If it were not public, or at least less prominent, fewer (potential) users might be bothered about it. $\endgroup$ – Mark Meckes Jun 29 '13 at 3:18
  • 2
    $\begingroup$ My reputation is the perfect example of a telescoping sum. $\endgroup$ – Fred Kline Jun 29 '13 at 13:12
  • 2
    $\begingroup$ @MarkMeckes Extrinsic motivation. People participate because of it. On Quora, credits are hidden (and used like an "economy"). The motivation to participate there is somewhat less. $\endgroup$ – Manishearth Jun 29 '13 at 16:16
  • 2
    $\begingroup$ @Manishearth: Actually, I do understand that. My point is really that I think it's not a compelling enough factor to outweigh the downside. I know of several well-known mathematicians who don't participate in MO precisely because of reputation and/or badges. $\endgroup$ – Mark Meckes Jun 29 '13 at 16:31
  • 1
    $\begingroup$ @MarkMeckes Oh, I understand your point too -- and yes, a "hidden" rep would be a nice option. I don't see it being implemented but it's worth a try. $\endgroup$ – Manishearth Jun 29 '13 at 18:24
  • $\begingroup$ @Mark: reputation is also a very rough measure of how much you should trust someone's answers (especially if they're using a pseudonym). $\endgroup$ – Qiaochu Yuan Jul 2 '13 at 19:47
  • 1
    $\begingroup$ @Qiaochu: Yes, very rough, and to my mind not worth losing the participation of excellent mathematicians who find the prominent display of reputation and badges to be too ridiculous, offensive, or otherwise off-putting to put up with. $\endgroup$ – Mark Meckes Jul 12 '13 at 14:41

I should admit I don't completely understand why reputation seriously bothers some people. I think (and I think others share this view) that reputation is just there to make the system work (analogous to funding money for research) and provide some feed back, but other than it does not really mean much (though I have seen it causing some participation addiction). I personally found statistics like average total votes for answers more interesting.

I think if we have a better understanding of the reason some dislike reputation, we may find solutions that work for them without altering the system.

First, let me point out that even if there was a reputation opt-out it would still be possible for others to calculate how much reputation you have (and it is not really difficult using the SE API).

If the issue is that seeing your or other users' reputation bothers or distracts you then you can use userscripts to hide them.

If the issue is that you don't want to have the privileges that come with reputation then you can donate your reputation on questions you like. This is particularly fun with the new bounty system which allows you to select the reason for your bounty. For example, you can put a bounty to award it to some already existing extraordinary answer.


After reading this discussion over tea I think I partly understand the motivation.

"I answer questions under random user names so people actually have to engage with the mathematical content rather than simply vote based on name recognition."

However, it seems to me it is very difficult to have the following two at the same time:

  1. retain the ownership of posts,
  2. not have any publicly accessible history of behavior.

I think the issue is not really reputation, at least not in this case, but not having any public history so the posts are judged purely based on their content. Users who don't care about 1 already can post completely anonymously.

I think we can request a feature where the information about the owner of a post is not publicly accessible (but it will be to moderators and maybe also to trusted users). I think the reason has merits and such a feature would be interesting. I don't know if SE would find it a reasonable feature-request, but it might worth discussing/asking for such a feature.

A negative side of posting anonymously would be lack of meaningful interpersonal interaction, I mean think about a math conference where people are wearing masks to remain anonymous to keep the focus on math. I don't think I would find it a pleasant environment.

