27
$\begingroup$

The question Examples of common false beliefs in mathematics, was one of the best and most successful questions in the history of MO. When we discussed closing it, it was argued that after 164 answers and a year and a half, perhaps any added answer will not represent "a common false belief." (I was against closing the question, which is still my view.)

Two days ago the question was locked and frozen in a very drastic matter: as a result, it no longer shows in the list of highly voted questions, no editing and voting are possible, and it is not even possible to add comments. The question does not appear in the list of all questions, or even in the lists of all "big list" questions or "mathematical education" questions (Those were its two tags). On top of that, the following message appears:

"This question exists because it has historical significance, but it is not considered a good, on-topic question for this site, so please do not use it as evidence that you can ask similar questions here. This question and its answers are frozen and cannot be changed."

This message does not express the common opinion we had, even among those who wanted the question closed.

(It is possible to lock a question without freezing the entire thread.)

I propose to delete this message and unlock the question. (To be clear, I do not propose to open the questions to new answers.)

$\endgroup$
  • $\begingroup$ In addition to the above comment on technicalities here is my opinion: I do think some question should be put in this state, the current one is however at least not on top of my list for this treatment. (It should however also be recalled that whatever the "common opinion" was the opinion that this question should dissapear best completely from MO was also voiced already, basically for the reason that is given now.) But likely the simplest is not to start a debate but to just close it again as it was for a while now. $\endgroup$ – user9072 Oct 13 '13 at 18:35
  • $\begingroup$ (I deleted two comments on "lock" vs "freeze" and will likely delete the third one later, first as this is a side aspect and also since I think it is based on a mutual misunderstanding. In any case I am now convinced I misunderstood the intent in part. Sorry for the confusion!) $\endgroup$ – user9072 Oct 13 '13 at 20:16
  • 3
    $\begingroup$ As a practical matter, the closing of the question certainly makes sense to me: once the answers start either repeating themselves or are becoming lousy, it's probably time to close. I think it's likely that happened here (my +10k x-ray vision shows some lousy deleted answers already). I am less sure about locking and freezing, but it may be there are practical reasons for it that the official message is not fully conveying. David White's answer touched upon some possibilities. $\endgroup$ – Todd Trimble Oct 13 '13 at 22:09
  • 11
    $\begingroup$ I'm the culprit, converting the closure to a lock. Personally I thought that this actually indicated a more favourable position on the question --- in any case we don't want further answers on it (as I think had been agreed), but locking for historical significance, rather than closing, explicitly says 'this was good in the past, but no longer'. Beyond saying that, I will stay out of the rest of this discussion, and I'm very happy if we work out a mutually agreeable position, regardless of whether it fits with what I did. $\endgroup$ – Scott Morrison Oct 13 '13 at 22:24
  • 3
    $\begingroup$ If anyone wants to re-read the old discussion, it can be found at tea.mathoverflow.net/discussion/1165/… $\endgroup$ – Scott Morrison Oct 13 '13 at 22:27
  • $\begingroup$ It's impossible to close or lock without giving some reason. The message there wasn't written by Scott, but is the automatic "historical lock" explanation. We used to use "no longer relevant" as the close reason, but that's no longer a close option. $\endgroup$ – Noah Snyder Oct 14 '13 at 3:54
  • 4
    $\begingroup$ There are three lock reasons a mod can choose from, "content dispute", "historical lock" and "off-topic comments". Only the historical lock has some additional effects like locking all answers and removal from the top question list. Moderators can not use any custom lock reasons, but it might be possible for the site to request a different lock reason to be added. $\endgroup$ – Mad Scientist Oct 14 '13 at 8:34
  • 1
    $\begingroup$ The general matter of 'historical looks' was already discussed long time ago; then there was no easy way to do it, so it never got implemented. But the general idea had some support then. See tea.mathoverflow.net/discussion/1122/… $\endgroup$ – user9072 Oct 14 '13 at 10:59
  • 3
    $\begingroup$ One more point: there is in principle a jsutifcation for locking the full thread, namely that with editing privileges, which for CW is not much of a threshold, one can undermine the closure. This happened to some limited extent on this question (though it was I think not the immediate motivation for lock) and also happens on other such questions were some edit in new answers into existing ones while question is actually closed. One can say this is not enough to justify it, but then I also do not see that much need for such threads remaining editable and votable after some point in time $\endgroup$ – user9072 Oct 14 '13 at 12:42
  • 4
    $\begingroup$ @GilKilai: if the message is the problem, which of the current lock or close messages would you prefer? Too broad? Off-topic? I understand that the historical lock isn't perfect, but it seems to me to be the closest of our available options. $\endgroup$ – Noah Snyder Oct 14 '13 at 13:55
  • $\begingroup$ @Noah, but mods can remove the notice after locking without unlocking the post. A question doesn't need to have any notice while locked. $\endgroup$ – Kaveh Oct 16 '13 at 0:25
29
$\begingroup$

