Couple thoughts on MathOverflow

MathOverflow was started roughly two months ago.   The administrators there have done a good job with the FAQ, it now does set useful rules.

I quite like the site, but find some features a bit akward, or even having the potential to turn away much needed experts.  I’d thus like to make a few personal remarks, I guess the tone is too general for being a meta thread so I put them up here.

So far MO has about 525 “active users” (folks who got more than one upvote).  Just glancing at the list it seems that 30% are grad students, 50% are junior mathematicians (Postdocs or young faculty) and 20% are senior professors.   Most people one would expect to see (i.e. those already involved in web 2.0 stuff like blogs and wikis)  are there, but not many other researchers seem to have registered. So maybe the site isn’t doing as well as it may first seem.  Here is my list of pros and cons to try and explain that.

Pros are:

– use of real names. Very important to avoid undelicate comments. Usurpation of identity is one issue, perhaps some kind of PGP signing (or email verification by the administrators) would settle this.

– the upvote/downvote system and the badges: allows to know what the community appreciates or not, and to figure out who has done what.

– the tag system, allows to select (resp. skip altogether) things one wishes to read (resp. or not)  e.g. “nt.number-theory” or “soft-question”, altough this should be publicized to those who haven’t registered yet.

Cons are:

– the reputation score:  it’s not helping at all, some folks seem to have asked question to inflate their score (kind of an ego thing) rather than post a genuine question that troubles them.    Better use some broad classes instead, with the precise score only visible for oneself, if one cares at all. For example one could simply use the software’s settings mentionned in the FAQ, I count 10 such classes : less than 15, 15 to 50, 50 to 100, 100 to 200, 200 to 250, 250 to 500, 500 to 2k, 2k to 3k, 3k to 10k and above 10k.

– the navigability: it’s not so easy to locate questions, some kind of wiki mentionning the title and linking to each of them (by arXiv area, say) might be useful.  In 12 months time the site will probably contain more than 20,000 questions at this rate, by then tags and keywords won’t necessarily be enough to search the database.

– some fringe communities could develop (people systematically upvoting one another yet not providing anything substential, just like people refereeing one another yet publishing results of little significance). It might be a problem for new users who don’t know each single 10k+ commenter.

– lack of visibility: how many non web-addicted experts know about the site anyway?

I’m guessing that as long as issues regarding reputation score, skipping particular subjects, and accessing older questions are not made very easily understandable and respectable by experts (especially those who are not so computer friendly)  the site will not see them register.


3 Responses to “Couple thoughts on MathOverflow”

  1. David Brown Says:

    “navigability”: I agree. I’m actually disappointed with how poorly search works. I was hopeful because search worked pretty well with stackoverflow and superuser, but most of the time I find a question by recalling who asked it and when, and then going to that particular user.

    “.. some folks seem to have asked question to inflate their score ..”: I’m a little skeptical of this. Do you have any examples?

    “visibility”: We’re working on this. While we have gotten a number of otherwise invisible-to-the-web people to use MO, most of the time I have to physically sit down with such and such person at a computer and walk them through it, and usually follow up later too. Emailing mathematicians that I know is less effective than I’d like (even if I email them a link to a question that I know they know the answer to..).

    “people refereeing one another yet publishing results of little significance”: maybe I’m naive, but does this really happen in real life?

    “It might be a problem for new users who don’t know each single 10k+ commenter.” I don’t see the connection of this to fringe users.

    Anyway, sorry to respond so critically; I’m quite happy to have stumbled upon your thoughts on MO.

  2. David Brown Says:

    By “real live” in my naive question above I mean in the context of research mathematics.

    One other thing: while we obviously want to grow MO, as a number theorist and geometer it is already an extremely useful research tool to me, and so from that point of view is already a success and does everything I want it to. (Of course its not useful for people in every field yet.)

  3. Thomas Sauvaget Says:

    Hello David, sorry for the late reply, European time zone here.

    It’s great that you agree with the navigability and visibility issues, hopefully this will be improved soon to make MO more attractive.

    As for score inflation it’s been acknowledged (see this meta thread). Problem is that possibly what happened on the site in the early days may have turned away some experts: they may have tought “well, this is a website for grad students (resp. algebraic geometers), not for experts (resp. those in my area), let’s leave”.

    The refereeing remark in math is true (I’m not talking overtly crackpotish papers, simply unimaginative ones or with some overlooked yet fairly obvious gap, after all that’s why journals have reputations and impact factors) and the same could happen on MO: some people could make a living, and eventually reach 10k, by simply upvoting one another. And this could lead to occasionnal problems if they become admins.

    I agree that MO is probably very useful to algebraic geometers already, sorry if I made it sound like it isn’t, I was thinking on a broader level, for expert lambda in area Z. I mean, kudos to you the current admins, obviously you’re doing a great and time-consuming job! Endeavours like those take time to reach their final form.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: