Not signed in (Sign In)

Vanilla 1.1.9 is a product of Lussumo. More Information: Documentation, Community Support.

    • CommentAuthorquid
    • CommentTimeOct 8th 2011 edited

    One of the most popular MO questions, by some metrics the most popular, Examples of common false believes was recently closed. It is now about to be reopened (4 votes at the time of writing).

    It is my opinion that for such a high-profile question it might be better to arive at a consensus or a solution via discussion rather than to send it through various open/close cycles.

    My opinion is: It was good it was open for a long time, but now I would prefer if it was/stayed closed.

    Here are the reasons.

    It seems to me the software is not really made to deal with question with that many answers. I am often on a slow connection with poor harware, it is really difficult to handle this and related questions (on technical grounds).

    There is an ever growing risk of duplicate answers and almost duplicates. Which is particularly large as there are only limited ways to structure and or search the answers (for the latter one could use google in a targeted way as a work around, but still); cf. the first point.

    Finally, and semi-seriously, if a false belief was not added over that long a time it can't be that common.

    Afterthought: In case some should really want to keep it open, one solution I could imagine is that one or a group of these users does for this question what Gil Kalai did/does for Fundamental Examples that is make it so that one can get an overview over the already given answers.


    I upvoted your comment, quid, and I hope others do the same. It would be good if those who voted to reopen would explain their reasons.

    • CommentAuthorgilkalai
    • CommentTimeOct 8th 2011
    I voted not to close.

    We dont have a general policy to close old questions (unless, in a few occasions that the OP himself asks for closing his question) and I dont see any reason to have such a policy, in general, or in particular for this question. (Or in particular, for big list questions, or for soft questions etc..)
    • CommentAuthorquid
    • CommentTimeOct 8th 2011

    @gilkalai: There is a policy. This is what Francois Dorais said some days ago on meta in a general context on closure

    Big list questions need to be closed after the big list has been generated and new list items are unlikely to provide any new insights.


    Yes, quid is correct. I didn't really explain the reasons why when I wrote this since that would have been off-topic. So here are some of the main reasons:

    • Users are unlikely to carefully read through long lists, which leads to many duplicate answers. (This particular big-list has a large amount of deleted duplicates.)

    • As time goes on, new list items tend to be more and more detached from the topic. (This particular big-list has been more on-topic than others, but this effect is still perceptible.)

    • Some big-lists are are inexhaustible (e.g. 1083, 5450). If left open, they would keep coming back to the front page with high frequency, which rapidly gets annoying to regular users. (One could try to make a case that this one is not of this type, but with that many answers...)

    • CommentAuthorgilkalai
    • CommentTimeOct 8th 2011
    In my opinion, a question should be closed only if there are good specific reasons to close it. (Actually I do agree that the big list question about jokes in mathematics should have been closed after some time, but this is not a typical big list question. ) Overall, I think it will be a bad policy and a major waste of time to actively go and close old questions.

    As far as I can remember, only one user actively seeks old questions to close them, without much success. This one got closed because of some new duplicate answers that brought it back to the front page.

    The policy strikes me as healthy and sensible: if we let big list questions get too large, they will become like bad journals: write-only. Users will write new entries and no-one will read them (hence the duplicates, btw). This question in particular has been very successful in attracting good answers, closing it will spare it the indignity of this slip into irrelevance.

    The problem with old "big list" questions is that the software isn't optimized for them. You'd want new answers to pop up near the front so that people could decide whether to vote them up. Since all new answers go to the end, they end up stuck there because no one sees them. In my opinion, once the first page is full, there's little point adding new answers to big lists. As Thierry says, those answers are write-only.

    @Noah: this is not accurate: you can choose to view the answers ranked in one of several ways, and newest first is one of the options, and the one I use. The software always puts the accepted answer (if any) on top, but sorts according to oldest/votes/newest and remembers your choice from session to session. I like newest first, though I guess it will not push up old answers that have been edited recently, which is unfortunate.

