@fpqc: I think Andy's point was that you would be well-served by moderating your toned (even it wasn't, it's true). It doesn't convince anyone of your arguments, and continues to shread whatever goodwill remains toward you in the community. I agree with you on many, though not all, points of this particular case, but your approach is not actually moving the discussion forward.

@Regenbogen: If you are worried about the reaction on meta, you can always email the moderators.

Your suggestions are pretty reasonable (though I would want to hear from more people before supporting them), and probably implementable, but Anton is the one who knows that.

Maybe I'm among the few who do this, but my browser is set to open MO on the questions tab. I'm not sure exactly why I prefer that, probably because I like problems over solutions. In any case, it has the visible advantage that the page is dominated by lower reputation users. Would it be a good idea to make questions the default main view, or just make it a more accessible option?

Yes, and in the one that I can remember, which is surely one of the two you remembered, I apologized for my incorrect and flippant response. This is also not competitive behavior. You can call it rude or dismissive, but it is not competitive. Words have meanings that you can look up in the dictionary. You can't change the meaning of the word competitive. That's not how language works.

I agree with Regenbogen's suggestiong that it would be nice to remove the recently awarded badges chart from the front page. While I understand that some people enjoy the badges, I think that having this on the front page emphasizes the competitive aspect of the site, while giving little in return. I don't have a sense of how controversial this suggestion is.

Another suggestion, which I would guess is controversial, would be to reorganize the User display to be alphabetical (say), rather than ranked by reputation. (I have in mind a typical alphabetized list, where as well as having links to the pages, you have links for each letter.) I am curious to hear if anyone besides me would support such a change. (I also have no idea if it's possible, given the constraints of the underlying software.)

Let me make an argument for these kinds of changes (to a large extent a rehash of what's already been said): I don't think they substantially effect the working of the site; I think they help improve the sense of the site as one for professional mathematicians to participate in, with the sillier aspects such as reputation and badges being downplayed; and they help make the site more neutral --- those who want to focus on reputation, badges, etc. of course still can, but those who are put of by these things won't see them as the dominant aspect of the site. Rather, they'll just see a site with a bunch of mathematicians asking and answering questions; people will have a reputation, but it will look more like what it is (in my understanding) supposed to be --- a merely technical measure of their level of involvement with the site.

In summary, these seem like minor changes that could make some kind of positive difference (although I appreciate that others may think that they are not minor changes, and/or that they will make little difference).

Dear fpqc,

In this discussion, as far as I understand it, "competitive" is standing in for a collection of related behaviours. As I wrote in a previous comment, there is a well-known style of mathematical discourse in which people try to demonstrate their technical knowledge in the brashest way possible, in order to impress, intimidate, or for other reasons. This behaviour may not be, strictly speaking, competitive, but it is certainly part of what I, and I suspect others (including Qiaochu) have been talking about in this thread.

I certainly don't think it is dominant on the site in any way, but my feeling is that even a small amount of such behaviour is very bad for the site. One hope would be that, by making the environment of the site less competitive, even in very superficial, cosmetic ways, such behaviour would be made even more rare than it currently is (by some osmotic effect, I guess; my hope is that it would simply it would be seen as evidently inappropriate by participants on the site).

Dear Emerton,

If there is problematic behavior that one wants to address, then one should be precise with one's labeling, since imprecision has the effect of doing nothing more than damage the user in question's reputation even further.

In any case, if this is the behavior that you would like to address, then it is not relevant to this discussion because the ways that one would deal with competitiveness are not the same ways one would deal with the behavior you describe.

There is an alphabetical user list, at http://mathoverflow.net/users?tab=name. Perhaps someone can suggest how to make it more useful? At present it's hopeless, but if we can think how to fix it it may be a good feature request to have this as the default instead of the reputation sorted list.

@andyputman:

1) is already implemented, but I like 2) and 3). However, I feel like this still will not be sufficient to filter out enough people. It's been noted before that somewhere near half of the people who have the fanatic badge (visited the site for 100 days straight) have extremely low reputations because they lurk instead of posting.

