Vanilla 1.1.9 is a product of Lussumo. More Information: Documentation, Community Support.
This is supposed to be a continuation of "elitism for elitism's sake," but I thought it might be good to start a fresh thread.
So we had some internal moderator discussion, and we all agreed on the following (I think important) point:
don't tell people their question is inappropriate unless you have closed or voted to close the question. Even then, don't comment unless you have a particular point in the FAQ in mind.
If you have suggestions for improving a question, by all means share them with the author. If you're confused by a question, by all means, leave a comment asking for clarification. I'm not suggesting people shouldn't leave comments.
But the Unwelcome Wagon has to stop. It's not good for the site, and it's a waste of time. When the person asking the question is a legit user who doesn't understand the site yet, jumping down their throat doesn't achieve anything, and if the person is a troll then annoyed comments are the sort of thing they love. If you see a post you don't think is appropriate and which you have no ideas for improving, there are three things you can and should do: vote it down, flag it for moderator attention (or as span/offensive if you think it is), and vote to close it if you have enough rep. You should have no reservations about doing things without leaving a comment.
The point of these mechanisms is to make decisions about posts as a community. No individual member has the responsibility or the right to decide when questions are appropriate, especially when the question is borderline. Of course, moderators have the technical ability to make these decisions unilaterally, but even that happens pretty rarely. I almost never close posts that don't already have several downvotes or votes to close, and that is how things should be.
From this point forward I consider comments that just put down a post to be offensive, and will flag them them any time I see them; I encourage other users to do the same.
With the proviso -- if there hasn't already been a helpful, constructive comment, referring to some part of the FAQ, explaining why the post is being closed, and you are casting the final vote to close, then it is your responsibility to write such a helpful, constructive comment. It could even be friendly and/or apologetic.
(I admit to failing to do this in the distant past :-)
I'm in no way in an opposing camp, but I fear I might have given the impression that I am. I think I feel more strongly than Ben or Scott that you should leave a comment if you vote to close a question (say), regardless of whether you're casting the final vote (spelled out in more detail in another post), but I absolutely agree with them that leaving no comment is better than leaving a non-constructive comment. If nobody is going to get anything positive out of your comment, don't leave it. "Not appropriate for MO" isn't ever useful to anybody.
One special case I'd like to draw attention to is non-salvageable bad questions. Just because the question is going to be closed with no chance of reopening doesn't mean that you can't say something useful to help the user understand why the question doesn't belong and where it might fit in better. A few of our top users have asked first questions that have been inappropriate for MO. What a shame it would have been if we'd just sent them away instead of explaining what MO is best for and why. However, I realize this takes a bit of energy; if you're not willing to put in that bit of energy, please don't leave any comment at all.
Maybe the balance between the idea of helping users and the need to close some questions could be reached by writing a new page, something like
Why is my question closed/voted to be closed?
If you were sent to this page, it means that somebody voted to close your question or asked moderators to close your question (by flagging). First: relax! This happens even to the best of us.
Now, the person who sent you here should have given you one of the reasons below, so please take a careful look:
What to do now? Well, if you had problems writing a clear text or were misunderstood, the best solution is to correct the question:
@Ilya: I feel like the "what not to ask" section of the FAQ already does that to a large extent. It even has a slogan in bold for each reason (and even an anchor for each reason, see http://mathoverflow.net/faq#meta). But the comment "Please read the FAQ" is pretty much just as bad as "Inappropriate for MO" (and about as common). I really deeply believe that 30 seconds of personal attention is vastly better at reaching somebody than the most heartfelt and well-intentioned canned text (which hardly ever actually gets read). To be sure, the canned text is a useful reference, but comments like "too broad. see http://mathoverflow.net/cannedtext#Item_1" just don't work in my experience.
It seems to me that the focus of the above policy should be "don't comment that a post is inappropriate without giving specific guidance as to what is wrong with it."
The way Ben puts it, "don't tell people their question is inappropriate unless you have closed or voted to close the question", seems like it is missing the main point. It is not failing to click the close link that it is harmful, it is commenting without useful criticism.