    Of course, I bet few users have newest first as their choice. Anyway, once the number of answers goes above a certain limit, there isn't any satisfactory way of viewing them anyway.
    • CommentAuthorgilkalai
    • CommentTimeOct 9th 2011
    Noah, when you choose to order the answers by "newest" (and not by votes) you will just see just the newest answers popping up in the beginning and the oldest answers at the end.
    • CommentAuthorquid
    • CommentTimeOct 9th 2011

    Since I might have caused a misconception above by quoting just a recent mention of the policy, but perhaps I am overinterpreting the 'will' in Gil Kalai's comment, I add that it is already in use since a long time; for example due to an edit Most interesting mathematics mistake? just reappeared on the front-page; closed with this rational in March, 2010(!), which also might qualify as a more typical big-list question.

    Finally, I do not quite understand why the wish that there should be a good specific reason for closure is in itself in contradiction with the policy or the reasons given by some (including me). For example, having aquired (too) many answers is something specific to a question and of direct practical relevance (as opposed to a policy or reasoning along the lines, after a year a question is automatically closed).


    I will - reluctantly! - admit that some of these big list questions have provided useful insight. But I feel that the MO engine is not optimal for recording this. It might well be the easiest way to gather the information, but it is definitely not the easiest way to use it. None of the sorting methods is useful in this regard.

    So what I would like to see from those who champion the big-list type of question is a little extra effort. Why not take the answers from this question and put them somewhere in a little more organised fashion. Speaking for myself, I'd be happy to see this on the nLab. Then one could order them by topic (for example) and link concepts to explanations on other pages.

    When the "counterexamples in algebra" question came up, I did this - though I've failed at keeping it up.

    If this were done, and maintained, I would even be happy at keeping these questions open because I would know that someone would be keeping an eye on the thread and ensuring that any new information was put in a more useful place. It would also reduce the number of double posts as a link to the ordered page could be put in the main question with a link "Check to see if your answer is here first". As it would be a page with some system, it would be easier for a potential contributor to see if their answer was already there. Of course, not everyone would check this but some would.

    • CommentAuthorgilkalai
    • CommentTimeOct 10th 2011 edited
    The apperant policy over stackoverflow, see is this: Good big list questions are kept open. Questions that are closed are only questions that are considered as bad (although earlier they may have been accepted), and people are explicitely warned against asking similar questions.

    **Edit:** Quid (next post below) challenged my impression from stackoverflow and indeed he is right.

    Going back to MO, MO's main role is not to create new information (although this rarely does happen) but to move existing information around and sometimes to organize existing information in various ways. Good big list questions were useful for this purpose no less than other question (perhaps even more), and this accounts for their popularity and for upvotes they gained.

    What is special about these questions (and perhaps a few other types of questions) is that they annoy some users, that especially dont like when these questions resurface on the front page. (I regard this particular question resurfacing on the front page once every few weeks as something beneficial to MO and to MO's users.)
    • CommentAuthorquid
    • CommentTimeOct 10th 2011

    @Gil Kalai: Could you please elaborate a bit. Sorry in case I am being dense. But I just clicked around there (on SO) on a couple of questions with many answers at the top of that list, and got a different impression. I did not find a single one in a somehow usual state.

    Many are closed or even locked. One (Hidden features of Phyton) is 'only' protected (something we do not have) and has a sort-of 'table of content' in the question. And another one 'Phyton IDE' while seemingly open has a huge warning not to ask something like this, and docuemnts a large amount of work in the question.

    So, I did not find anything comparable to the present situation. That is an open question with that many answers where there is no table of content like infomration in the question nor any other information edited into the question that sofar was obtained from the list.