I am against hiding this. If the answers actually are of value then they should be displayed not hidden. More than that I would like the list of highly voted questions to actually be the list of highly voted questions.

$\endgroup$
  • 3
    $\begingroup$ Could you clarify in what sense you consider the question hidden? Is it just that it doesn't show up on the list of highly voted questions? $\endgroup$ – François G. Dorais Oct 14 '13 at 15:50
  • 6
    $\begingroup$ Francois, as far as I can see, the question does not appear at all in the list of questions: e.g. when you look at all "big list" questions or at all "mathematical education" question. $\endgroup$ – Gil Kalai Oct 14 '13 at 20:39
  • 7
    $\begingroup$ I don't think it shows up on a list of highly voted questions. I tested this. In general I would consider a question hidden if I do any type of search and it doesn't show up where it would be displayed if it were not locked or frozen or closed. $\endgroup$ – Kristal Cantwell Oct 14 '13 at 20:43
  • 1
    $\begingroup$ Kristal, the question does not show at all in "math education" or "big list" lists which is a shame. But if you know what to search: say search under "false beliefs: then you can find it... $\endgroup$ – Gil Kalai Oct 14 '13 at 20:45
  • 1
    $\begingroup$ Now that the question is unlocked and protected I see it is now showing up in the list of highly voted questions. $\endgroup$ – Kristal Cantwell Oct 15 '13 at 16:42
16
$\begingroup$

It's not really clear that there's much consensus, but here's a concrete proposal.

We unlock the question, but protect it.

This will prevent entirely new users from posting answers (although users with an association bonus from another SE site will be able to post answers), but otherwise everyone will be able to edit old answers, add comments, or vote, as per the usual rules.

If there's a further problem, we'll deal with it as necessary (e.g. by a feature request to increase the reputation threshold for protected questions, or asking for a new closure or lock reason).

(There is clearly reasonable disagreement on the correct course of action here, so I'm suggesting we take the relatively 'optimistic' one, and presume that, especially with the extra attention this thread has brought, we can revise the plan later if needed.)

$\endgroup$
  • 3
    $\begingroup$ I thought the association bonus was ignored when counting the required ten points for answering locked questions. Is that statement wrong? $\endgroup$ – Ricardo Andrade Oct 15 '13 at 0:38
  • 2
    $\begingroup$ Here's a reference for @Ricardo's statement: meta.stackexchange.com/a/52765 $\endgroup$ – Jonas Meyer Oct 15 '13 at 1:37
  • 3
    $\begingroup$ I went ahead and protected and unlocked the question. $\endgroup$ – François G. Dorais Oct 15 '13 at 2:36
  • 4
    $\begingroup$ This looks like a very good proposal, Scott. Thanks a lot. $\endgroup$ – Gil Kalai Oct 15 '13 at 2:46
  • $\begingroup$ I do not see much reason for being optimistic; the best case realistic scenario is that simply nothing much will happen, so that it would have just been as good had it stayed closed all along. But well as I said let us perhaps not debate too much. But, I take it now that you (moderators) unilaterally reopened a community closed question you will also intervene in this form if there are problem with duplicates or otherwis and close it again without much need for debate. (Notifying also @FrançoisG.Dorais) $\endgroup$ – user9072 Oct 15 '13 at 10:14
  • 3
    $\begingroup$ @quid: That question (like many others that have been the subject of a close war) has been on moderator management for a long time. Yes, we will close it again if there are problems. $\endgroup$ – François G. Dorais Oct 15 '13 at 11:44
  • $\begingroup$ @FrançoisG.Dorais thank you for the confirmation. $\endgroup$ – user9072 Oct 15 '13 at 12:05
15
$\begingroup$