  • $\begingroup$ Unless you're rep-capping every day, it anyway takes days to reach 500 rep. I don't see an issue with 500-rep-bountying one interesting question every few days :) $\endgroup$ – Manishearth Jun 29 '13 at 9:50
  • 6
    $\begingroup$ About anon posting: It exists on Quora, an it permits one to say wrong things in a convincing tone, among other things. On Physics.SE when I'm not sure of something I mention that there's a significant chance my answer may be wrong. On Quora it sometimes is feels easier to just post anonymously (and I've seen that many anon answers are wrong). IMHO anon posting may lead to a greater influx of wrong answers. But I'm not sure of this. $\endgroup$ – Manishearth Jun 29 '13 at 9:52
  • $\begingroup$ It seems to me that an obvious solution to the conundrum you raise is what I suggested above: reputation is recorded by the system internally, but doesn't appear in any publicly viewable way. $\endgroup$ – Mark Meckes Jun 29 '13 at 13:07
  • 7
    $\begingroup$ My personal preference: participate normally, but ignore your reputation (even if it appears for others to see). This requires no change in the system, and anyone can adopt it immediately. The advantage for others is that I don't post rants about those who downvote me. $\endgroup$ – Gerald Edgar Jun 29 '13 at 13:13
  • $\begingroup$ @Manishearth: I don't know Quora works, but to avoid potential confusion, I let me stress that I think there is a huge difference between posting something (isoloated) with an anonymous throw-away account and having a (one!) pseodonymous account on MO for an extended period of time. By contrast I'd say if pseodonymous one rather needs to excercise more care to "make sense" on site, else everything will be ignored. There is nothing but what is on the site to have credibility. For the user who's actions started discussion, I think one must agree that maths is beyond any reproach. $\endgroup$ – user9072 Jun 29 '13 at 15:23
  • 2
    $\begingroup$ @quid This has to do with a "post anonymously" option per-post. I've found that the inclusion of such an option leads to less reliable answers. Anonymous psuedonyms still post reliable things, because the posts are still linked to their anon profile (and thus they have a reputation to keep up on the site) $\endgroup$ – Manishearth Jun 29 '13 at 16:03
  • $\begingroup$ @Manishearth: thank you for this clarification. In this form this makes a lot of sense to me. $\endgroup$ – user9072 Jun 29 '13 at 16:07
  • $\begingroup$ @quid Of course, the people on MO seem to be a very nice and trustworthy bunch, and there's a high chance that the post-this-as-anon option, if present, wouldn't be abused. :) $\endgroup$ – Manishearth Jun 29 '13 at 16:09
  • 1
    $\begingroup$ @MarkMeckes, not showing reputation (except in profile pages) seems interesting, you should post it as a new feature-request post on meta. $\endgroup$ – Kaveh Jun 29 '13 at 19:16
  • $\begingroup$ @Manishearth, to deal with the issue of increased number of wrong answers, anonymous posting can be restricted to users who have high reputation! :) $\endgroup$ – Kaveh Jun 29 '13 at 19:17
  • $\begingroup$ @Kaveh Hahaha that would get confusing... :P $\endgroup$ – Manishearth Jun 29 '13 at 19:23
  • $\begingroup$ @Kaveh: How is that different than the current one? $\endgroup$ – François G. Dorais Jun 29 '13 at 19:36
  • $\begingroup$ @François, there is no reputation out-out, but users can opt-out from their reputation being displayed except on their profiles to other users (or even do this for all users on the site). Others can check their reputation from their profile page but reputation is not displayed all over the site any more. I think this is roughly what Mike is suggesting. $\endgroup$ – Kaveh Jun 29 '13 at 19:55
  • 1
    $\begingroup$ @Kaveh: If by "Mike" you mean me, what I'm really suggesting (knowing this suggestion will go nowhere) is that nobody's reputation should ever be visible to anyone anywhere (with the possible exception of moderators). $\endgroup$ – Mark Meckes Jun 30 '13 at 16:02
  • $\begingroup$ Here is link to the same query for MO (which is now also available in SEDE): data.stackexchange.com/mathoverflow/query/106250/… $\endgroup$ – Martin Sleziak Jul 30 '19 at 4:47

I have a different suggestion: register as suitcase381, onionbag, probterob, or some other anonymous name, which you can change each year (if not each month). If you adopt an alter ego on MO then I don't see how reputation points would remain a concern. Other people have done that successfully.