    As I said right at the start, in my afterthought, to me it is to a large extent the lack of any overview information that makes it difficult to find this question still being open useful. (I did find it a good question, and I will continue to do so, after all it woud still be visible; only an ever continuing stream of duplicates and semi-on-topic answers is something I do not find useful. [At least the former could be avoided, by this overview information.] And, Andrew said something similar, and Thierry statement is in some sense not that different,

    Or, for comparison: recently somebody (I believe David Roberts) suggested to close Proofs without words; Mariano said he would like the question to remain open and said that he pays attention that there is no accumulation of duplicates and not really on-topic answers.

    For this question, so far, nobody did or said anything along these lines. Only five users anonymously clicked reopen and you argue mainly abstractly.

    Before your last comment not a single person said "I find it useful that the question is open because ..."

    • CommentAuthorgilkalai
    • CommentTimeOct 10th 2011
    "Only five users anonymously clicked reopen" I don't know that there was anything anonymous about those who clicked to reopen. Ususally such votes are completely public. (On the other hand, **you** are anonymous, quid, arn't you?.) Anyway, I dont think we should, in general, close good old questions.

    Let me copy the entire relevant part of the post that quid quoted above:

    There is a general misunderstanding of what closing a question is supposed to do. Closing a question prevents new answers to be added, and has absolutely no other effects. (Closed questions are just as visible as open ones, they can be voted on just like open ones, they can be edited just like open ones, and comment threads remain open after closing.) So why would you want to close a question? There are three basic reasons (and multiple variations):

    • Off-topic questions need to be closed because such questions should not be answered here. In this case, it is better to include a link to a more appropriate resource.

    • Big list questions need to be closed after the big list has been generated and new list items are unlikely to provide any new insights.

    • Questions that are too vague, lack context, or otherwise need serious editing should be temporarily closed until appropriate edits are done. This is because answers appearing before the edit are very unlikely to correctly answer the final form of the question, so it is best to prevent such answers from appearing altogether.

    Closing of type 1 are permanent; closings of type 2 are permanent but usually occur several months after the question has been posed; closings of type 3 are intended to be temporary, but if the edits are never made they become permanent.


    The voters are not anonymous:

    • CommentAuthorgilkalai
    • CommentTimeOct 10th 2011
    I oppose to the second item in Francois list. It gives us enough work to decide on closing new questions and I dont see any reason to add the burden of closing old good questions. I also dont think there should be a special rule for big list questions.
    • CommentAuthorquid
    • CommentTimeOct 10th 2011

    Regarding the technical point of 'anonymously':

    François thank you for providing this link; I was unaware of this possibilty. (OT: Is there a clickable option to get there for any question?)

    What I always new is that typically one can see this via the revision history if there is one. However, for this question there is none. Of course, I could have created one via making an edit, so that I always knew that I could find out and the 'anonymously' was not meant as 'secretly' or with any particularly negative connotation, but mainly as without giving an explanation or reason; sorry for the poor choice of words.

    Gil Kalai, I did not mean the anonymously in the way you perhaps understood it, sorry for any misunderstanding. However, what I would still be curious about (and as you can see I already invested some time to look around there and to understand) is what precisely you had in mind with your comparison with SO.

    @gilkalai: "I oppose to the second item in Francois list.". Fair enough but I guess other people do think differently. For example I would say I was pretty much completely opposed to all big list questions even being on the site at all, far preferring a precisely-posed and perhaps technical mathematical question at graduate level (perhaps reminiscent of a year or so ago when I was far more active on this site). I appreciate that this view is probably very different to yours but I'm just flagging that people with this view exist. I would censor this site far more. I would rather the site comprised only precise and perhaps technical questions because these ones give me by far the most pleasure. Not all questions I've asked have been of this nature, but most have, and I do think it's a bore when old chestnuts pop up again and again just because the software thinks it's important to flag that someone has given an answer to a big list question. The problem, as has been said already, is that when there are >= 10 answers or so, answerers don't even bother to read them, and just often post duplicate answers. Closure solves this effectively.
    • CommentAuthorgilkalai
    • CommentTimeOct 11th 2011 edited
    Hi Kevin, of course I realize that there are people, well (over)represented over meta, who do not like big list questions, and there is also some opposition over meta to applied math questions, to questions with philosophical ingredients, to soft questions, and more. I was not aware that your own view is so extreme in this direction, and that you support far more censoring (in what direction?, by whom?) on the site. I note that your expressed view is not entirely consistent with your own questions and answers.