I'm actually not sure at all that such questions should be closed, let alone locked. How sure are you than another brilliant and interesting answer will not emerge?

Rather than closing or locking, how about setting the threshold to answer higher for soft questions which already have many good answers and which are beginning to degenerate- say at 1000 rep (reasonably established users are not as likely to post garbage or to repeat one another).

This is a feature request, but it looks to me like it does the job much better than all of the close/lock options, which in my opinion are a bit ridiculous for a good question which had many interesting answers in the past, and could generate more in the future if given the opportunity to do so.

$\endgroup$
  • 5
    $\begingroup$ This is basically protecting the question with a higher treshhold. $\endgroup$ – Michael Greinecker Oct 14 '13 at 19:29
  • 10
    $\begingroup$ Now assuming these brilliant answers might come, why aren't you worried they'd come from a new user not yet above this limit? (Seems at least as likely then that some regular user will have new idea.) Also, I would find it odd to have a high threshold to answer some questions. The very low one for protection is justifiable on the grounds to exclude posts from people that do not yet "understand" the site, but frankly to tell effectively some, say, 754 point user you are not yet 'good' enough to answer the question feels a bit strange to me. $\endgroup$ – user9072 Oct 14 '13 at 21:14
  • 3
    $\begingroup$ @quid : It's less extreme that locking and closing, and it achieves the objective without hiding or defaming the question or preventing all new good answers. There's nothing sacred about the number 1000, except that it seems to me to be high enough as to reduce the "risk" of a repeat answer or of a non-answer below an acceptable threshold. $\endgroup$ – Daniel Moskovich Oct 15 '13 at 0:12
  • $\begingroup$ It is quite different in that sense I do not think it is justified to say it is "less extreme," as in another sense it is very extreme as it defames many regular users as being incompetent to judge whether to answer or not. In addition, when the question was closed people could have left their examples as comments. The 1000 is not key, the point is really that I find any non-minmal limit a prblem, and with a minimal one it is just protection. $\endgroup$ – user9072 Oct 15 '13 at 9:59
  • 2
    $\begingroup$ I completely agree with quid. An old question, that by now got answers from all users who had something interesting to say on the subject, can possibly receive a new interesting answer especially by a brilliant new user. For this reason I'm not sure it is a good thing putting a threshold for posting. Rather, I'd delete answers that received a score below some negative threshold. $\endgroup$ – Pietro Majer Oct 21 '13 at 10:01
6
$\begingroup$

I think we should ask for a custom lock (without freeze) saying roughly "This question is no longer likely to receive high quality answers" that we can use for successful soft/biglist questions that have outlived their usefulness.

They may say no, but it would be nice to actually have a tool that does exactly what we want instead of trying to adapt one of the wrong tools.