  • 1
    $\begingroup$ That's OK as long as the user identifies himself or herself in the profile. $\endgroup$ – Bill Johnson Jun 29 '13 at 0:40
  • 15
    $\begingroup$ Why does a user have to identify himself or herself in the profile? There's no identifying information for Frictionless Jellyfish in its profile, for example. $\endgroup$ – KConrad Jun 29 '13 at 1:19
  • 3
    $\begingroup$ An issue I could imaginge with having several accounts is that it can be difficult to keep track of 'late' comments, make edits to ones own posts later and other things like casting multiple inadvertent votes on the same things. This seems like a possibility but feels quite limiting and/or inconvenient in some ways. If one where to keep several accounts active to partially sidestep the problems I think it might be good to somehow make this clear at least to mods to avoid the risk of the perception of some fraud going on. $\endgroup$ – user9072 Jun 29 '13 at 1:47
  • 14
    $\begingroup$ Giving real names or at least identifying information is encouraged on MO. IMO it should be required. $\endgroup$ – Bill Johnson Jun 29 '13 at 1:47
  • $\begingroup$ I should perhaps add to my preceding comment that personally I have no interest to opt out of the reputation system, as I find the point-system serves a practical purpose, and also but this is still more personal in the end I find it somehow fun to get ever more points and badges (as a by product, to optimize ones behavior for it though would seem a bad idea to me as well as assigning any actual relevance outside the site to them). $\endgroup$ – user9072 Jun 29 '13 at 2:19
  • 3
    $\begingroup$ One could have a sequence of user names, e.g $BC1, BC2, \ldots, BC n$ (to borrow a real world example), and just move to the next one if too much rep accumulates $\endgroup$ – David Roberts Jun 29 '13 at 3:40
  • 4
    $\begingroup$ Not using your real name can be quite an inconvience for someone wishing to cite your question/answer. $\endgroup$ – Fred Kline Jun 29 '13 at 13:14
  • 3
    $\begingroup$ @Fred: This has actually happened to me. Recently, I posted a solution to a question on another SE site without realizing that the question was effectively a small open problem (a related open problem having been stated in a fairly recent paper of Yuval Peres). The OP of that question has submitted a paper for publication using my answer, listing me as a coauthor under my SE username! While I take no extreme measures to remain anonymous (and, though, some may not realize it), there are, indeed, real-life circumstances that do not admit the ready use of one's name on this or other SE sites. $\endgroup$ – cardinal Jun 29 '13 at 15:02
  • 1
    $\begingroup$ @FredKline: to echo cardinal, I do not see the inconvenience this causes; not a coauthor but also quid is acknowledged in a paper how is this more or less convenient than my real name, actually quid is rather shorter than most real names (including mine). $\endgroup$ – user9072 Jun 29 '13 at 15:09
  • $\begingroup$ I was cited by OEIS as Rudy Toody on Mathematica.SE and was asked by Neil J. A. Sloane to change to my real name in case of other citations. $\endgroup$ – Fred Kline Jun 29 '13 at 15:14
  • 2
    $\begingroup$ @quid: I wouldn't refer to an internet pseudonym in a paper -- simply because such pseudonym may change and may also be used by anyone. $\endgroup$ – Stefan Kohl Jun 29 '13 at 15:55
  • 1
    $\begingroup$ @StefanKohl: just real names are also not unique identifiers, they can also change. If one wants to quote content one will anyway need to do it not just by the displayname, but referring to the relevant post. In addition there is also the more stable user number for MO. That being said, I can see reasons why somebody would not want to quote an internet pseudonym in a paper; my reponse to the inquiry how to acknowledge me was with pseodonym or just not at all (the information I gave was anyway available). $\endgroup$ – user9072 Jun 29 '13 at 16:04

There's a third option, which used to be used by ChrisF on Meta Stack Overflow for quite a while: He would post all of his answers as Community Wiki, thereby rarely gaining any reputation.

This could be used as a viable workaround by those who do not wish to gain much reputation. Getting rid of all current rep can only be done via bounty. (or mass deleting answers, but that's not allowed and ought to be undone by mods)

Note that this is an abuse of the CW system -- it's not supposed to be used that way. But I doubt that it's anything serious and it won't make a difference if your users do this.

  • $\begingroup$ We used to discourage this plan since new answers to cw question automatically become cw, which is unfair to the author of the answer. However, since (non-moderator) users can no longer make cw questions, this is not a problem anymore! $\endgroup$ – François G. Dorais Jun 28 '13 at 23:56
  • 1
    $\begingroup$ How do you do a footnote? $\endgroup$ – Will Jagy Jun 29 '13 at 0:00
  • 3
    $\begingroup$ @WillJagy meta.mathoverflow.net/revisions/… . Use superscripts. $\endgroup$ – Manishearth Jun 29 '13 at 0:05
  • $\begingroup$ There are some MO users who proceed like this, Jason Starr is a prominent example. $\endgroup$ – user9072 Jun 29 '13 at 0:16
  • $\begingroup$ Manishearth: There is no +15 for accepted cw answers. Also, there is no +2 for the user who accepts a cw answer. $\endgroup$ – Jonas Meyer Jun 29 '13 at 2:59
  • $\begingroup$ @JonasMeyer K, thanks, edited :) $\endgroup$ – Manishearth Jun 29 '13 at 3:00

I think the real problem is that "reputation points" on this site are not exactly about "reputation" among professional mathematicians. People discard they "reputation points" not to drop the "reputation", but to show they disagreement with the system measuring "reputation".

I have the following proposal --- let us rename "reputation points" to "administration points" and almost everyone should be happy (except people that are only interested in collecting "reputation points"; but who cares?).


You must log in to answer this question.

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