@Anton, I took a look at the last closed question — to which, it turns out, you left the comment
@dassouki: your question was closed because it looks like a homework problem that you haven't put much effort into. I hope you can appreciate that we have to discourage such questions. If you want the question reopened, you should google for statistical significance and small sample techniques, and edit your question to explain what approach(es) you've tried and why you're not satisfied with them. – Anton Geraschenko 8 hours ago
I agree this is better than any boilerplate text, but in my experience 75% questions are closed with less explanation; and the combination of short comment
Please take a look at [boilerplate#homework], and consider googling for statistical significance and small sample techniques first.
and boilerplate text
homework Your question was voted to be closed because it looks like a homework problem that you haven't put much effort into. I hope you can appreciate that we have to discourage such questions. If you want the question reopened, you should google for the topic of your question, and edit your question to explain what approach(es) you've tried and why you're not satisfied with them.
would be 70% as good while the comment is much easier to type so that it can be done by more people!
Or, for example, A good place where to learn about derived functors.
It wasn't closed, but it's still a typical example of question with not enough background.
The commenters to the question, for example, didn't say that the question should be edited to provide more background, rather then the comments should be added. This is not surprising: right now we're essentially having many people typing their own suggestions to each question, and it's easy to miss a thing or two every time.
What's wrong with giving a boilerplate link to questions like that?
@Ilya: There's nothing wrong with linking to boilerplate text. The anchors http://mathoverflow.net/faq#homework (for your first post) and http://mathoverflow.net/howtoask#specific (for your second post) are there exactly for that purpose. What I was getting at is that people like to interact with other people and don't like the feeling of talking to a robot. "Please look at http://mathoverflow.net/faq#homework and consider googling for statistical significance and small sample techniques first" is okay, though it still feels dismissive to me, but just "http://mathoverflow.net/faq#homework" or "Homework. Read the FAQ." are needlessly dehumanizing.
David- part of the point was trying to make is that "This question is inappropriate for MO" is a statement you should be very loath to make. If you're making in the context of explaining why a question is closed, that's one thing, but otherwise you're just being incivil.
But your focus is bizarre. Someone who just makes that statement, unadorned, and votes to close is being harmful. They are making a harsh statement without backing it up. Someone who leaves a comment like Anton's example above, but then doesn't vote to close, is much better.
@David- I feel like this is an a implies b = b implies a problem here. I'm saying I think it should be a new norm on MO that you don't tell people their questions are inappropriate, even if you think they are. Now, obviously there needs to be exception to this, which is that when questions get closed, it needs to be explained that they were closed because they were inappropriate, which necessitates at least an indirect suggestion that they aren't. That's not the same as saying that people have carte blanche to say what they want when they are closing a question.
I think what David is getting at (or at least what I'm getting at) is that casting the final vote to close a question is not a special situation. "Inappropriate for MO" isn't ever a useful complete comment, so nobody should ever leave it. If your question is closed, it's clear that at least some people thought it was an inappropriate question; the person who casts the last vote doesn't have to say so. But it would be a lamentable situation if five people familiar enough with MO to be able to vote to close questions all voted to close, and not one of them bothered to leave a note about the (presumably well thought-out) reason they voted to close.
Douglas-
Comments that add to readers' understanding of questions are great. If something is a definition check that takes over a minute (!) then posting an answer how to do it probably wouldn't be out of place. On the other hand, if you think a question is unsalvageable, the thing to do is flag it and get it closed, so people don't have to click through at all.
What doesn't help anyone, though, is writing things like "This question is way below the level of interest to mathematicians," even if you think it's true. Remember, everything comes out meaner than you intend it when you're on the internet talking to people you don't know. It's worth going out of your way to be polite on MO, if only because it will improve the regulars' opinion of you.
All the comments on that look fine to me. If I were Scott I would probably have deleted the last comment just because Anton had just said the same thing.
I would say they are "appropriate" because they were directing the person, and explaining to them why the question had been closed.
I do hope you will stay on MO Douglas. I didn't see the comments in question, so I'm not in a position to judge, but in general I do think you are a valuable member of the community.
Douglas-
I do apologize for the way I wrote the message above. I especially should have been more careful about my use of the pronoun "you." I honestly did intend the "you"s in the last paragraph to refer to a generic person, not to you specifically. I recognize that was probably not how it sounded when I had used an example from one of your comments, but it is really how I intended it to be read (a perfect illustration of how everything one sends sounds meaner on the internet than it really was intended). I certainly never meant to imply that I don't respect your contributions to the website, and I certainly wouldn't be pleased if you quit the site. Rather, I was just trying to point out that if any user were to get a reputation for making rude or mean comments, it would adversely affect how a lot of people on the site viewed them, and that it is easy to make comments that come off that way without realizing one is doing so (since it is so easy for comments we make on the internet to come off differently than we intend them).
