Vanilla 1.1.9 is a product of Lussumo. More Information: Documentation, Community Support.
I've noticed that I am spending less and less time on mathoverflow. Somehow, I'm not getting as much enjoyment from the site as I used to. I talked to a few other people who seem to agree. I think it wouldn't be a waste of space to see if other people are noticing this, and (more importantly) whether it can be helped.
I think this is important for the following reason: we all know that this site is by far the best place to get some mathematical questions answered. So, when I'm stuck on some trivial lemma in some paper, I'll of course come here. However, this is not as useful to the well-being of the site as me answering questions. Now that I'm reading this site only rarely, it's much less likely that I'll find a nice question that I want to answer. If too many people start acting like me, the quality of the answers will decrease. The worrisome thing is that this begins a vicious cycle: as the quality of the answer decreases, fewer people start browsing the site to read them, they become less likely to give their own answers, and so on.
A few disclaimers:
1) Not that this needs saying, but this is still a wonderful site, and it's amazing that there is now a place to get quick answers to questions that would give us so much pain before. To the creators: thank you very much for spending so much time on this.
2) A big factor in this might be just that this site is not so young any more, so it's hard to keep the up excitement of this amazing new thing appearing. I don't know what to do specifically about this, but I think this is not a reason to give up. Also, this might not be so hopeless: so many people still don't know about the site, they can be made excited about it just as we were when we first saw it.
3) I have a habit of wording my opinions rather strongly and directly, but I certainly don't mean to insult anyone. I think it's amazing how well-intentioned pretty much everyone on this site is, and I appreciate all of you guys. Also, there might be a bit of hyperbole involved - I don't really think the site is in danger of imminent destruction, but I think this is something good to think about, and hyperbole is good to make one's point clear.
The following is a guess/opinion on one reason this might be. I'm posting it separately so as not to make the thread limited to discussing this one opinion. I also expressed a version of this opinion in an old thread, but it didn't get any response.
Basically, I think there is a big downside to the policy of keeping mathoverflow strictly on topic (of research questions). I really liked the way the old version of the FAQ said (and the StackOverflow FAQ still says) that pretty much all questions are allowed, as long as they are "of interest to research of mathematicians". I really enjoyed questions that are either closed now or would likely be quicly closed if they were new: "Cocktail party math", "Better to walk or run in the rain", "Favorite books", etc., etc.
Apart from my personal warm and fuzzy feelings towards such questions, I think there is a serious advantage from them. They make more people spend time on the site, which makes them more likely to answer questions. I think this site should be structured like a magazine (specifically, Science is structured like this): there should be serious articles, but then also a few more general-audience ones in front and back. There might even be a couple of comic strips, all so that more people pick it up.
So, I'd recommend not only to allow borderline questions, but to try to specifically nurture questions that are interesting and amusing for as many mathematicians as possible. I'm not sure how to do it, but first I let's see if you agree with the general principle...
I see a couple of arguments against this:
1) Sometime in the winter, the community's consensus clearly shifted in the opposite direction. To be honest, I do not quite understand why people feel so strongly about a few questions they don't like popping up on the front page (maybe it's more annoying if you use RSS? Maybe it's 1.5 below?). However, I imagine this opinion will be unpopular, if only because the people who share it are less likely to read this (they might not find the meta that exciting).
1.5) We might not like to have "too many" questions, that would make the site difficult to browse. I don't really know what to do about this, but this is a danger in any case if mathoverflow becomes "popular enough". StackOverflow people must know a lot about how bad this is, and whether it's possible to deal with.
2) There is a danger of attracting the "wrong" people here or "too many" of them. I don't think the danger of this is too great (it's not that much fun to troll mathematicians, and I think it was always clear we don't like homework questions), but I might be wrong.
3) The best argument I see against my idea is that we might not want the answers from the people who are attracted by soft questions. The quality of the answers from "random" people will likely be worse than the answers from Terry Tao or (pick your favorite mathematicians posting here). However,
-- We can continue to mercilessly moderate bad answers (or moderate them more mercilessly).
-- Even very smart people sometimes want to spend some time mindlessly, and mathoverflow used to be great for that. Some of the "random" people attracted by this might actually be very good.