    When it comes to good big list questions it looks that the majority of users like them, as expressed by number of visits and upvotes. I bet that even the latest answers to the false beliefs question (which were nice) were read by far more users than an average answer to an average question. (Also I am not sure duplicates is a real issue.)

    It is understandable that you would like to see those questions that give you by far most pleasure, but it is not clear why you want to close questions that give pleasure to other mathematicians, with a different taste or different areas. Note that the kind of questions that please you less may be more appealing for mathematicians who are not working in the few mathematical areas that most MO's questions are about.

    I'm also largely a fan of the focused, technical question. I have soft questions and big list question on my ignore list, as I find too many of them to be gossipy and low in substance. That said, I have contributed to plenty of soft question threads, but since they're on my ignore list I tend to find out about them rather indirectly.

    • CommentAuthorvoloch
    • CommentTimeOct 11th 2011
    This discussion is kind of pointless. If five users with enough rep think that a question needs to be closed, it gets closed. If five others disagree, it gets reopened. If the general opinion is skewed one way, it will eventually stabilize that way. I am not going to refrain from voting the way I feel just because somebody disagrees with me.
    @voloch: fair point, but it cannot hurt to build a bit of consensus around policy first. One rationale, which has been specifically invoked in regards to the question being discussed here, is to avoid opening/closing wars when there is a lack of consensus. As MO gets older, more and more users have opening/closing privileges, making the odds of such a war higher and higher. It's a good idea to try to reach a consensus beforehand, rather than eke it out of countless battles over each questions.
    • CommentAuthorStorkle
    • CommentTimeOct 11th 2011

    voloch is right. The discussion is pointless and a waste of energy---no one is going to change his or her mind and we are all going to continue to be upset every time there is an open-close-open-... cycle that doesn't end the way we want.

    In fact, the whole tug-of-war between the hardliners and liberals should be irrelevant, except that the software has a very serious flaw: it doesn't allow for effectively ignoring questions with certain tags. If ignoring a tag had the effect of not seeing questions with at least (or only) that tag at all (instead of just graying them out), the two communities would never have to interact like this. But of course we can't change the software. Is it likely that migrating to SE would remedy this?

    • CommentAuthorgilkalai
    • CommentTimeOct 11th 2011
    Felipe, I dont understand what you are saying. The question was closed, then it was open by five users and then it was close again by five other users. How such a process is going to stabilize and what is the meaning of the "fix point"? It is also clear that there is no consensus. My main suggestion is to reduce the decisions of closing and opening(of any kind) to new questions and when there is a good question that is already open to keep it open.

    Storkle, there is a 'Hide Ignored Tags' button in your preferences to do exactly that.


    @Francois: small point, but if thread has a hidden tag and also an "interesting" tag attached, then the question still appears, but is greyed-out. IMO that's a pretty good compromise. If it's hidden but has no "interesting" tag, then it does not appear.

    • CommentAuthorvoloch
    • CommentTimeOct 11th 2011
    Gil, if I understand the software correctly, a user cannot vote to close or reopen twice. So the close/reopen cycle is in effect an election, with a slight advantage to the "close" party, among those who care enough to vote.
    • CommentAuthorquid
    • CommentTimeOct 11th 2011

    Since I started the discussion, and in view of Storkle's last comment, I would like to say that I was, and still am, willing to change my mind.

    I even sketched out a precise scenario under which I would (this was just a suggestion, not the only possibility), and Andrew said something similar.