I just want to emphasize as a general point here: homework questions are not a big problem on the site at the moment. That is a really important context for this conversation. At the start of MO, we were really worried about being deluged by HW questions, but that deluge never really came. The average number of closed questions per day over the last month is under 5.
So at current trends, it would be essentially physically impossible for them to take up a significant portion of front-page real estate or page views. Most users probably wouldn't notice if we simply stopped closing questions. I just thought I would mention that.
I think most of us agree with Douglas' comment that "being helpful is not the same as pretending everything is wonderful". The point is that comments shouldn't be unnecessarily mean-spirited, especially if they're also not being helpful.
I'm pretty surprised that Douglas' comment was deleted because it was repeatedly flagged. I don't remember exactly what it said, but I do remember that I read it and didn't think anything was wrong with it.
I would encourage people to be fairly reserved in their flagging of comments. Please only flag comments that are both lacking any real content and are rude. Though I don't care for them, I don't like the idea of comments being deleted just because they lack content or just because they are a bit rude (but have some useful content). Not even moderators have access to deleted comments, which makes it very difficult for a moderator to do his job if any comments in the thread have been deleted (btw, please vote up this request if you can). Moderators should be especially careful, since a comment being flagged by a moderator results in immediate deletion. Hostile comments create a hostile environment, but overzealously flagging comments also creates a hostile environment, and the latter situation is much more difficult to fix because the trail of evidence disappears.
Let me also try to restate what I am saying:
The internet has a bad reputation, and preservedly. A lot of people say and do a lot of rude things on it. MathOverflow is trying not to be like some other internet communities I can think of. Now, part of doing that means not letting the site be taken over by undergrads and high school students trying to get answers to their HW. This has been a problem at other internet math communities, and is the fear most people express about whether MO will function. But at the same time, this means that someone on the site has to make judgements about what are good questions for the site and which are not. The point of the SE software is to not leave this judgement entirely up to a group of dictatorial moderators, but rather to give all the participants of the site a voice in the decision.
So, I think it important to keep in mind that when you look at a question and say to yourself "this question shouldn't be here," there's a decent chance that a lot of reasonable users out there on the site who disagree with you, even in some cases where you might not suspect it. Just as an example (brought up by Douglas earlier) at http://mathoverflow.net/questions/14120/problem-with-pearson-correlation-coefficient-closed, I honestly disagree with Scott's decision to close the question. I respect other people's opinion the other way, and see the logic for it but I feel like that was a straightforward question with a simple answer: "You can't."
So, I am suggesting that users err on the side of caution when commenting on questions. As I have said consistently, if you have a suggestion for improving the question, by all means make it. If there is something you think should be clarified, please say so. But I think the risk of creating a hostile climate outweighs any benefits from comments that just say that a question is inappropriate. It is just too easy to say things that offend people without meaning to.
Harry- I would, but he didn't put in his email in his user info.
Oh, and in the case of the missing comment. I'll admit that I flagged it, though I forgot that this would delete it (and didn't realize it did so irrevocably that other moderators couldn't see it). I certainly meant it as no disrespect to Douglas.
I'll note that there was no way for me to have discussed it with Douglas except starting a meta thread, since he didn't put his email in his profile.
Douglas has mentioned in the past a seeming inconsistency about some things being jumped on, some things staying open, and some (meta-)questions/issues not being addressed for a long time. Am I being too naive in chalking this up to time differences, working routines, a front page that fluctuates with edits, rather than some "unwelcoming behaviour by the community of regulars"? (I am not thrilled by the implication that I am part of a herd of closed-minded, my-corner-of-pure-mathematics-is-all-that-matters-and-applied-maths-sucks, yahoos; and I don't think other people here would be. If anyone can divine my research interests and general mathematical interests/attitudes from what little they see of me on MO, I'd be impressed.)
If I don't comment on applied probability questions, positively or negatively, that would be because it's been over 8 years since I touched the stuff, and I see no reason to leave ill-informed comments rather than leave none. I suspect the same goes for other people on the site, too.
Lastly: I'm sorry Douglas felt things had come to such a pass that he might as well delete answers he's given (to Profs. Gowers and Stanley no less, dont'cha know). That would seem not just a shame, but a sign that undue offence had been given to a degree not hitherto realised; and that's the greater shame.