-- I think the benefits outweight the risks. The main benefit, apart from the above, is that we'll get answers to the questions that Terry Tao might find boring, and some of them might be good.
-- This will be very good for the education of all of those people who dare to answer questions. Even if their answers aren't very good at first, they might improve.
I have to admit to being very dubious. Of course it's natural that any individual users interest may show an initial burst of interest, followed by a gradual decrease. This isn't anything to worry about (indeed, if this didn't happen we might have to seriously worry about wasting so many people's time :-).
Looking at the statistics, it seems the site as a whole is very healthy, on the other hand.
Are there stats on how much people return to the site?
@Mariano: Google analytics has a "visitor loyalty" section. I've pasted the most relevant table below. The columns are Count of visits from this visitor including current, Visits that were the visitor's nth visit, and Percentage of all visits:
1 times 59,191.00 30.12%
2 times 9,128.00 4.65%
3 times 4,804.00 2.44%
4 times 3,261.00 1.66%
5 times 2,587.00 1.32%
6 times 2,109.00 1.07%
7 times 1,846.00 0.94%
8 times 1,653.00 0.84%
9-14 times 7,531.00 3.83%
15-25 times 9,090.00 4.63%
26-50 times 13,507.00 6.87%
51-100 times 16,923.00 8.61%
101-200 times 20,078.00 10.22%
201+ times 44,793.00 22.80%
I'm pretty sure that by most other measures, site traffic is quite healthy. Number of visitors, pageviews per visit, and duration of visit have been pretty steady. Number of questions and answers seems pretty steady too, but I haven't looked into it too carefully. I've posted a fresh database dump, so you can have a look for yourself.
All that said, I definitely don't spend as much time on MO as I used to. But when I do go there, it's awesome. I think a low density of "fun" questions is okay, but I would rather MO be a small site which everybody knows is specifically for precise technical questions and answers than a big site whose purpose is unclear, or whose purpose is to be fun. The fun is supposed to be a side effect of actually getting things done. Sure, everybody likes to spend some mindless time on the internet, but I think that there's a huge benefit to separating people who are online to goof off mindlessly from people who are online to solve problems (which can also be a form of goofing off).
In this precise moment the top 5 questions are "Fundamental group of a formal punctured neighborhood", "How can I understand the “groupoid” quotient of a group action as some sort of “product”?", "Failure of Fin. Presented and Fin. Generated Modules to be Abelian Categories? ", "Relationship between geodesics and curvature lines on surfaces?" and "Is there a good account of D-affinity and localization theorem for partial flag varieties?". From the titles they all seem to have nontrivial mathematical content.
@Dror: Can you draw a picture of what MO looked like when it was more interesting? I'm interested because it's not totally clear to me what you're comparing to, and I'm also wondering if your perception is area specific.
Yeah, I don't know what you're talking about, Dror.
(Edit: it looks like fdorais made pretty much the same point while I was writing this)
@Dror: Could you clarify a bit? It seems like you're saying that you're generally not interested in asking (or reading/answering?) technical questions. On the other hand, you're saying that most of the questions you see are not really technical, and are uninteresting to you for other reasons (or perhaps because they are not technical?). From this, I would conclude that MO is not interesting or fun to you, but I don't understand why it has become less fun. What used to be different about it? You say that the questions interest you less and less. Could you explain what sorts of questions you found interesting before?
And what are you guys talking about? MO has become much more fun within the last 24 hours. Hadn't you heard?
Yes, this is a rumour I started to trick people into thinking that MO was still fun.
Harry, that must be a lie! If you really had started it, then Pete would never have heard it...
All hail the new king!
Seriously, I'm impressed Pete. Good job!
Now, back to the topic... How did you manage to keep it fun all this time?
Yay verily and so it was spoken. On the 27th of April, a new King shall ascend to the throne, and there will be much merrymaking.
Just a temporary blip. I still have a fund of questions that only David Speyer can answer (I started collecting them when he was an extremely high-rep user with no silver badges. He got one without needing my help so I still have the list of questions).
For the true, faithful MOwers, there is only one King.
Are MOwers a threat to MathOverFlowers?
1 to 20 of 20