Edit: If the reputation cap is somewhere around 500 as Noah noted, then this should be enough, though.

Dear Noah,

Regarding the point of the user page:

I often use the user page as an interface to make my way to a particular user's page, so that I can look at their latest activity. Now that I know the names of the people I follow, I can normally find them just via a search. But there was a time when I used to look for people manually via the list of users, and an alphabetical order, while not perfect, would have been more helpful than a reputation ranked (and hence highly non-constant!) order.

But even now, when I go to the user page in order to search for a particular user, I think it would look less like a contest if the display was alphabetized rather than ranked by reputation. (Yes, this is a remark that is entirely about superficial appearances, but hopefully this is in the spirit of the discussion!)

I dislike the idea of listing users alphabetically by name, just because it conveys no useful information (if you're looking for a particular user, you should search anyways). Listing by most recent login/activity, however, seems quite reasonable.

The only problem is that the users list will be constantly changing every time somebody makes a post.

My opinion is that there are probably plenty of mature mathematicians out there who might regard all the talk of reputation as somewhat puerile. I think it would be great if mathematicians of all persuasions were given the impression that MathOverflow is a professional place for "real" mathematicians to interact, rather than some new-fangled web 2.0 thing dominated by younger mathematicians and grad students. The use of the internet brings with it some great potential but also a few annoyances (just look at how much time users of MathOverflow have spent discussing and arguing in meta!) and I think that some professional mathematicians may be turned off by some things that might seem quite natural to us younger guys (or maybe more accurately, annoyances that us younger guys are more used to dealing with---e.g. internet arguments).

Very well said. It wouldn't surprise me if the term "reputation" struck people as somewhat tasteless. Perhaps we could just replace the word?

Anton changed the rep count font?

I'm not convinced by many of the suggestions that have come up. Firstly, I think badges are fun and seeing reputation is useful. The reputation sorting of the user page is the most useful to me (the user I'm looking for is almost always on the first or second page). (Aside: sorting by last seen time with a lower bound on rep might be interesting ... somebody would have to request that on meta.SE).

Secondly, and more importantly, I don't see what real benefit would come from hiding or downplaying badges or reputation. If there were no badges and reputation were completely hidden (or didn't even exist), the site would be a lot less active and less fun, but it would still be competitive (equally competitive I think). Any site whose point is largely to post really good questions and really good answers is going to play to peoples' competitive instincts, especially if those posts get voted on. I find it hard to believe that anybody who objects to MO because it's too competitive would find it any more palatable if reputations were only displayed on user pages. If the world is divided into competitive people and non-competitive people, I don't think MO can reasonably try to target both of them.

However, I don't think the world is divided that way. After all, if people who objected to MO being too focused on reputation are clearly themselves too focused on reputation, because MO has lots of awesome content and it seems to me like the users are generally very nice to each other. I think it's a good thing to be in a competition with yourself when you compose a post (i.e. it's good that people try to get as many votes as possible, so long as they're trying to do it by making the post as good as possible or by responding to somebody's question very quickly). I also think it's a good thing to be in competition with others a bit: it makes it way easier to get better. When taken to extremes, these behaviors can be harmful, but I think the level on MO is not nearly as extreme as this thread has inflated it to be. For the most part, I think there are two good options for people who find MO too competitive:

• Get over it. It's only a competition if you make it one. By simply ignoring reputation altogether, you can reap the benefits of other people competing to answer your questions. You're able to deal with the fact that people are competitive about more serious things like salary, number of publications, official title, sexual partners, and alcohol tolerance; this is just like that. If you don't want reputation points, what does it matter?†
• Don't use MO. I know it's nice to be as welcoming as possible, but if somebody has difficulty finding value in MO because of the reputation points, I don't think any superficial (or even deep) change is going to help. It will only make things worse for people who like MO as is. To draw a parallel: we've often discussed that MO is bad for unfocused discussions, and now I'm claiming that it's bad for über-non-competitiveness.