    If somebody gives me a good reason why the question should be open, I will vote to reopen. But what I heard so far is not very convincing; not only but also, as it is often based on assumptions that turn out to be inaccurate.

    That being said, at the moment, I too feel this particular discussion is pointless. Sorry for having started it.

    • CommentAuthorStorkle
    • CommentTimeOct 11th 2011

    François, thanks! That is, by far, the most useful thing I have learned on this thread. Out of curiousity, when did this happen? I have not been paying attention for a while.

    Second, can someone tell me why it is still necessary to have fights over the closing of big-list questions? If you don't like them, do what I just did: add big-list, soft-question, and mathematics-education to your ignored tags and then select "hide ignored" in your prefs.

    • CommentAuthorquid
    • CommentTimeOct 11th 2011

    Storkle: regarding your second, for example, because I do not have anything against big-list questions in general. I only think that this particular one reached a point where it now should be closed. And the discussion is, at least so it was conceived, about this very precise question and what are the advantages and disadvantages of keeping it open or not. (That some people do not stay on topic, oh well, I said I am not happy with the discussion.)


    @Storkle Another reason might be that people want their technical questions to be prominent. Let's imagine that no questions are closed any more, especially not the most popular ones. Then, the front page will be largely populated by big-list and homework questions. Most new users will not know that you can make such a selection of visible questions and will not bother, even if they knew. They will just come to the site because somebody told them to have a look at it, they will look at the first page and will come to the conclusion that it's just a gossipy site overrun by undergrads and cranks.

    In other words, the rationale behind unwanted questions is not just that the closer doesn't want to see them, it's about what he would like this site to be and to do. Not least, it's about the impression that the site will create upon a famous mathematician, say, who comes here for the first time, not knowing what to expect, possibly slightly sceptical in advance, since after all, this is "just an internet forum".

    • CommentAuthorgilkalai
    • CommentTimeOct 12th 2011 edited
    Another good big list question that was closed without good reasons is this one
    PS I am also largely a fan of the focused, technical questions.
    • CommentAuthorquid
    • CommentTimeOct 12th 2011

    Gil Kalai: thank you for providing one more example for the existence of the policy whose existence you initially ignored.

    If your actual intent should be to further discuss question 2144, or big-list questions in general, or said policy, I suggest you create a thread to this end. Thank you.

    • CommentAuthorgilkalai
    • CommentTimeOct 12th 2011 edited
    Hi quid, indeed there is some "quasi policy" of this nature which I was not fully aware of. As I said, I personally do not agree with this policy and I also dont understand the gap between the vast popularity of the questions we discussed and good big list questions in general, and the hostility towards them in these discussions.

    I can see that in "real world mathematics" some mathematicians would almost certainly turn their nose up at some activity which is les than full hard-core mathematical research and even at different areas or at lower levels of excellence. Such an attitude can be very instrumental for some people in order to focus on the important things they are doing, I certainly do not see something wrong with such an attitude. (But there are also mathematicians with various different attitudes.)

    But MO consists mainly of outreach and educational activity in research-level mathematics, and it includes mathematicians of various levels, qualities, and interests, so I dont think MO is a place to demonstrate a very mathematically puritan approach.

    In the case of the single paper question the initiative to close looked completely arbitrary: in the middle of some discussion about another question that the OP himself asked to close somebody offered, out of the blue, to close this one.

    Dear Gil,

    From my perspective, the main issue is the negative connotation you associate with closing. In the case of big-list questions, the reasons for closing are almost never about the question itself, but about some of the negative side-effects it may inadvertantly generate.

    Question 2144 is a good example of this. If I recall correctly, it was closed right after some spammy user kept posting religious propaganda in answer to this question. That was obviously not a judgement on the question.

    I personally believe that big-list questions are important for MO. For example, they are very effective at attracting new high-level users to the site. However, I can't ignore their bad sides. I believe it is better to have A regular flow of new big-list questions than to have a long queue of good oldies popping up all the time.