Yemon-
The time differences thing is huge. I'm the only moderator in the Eastern Time Zone (which is part of why I was promoted; it certainly wasn't for my way with words. Originally all the moderators were in California), and I'm not an early riser. Thus it's not unusual that no moderators have looked at MO before 10 or 11am EST. Also, the vast majority of the users with vote to close privileges live in the Americas (Andrew Stacey, Jose Figeroa-O'Farell, Kevin Buzzard, Tom Leinster, Ilya Nikokshev, Dimitri Panov and Georges Elencwajg seem to be the only residents of the Eastern Hemisphere with these, so it takes all but two of them to close a question), and all of us are busy. I didn't answer Douglas's comment earlier because I'm finishing two papers, going to Austin next week and seem to be coming down with the flu.
It's also just a really hard question which questions should be closed. For example, here's one that was just flagged and has several close votes: It could turn into a pure vs. applied food fight as PLC suggests but maybe somebody has something really valuable to say on the subject. By the way, I think Pete's comment is a great example of positive input on a question he still says he's voting to close.
Technically, Tom Leinster is in the western hemisphere and Kevin Buzzard probably oscillates between the two on his drive to work - assuming he can afford to drive to work.
@bwebster: Yes, that roughly accords with what I expected. I have only started spending large amounts of time here owing to massive work-avoidance and a fiddle-while-Rome burns approach to a couple of deadlines; as they actually start to bit I expect I'll be dropping off the site for a while.
@Andrew- This is what I get for trying to not be continentist. I should have just stuck with "the Old World."
Harry- There's no reason we couldn't, but I think it require an Oldworldian nominating themselves to the moderators.
Reading the above in chronological order, it looks as though Ben is agreeing with Harry's idea that we all go and jump over the meridian laughing manically.
Sounds good to me. Get a few more involved and we could make it into an event at the first "eastern hemisphere" MO conference.
actually, it just looks like I'm saying there's no reason we couldn't jump across the meridian maniacally, which is a statement I will stand by.
On S[OFU], they hold moderator elections when they need more moderators. I was thinking we should do that, but wasn't planning on it anytime soon.
some things being jumped on, some things staying open, and some (meta-)questions/issues not being addressed for a long time
I expect that this is how things will always work. Many people may not even realize that they have an opinion or that a problem exists until somebody says something. Can somebody give an example of an issue that's not being addressed? I feel like we're being pretty thorough here on meta; if somebody brings up a problem, it's usually followed by a pretty good discussion where minds are changed and decisions are made. Sometimes there isn't a pretty resolution and the issue comes up again later, but I don't feel that would change if we added another moderator. After all, there isn't really any difference between moderators and normal users on meta.
As for individual questions being closed or not, moderators and 10k+ rep users have access to a list of all questions that have recent votes to close, all recently closed questions, and all recently reopened questions. If you're a mod or have 10k+ rep, you can see that information here. I try to keep an eye on that page, and I hope the other mods do too. That should mitigate time zone issues.
Like Ben said, it's also become harder to decide when a question should be closed. In the early days, it was the norm for moderators to single-handedly close questions and moderate heavily because we had to carve out MO's niche. Now the niche is more defined, so there are more likely to be more actually borderline questions and fewer blatantly awful questions.
@Harry: I do agree that it would be nice to have a moderator between UTC+9 and UTC-3. I wanted to hold off a bit longer on moderator elections to get a larger pool of candidates, but maybe that's not necessary. Alternatively, we could have a very informal (compared to SO) election process here on meta to appoint a single new moderator (in the manner they elect a new pope).
Here's some initial thoughts:
So if (that's a big if) a moderator is needed in the "Old World", may I recommend something that I suggested a while back: that the person(s) have two accounts: a normal one and a moderator one. It is extremely simple to have two OpenIDs and to log out and log back in as another if one sees Bad Behaviour. Indeed, I do this on the nForum: most of the time, I'm "Andrew Stacey" but just occasionally I need to be "n-Forum Administrator".
My original suggestion didn't get anywhere, but I think that was partly due to the fact that all of you moderators already had your accounts. But the proposal to have a new moderator would involve creating something new and so there's no reason why it can't involve a new account. At the very least, it would be worth doing the experiment to see if having two accounts with different privileges was reasonable.
(I know that Anton has requested a feature to make this superfluous, but we have no way of knowing when that will be implemented, or even if it will.)
@CSiegel: I think that that is the best long term strategy. However it may be felt that there is a need for a short-term fix until that is viable.
@Andrew and Charles: Perhaps some people outside the US could be persuaded to take up temporary administrative duty until more people hit 10k, if this is needed.