†Okay, I realize my examples aren't quite perfect. Authors' names on the arXiv, for example, don't always display with the number of articles next to them, and people don't wear t-shirts prominently displaying how many shots of tequila they can stomach (though official titles are purely for display). I've already argued that I think displaying reputations next to usernames is actually useful, so I'd really need to be convinced to remove them, but I can see the argument that they are displayed too prominently. To that end, I've overwritten the default css to decrease the weight of the font. What do you think?

Dear Anton,

I prefer the new font for reputation to the old; thanks for making the change.

As for the more general issue of the MO environment, my view is the following: in mathematics there are many hard questions, and a shortage of people who can answer them. We need as many people as we can, with as many different view-points and styles of working as possible.
To this end, the mathematical community should be open to as many different personality types and working styles as possible.

Now MO is not the whole mathematical community, but it is a part of it, and so I am applying the same principles to it. (These are the same principles I would try to apply to a department, to a class, to a fellowship program, ... .) I don't really agree with arguments of the "get over it" type. They apply to private gatherings, but I don't think that they should be applied to something as broad as MO. For better or worse, it is not a private gathering of a few mathematicians; it is more extensive than that. It may already be being funded by external granting agencies (I'm not sure about this), and if not already, surely will be in the future. It may well become a permanent feature of the pure mathematical landscape. For these reason, I think that it's worthwhile to analyze the environment on MO, and to try to optimize it to be as open as possible.

A little off topic from the current question of reputation, I would like to mention an example of what I perceive as extreme competitiveness.

In several occasions I have seen a remark on a selected answer saying something like: "Why was this answer selected? This is clearly not a full answer to your question!" To be clear, I am not talking about cases where the comment was intended to point out that a wrong answer was selected.

Whenever I see that a user answered a question and then left such a comment on the selected answer of another user, I interpret it (perhaps by mistake) as "hey, my answer was better, why did you prefer his over mine?" It also seems to me as disrespectful to the OP's choice to select a certain answer.

I think that such behavior should be discouraged, and I was wondering what do others think about this issue.

[Please note that the particular comments that I have in mind were not left by any of the participants of this discussion. Indeed, I would not make the above remark had I thought that it may be perceived as a personal accusation. So, if anyone in this discussion did leave such a comment, please don't take this personally.]

Thank you for your answer, Prof. Suárez-Alvarez. Funnily, I wasn't referring to the comments you mentioned (which I didn't see). In any case, I realize now that my interpretation may be wrong :)

I have to confess that the longer this discussion goes on, the more confused I get as to what the phrase "competitiveness" is referring to.

I agree. This was a point I made above.

Perhaps it's just the parent in me, but when someone says something like "I don't like MO because it's too competitive." then I wonder if the word "competitive" is just being used as a one-word summary of a much more complicated experience, and that by focussing on that word and assuming that it means what I think it means, I'm missing out on the real problem.

According to Scott, Emerton, and andyputman, the word "competitive" has degenerated to simply meaning "bad". This is misleading for a number of reasons. First of all, competitive behavior is not categorically a bad thing, as Anton noted. Second, it makes it harder to make progress and combat the actual causes. If the problem is behavior that the community thinks is bad, calling it "competitiveness" and trying to come up with solutions to that problem are missing the point. If the problem is some other sort of negative behavior, we should call it what it is and try to find relevant solutions.

@Everyone:

If the problem is actual competitiveness, then, like Andrew, I would like to see evidence of competitive behavior. However, the current examples leave me unconvinced.

Also, as an allegedly competitive person, I can say that I have not cared about gaining reputation since I hit three thousand points and that I was motivated exclusively by the promise of gaining the ability to vote to close. Certainly doing something about these sorts of incentives would be more effective than decreasing the visibility of reputation score.

@Sonia: With respect to your last paragraph, do you mean something like an RSS feed of the best questions of the day for your pre-set "Interesting Tags"?

The fact that users with very good questions can get reputation points quicker than users with very good answers encourages that.

Actually, this doesn't really seem to be the case on MO. I regularly see answers with 20-30 votes, but I rarely see questions with that many votes that aren't community wiki.