    • CommentAuthorquid
    • CommentTimeOct 13th 2011

    Hi Gil Kalai, while as said I would prefer to discuss such general matters in a separate thread I 'give up' and answer here. Also, I should have written my last comment less harshly, but I got a bit frustrated with this discussion.

    One reason for this frustration is that you state various things as if they were established facts, but that in my opinion are not, sometimes going well beyond the subject matter at hand. This puts me in the position that in some sense I have to write an analysis of all these claims to point out why I beleive they are opinions and not facts or even inaccurate.

    To wit, the SO-policy. It took me about half an hour to familiarize myself sufficiently much with the situation on SO and then to write down what I have learned. Now, to me it seems clear that the situation on SO does not at all support the idea of just keeping open big-list questions indefinitely. It would if the questions that are locked--not visible in the title--were open, but they are not; somehow I suspect you overlooked this fact that these question are not open, but in a worse state than being closed, to arrive at your conclusion. So it seems to me the SO-policy actually is not unlike the policy we are discussing. Now, if to you SO-policy is relevant to MO-policy (if it supports your point), I would assume that it stays relevant if it turns out actually not to support your point.

    Likewise, and even more drastically, if the only argument against closure you initially made is that there is no policy (under which this closure is justified) and therefore the question cannot/should not be close, then it would seem obvious to me that if there actually is a policy this should be of significant relevance to you. But, you mainly chose to change the subject.

    And, I could continue this list, analysing various things you said to Kevin Buzzard and to me. Not to make this sound overly critical; I think it is actual a smart way to discuss, in particular as it puts a significant burden of argumentation on my end, which is a bit time consuming in writing, and then I do not want to bore the meta.MO community at large, with pages full of analysis as to whether it is justified to say that there is opposition towards applied math on meta (where from the context this has to be understood as more so than on main) and so on.

    But just one more thing for Question 2144 (not the present one). Already many month before the question was closed comments critical to the question and/or making clear that its closure might be imminent where posted (with many up votes). It is true, the precise point in time when it was closed is somehow arbitrary. However, I do not rememeber this precisely (and was not even active in the earlier times), but from the visible answers it seems, that at the end of 2010 (each time the question reappeared) something critical was said. And then in July 2011 when it reappeared after a long 'sleep' it was evetually closed for real. So, in some sense closure was long overdue and it just was 'forgotten' or not relevant when the question is idling somewhere. Also, I disagree that it is a good question; actually the question is not so bad, though quite subjective, but many answers do not match question but answer something else, which one could also ask, but it was not asked in this question. Finally, there was essentially no opposition to the closure (at least none I noticed or is still visible on the question). The only one was you saying 'Vote not to close.' Now months passed, and still noone of the silent majority you invoced complained by a comment or in any other way.

    • CommentAuthorquid
    • CommentTimeOct 13th 2011


    Now, I said silent majority, you might say it is not silent the votes and view are visible. For the latter, the views, I would really give very little significane to this. If one watches the front page. Often questions with negative vote count have the most views as people look, iassume, what is going on. So going by views one could make an argument for posting pure spam as it gets many views. Okay, this is a short time effect and would not last. But, it is also in some sense clear that old and visible question have many views. If a question has many votes it isclose to the top of the question listing by votes. And, lately I was visiting a couple of SE sites out of pure curiosity how they are run and look. What did I do, I checked some random questions, and then those with most votes. Contributing a view. Yet this/my view is 100% irrelevant (except perhaps for somebody wanting to sell adds); in most cases I cannot even rememeber what the question was about on, say, the cooking site or the english site.

    Now, votes is a bit more subtle. Yet, still they way voting works on MO is in my opinion also quite arbitrary. Recently a picture of a cat with a somewhat fitting but not overly informative sentence (a LOL-cat or whatever the precise name of this type of internet-trivia is) was posted as an answer almost at +60 now. I really do not think we should encourage posting animal-pictures on MO although it seems highly appreciated by the community. I also have nothing against the fact that this picture was posted, all I am trying to say is that because occassionally something is appreciated by the community does not mean that when the volume of this type of content would be increased it would still be appreciated. By analogy, if I am at a math conference and a speaker tells a joke, and I find it good, I appreciated it and it can make the talk better or more memorable;a and if I do not find it good I do not mind. If however some conference organizer should make the observation that it seems the most appreciated part of the talks are in fact the jokes, and thus arrives at the conclusion it would be best to invite comedians instead of mathmaticians as speakers that will tell nothing but jokes, then the conference can take place without me the next time.

    So, I and I believe not too view others either appreciate or at least do not mind so much the occassional big-list, soft, semi-serious or whatever question. But, it should remain the exception. This is also an argument for sometimes closing an existing one. In a rough sense my goal is to have a constant number of open ones. So either I occassionally endorse to close an old one or I endorse to close each and every new one.

    For reasons explained by Francois, I actually think it is better to occassional have a new one (both for new and existing user).

    How long exactly any particular big-list question (and the same goes for questions with similar characteristics) should remain open, in my opinion depends on various factors such as appropriateness for the site (examples of courrse much better than say jokes), quality and quantity of the answers (if a list grows slowly one can keep it around for longer than one that explodes, if something stays on topic one can keep it longer than if it degenerates into a collection of loosely related pieces of information dressed up as answers), and finally also how many poeple still actively care about the question being open.

    • CommentAuthorquid
    • CommentTimeOct 13th 2011


    Back to the starting question: as I said I find it a good big-list question for MO and the quality of the answers stayed solid for quite some time (though I believe there is a negative trend) and it meanwhile really has very many answers, and due to the nature of the subject (my original point under 'semi-seriously' I find it very unlikely if something amazing would be added in five months [for a question asking about something extremely rare this would be different, but it is about something common]).

    So, due to its quality it stayed open very long (in my opinion the right measure here is not time, but passages through the frontpage), but now some people including me thought it was perhaps long enough. (In recent month the question was idle, now it resurfaced, chance are if it was not closed it would actually be a more frequent visitor to the frontpage in the weeks to come.) What I wanted to know, and to me is the point of this thread is to know how many people still actively care about this precise question being open.

    If a few people or also only one person would have told me something like: "I would really appreciated this question would stay open, because [reason]" and this is somehow credible (the reason would not even have to be overly serious), then I would have said, okay, let us keep it open. (You can check the formulation in my original post is only 'I would prefer' and I meant it like this, it is not that I see any urgent need to close it.) However, during this entire discussion noone said anything to that extent. The closest to this were you stating that you consider it as good if the question occassionally appears on the front-page, which to me is some reason but for this particular question a particularly weak one as it is anyway at the very top of the question list by votes. So, that it can hardly be overlooked by anybody new to the site looking around a bit, and everybody else knows it anyway. So the only point of this could be a reminder for long time users that the question exists. To be honest, I am not sure we need this.

    In the very end I howver think that our view on what should be an should not be on MO are not that different.

    I am not hostile towards big-list question, I just do not want too many of them; and since you said that you are laso a fan of focused, technical question there seems to be not so much dissent here. Only, what I do not want is some sort of automatism that you seemed to advocate that tells that just because at some point a big-list or soft or related question was not closed right away or for a long time, it will have to stay open indefinetly.

    Indeed, if I may say so, if your goal is to keep MO open for such content, I believe your are doing this goal a diservice by insiting on this. Because if the situation were like you suggested I would be much more critical towards such questions at the start. Only because I now that we can still 'pull the plug' later in case things go wrong, I am willing to take more risks at the start.

    • CommentAuthorgilkalai
    • CommentTimeOct 13th 2011 edited
    Hi quid, thanks for your detailed analysis. Regarding SO, as I remarked above in an edit to my relevant post, my initial impression was incorrect.
    • CommentAuthorgilkalai
    • CommentTimeOct 15th 2011
    A few more points:

    To quid: The cat answer is excellent. The demand to supply *proofs* to claims made is one of the most important contributions of mathematical and scientific thinking.
    (Overall, I tend not to agree with what you wrote above about votes and popularity.)

    I agree that we usually have similar views on policy questions for MO, and even when we dont agree I find your views reasonable and I appreciate your endless willingness to discuss them. My view regarding
    anonymous *participation* (an issue discussed here before) is also highly permissive. However, I would feel very uncomfortable with anonymous *administration* of the site since I think administrative decisions and the identity of people making them should be transparant.
    Indeed, moderators are not anonymous and I think this should be the rule. Therefore, I feel also somewhat uncomfortable when anonymous users (including you) are very much involved in (actually) closing and openning questions.

    To Francois: In 99% of cases closing a question has a bad connotation, I dont see an easy way around this.

    Overall, as I said, I dont support a policy of closing good big list questions, and I dont feel comfortable that this question was closed, partially because it was a "flag ship" of successful MO questions, at least in one direction, and partially because the answers it attracted are still rather good, so the decision to close looked (also in other cases) arbitrary. Of course, I realize that other people, even those supporting big-list questions, did not share my view.

    I'm perfectly willing to occasionally reopen old big-list questions when new answers come up. (Though I would expect the question to be re-closed after another while.) Just flag for moderator attention and explain why you want to reopen the question.

    • CommentAuthorquid
    • CommentTimeOct 15th 2011

    Gil Kalai: regarding the cat possibly I paraphrased that precise answer too negatively for effect, however if you look at the comments of that answer I would assume you agree that the second comment is quite a bit more substantial and relevant to the actual subject at hand than some of the other comments, which is not at all reflected in the votes by a huge margin (3 vs 20+). Let me say it like this: I think votes have some meaning, but one should be careful to only compare similar things according to their relative votes or other popularity count. A big-list question will in almost all cases win against a precise mathematical question with the big-picture and related somewhere in between.

    Thank you for your remarks on my meta contribution. In my opinion your remarks related to anonymity are interesting and legitimate. For this subject however I would really strongly ask everybody to not continue this subject in this thread. I have nothing against this being discussed, after thinking about it for a while I might even open a thread to this end myself, but please let us not mix this into this discussion.

    In any case, I did (so far) not officially vote on this question, and won't.

    • CommentAuthorgilkalai
    • CommentTimeOct 15th 2011
    Quid, after you quoted the OP of the maiden-names question who so galantly worried not "to keep us from our research any longer," maybe we should learn something from her and take it easy on our endless discussions. In particular, let's not open a discussion on the meanings of the votes to the *comments* of Felipe's answer. (I appreciate you thinking for a while on the anonymity issue, there is no hurry whatsoever.)

    IMO, Tim should have put this question on his blog rather than on MO. I think that this and other interesting "big list" questions detract from the main purpose of MO. If a new user checks to see what questions on MO have gotten the most votes, the user finds that Tim's question is #1. This gives the wrong impression of what MO is all about.

    Is it possible for the thread to be moved to Tim's blog? I have not asked him if he would accept that, but would do so if movement is possible and there is general agreement with my position. The thread itself, unlike most "big list" threads, contains so much useful information that I cannot bring myself to vote to delete even if I want it off MO.

    • CommentAuthorquid
    • CommentTimeDec 19th 2011

    @Bill Johnson, in case you care about a remark from an anonymous user, regarding organization and further maintainance it seems to me it could be simpler and more efficient in the long run to host this on some wiki rather than a blog.

    Above, Andrew Stacey mentions that some of the counter-examples in algebra were moved to the nLab. Another candidate, that could perhaps be also somewhat fitting subjectwise, could be the tricki (incidentally created on suggestion of or even by Gowers).

    ps: this is explictly not a contribution to the dicussion whether or not this should be considered at all.