$\endgroup$
  • 1
    $\begingroup$ To "lock without freeze" sounds a lot like plain closing. Could you clarify the difference? $\endgroup$ – François G. Dorais Oct 14 '13 at 14:04
  • $\begingroup$ Good question @FrançoisG.Dorais. Personally I don't care much about close/lock/freeze, and was basing my request off what most people seemed to want. $\endgroup$ – Noah Snyder Oct 14 '13 at 14:26
  • 1
    $\begingroup$ Hi François, look at the following question: mathoverflow.net/questions/22299/… . It is locked but the entire thread is not frozen. Comments can be added and answers can be edited. The question also appears when people are looking for questions in general or in this particular tag. (In fact, it was locked by you :) .) $\endgroup$ – Gil Kalai Oct 14 '13 at 15:08
  • 1
    $\begingroup$ @GilKalai yes but the question (in the sense of thread) there is closed in addition to the question (in the sense of just the question-post) being locked. I think if it were not closed in addition then one could even still add answers when it is locked in that way only, which then brings us back to the question ask to you by Noah Snyder, yet not yet answered by you, which of the current closing reasons should be used. The one we always used 'no longer relevant' having been sort of replaced by that historical lock. $\endgroup$ – user9072 Oct 14 '13 at 16:28
  • 2
    $\begingroup$ Why without freeze? Put differently, why is it so relevant the answers to the question can be edited. Basically that this is possible leaves a loophole to sneak in new answer that is used by some. I would propse to just have the delisting optional and leave the rest as is (possible with a different text). $\endgroup$ – user9072 Oct 14 '13 at 18:31
  • $\begingroup$ @FrançoisG.Dorais Locking stops regular users from reopening the question, this is one of the major reasons for the historical log on SO. The other side effects of the historical log are supposed to prevent the question from being bumped to the frontpage by edits and prevent flags from being cast. If the community does not try to reopen such a question and avoid edits to the question and its answers, a lock might not be necessary. $\endgroup$ – Mad Scientist Oct 14 '13 at 20:35
  • $\begingroup$ @MadScientist there are no major problems of the type you mention, but as I said edits that simply sidestep the closed status and add "new answers" do happen. (I can provide examples upon request.) $\endgroup$ – user9072 Oct 14 '13 at 21:29
2
$\begingroup$

I am in favor of the closing and locking. When that question was asked, most people on MO were professors and advanced grad students. There weren't that many questions and so people read the question and answers carefully before responding. Nowadays MO seems overly populated with off-topic questions, and it seems we have many people asking/answering questions without reading carefully. The common false belief question seems unlikely to get good new answers. I imagine most new answers will probably have already been covered by old answers or they will not have read the problem correctly and will leave a crank answer questioning the validity of math as a whole. Similarly, hiding this question from the "top-rated" list seems like it might help with some of the confusion for newer users about what is and is not on-topic, and might help reduce some of the crank questions about validity of math. I'm sorry, but right now I can't think of too many examples of these crank questions/answers I'm worried about (the whole Trust-God fiasco springs to mind), but I know I've voted to close a number over the past month.

$\endgroup$
  • 10
    $\begingroup$ Hi David, the issue is not closing the question (it was closed years ago) but locking and freezing the entire thread. "..hiding this question from the "top-rated" list seems like it might help with some of the confusion for newer users about what is and is not on-topic, and might help reduce some of the crank questions about validity of math." This statement seems baseless and rather absurd. $\endgroup$ – Gil Kalai Oct 13 '13 at 22:05
  • 8
    $\begingroup$ @GilKalai: I take it you don't find my observation about an increased number of off-topic posts to be baseless and absurd? If so, then let's start there. I don't know why we have more off-topic questions, but it's a problem. One point of confusion for a new visitor to the site is that the notion of what does and does not belong has changed over time. To reduce this confusion I would like to reduce the visibility of off-topic questions, even if they were formerly okay. I'm especially concerned with this one because the title could be misinterpreted by an outsider as a broad criticism of math. $\endgroup$ – David White Oct 13 '13 at 22:38
  • 8
    $\begingroup$ @GilKalai There was a thread recently (by which I mean it may have been a month or two ago) about whether the number of closed questions had increased, with some data, see meta.mathoverflow.net/questions/302/… . As to the on-topicness of the question, it is debatable whether it would be considered on topic today. Generally soft and big-list questions are less tolerated than they used to be. $\endgroup$ – Tobias Kildetoft Oct 14 '13 at 7:43
  • 27
    $\begingroup$ "because the title could be misinterpreted by an outsider as a broad criticism of math." Huh? May be we should also censor "Killing form", "special forcing", and "joint distribution" on the pretext that the "outsiders" may get the impression that we are a violent group of ex-marines who smoke marijuana in their free time? $\endgroup$ – fedja Oct 14 '13 at 9:28
  • 10
    $\begingroup$ @fedja: I heard that those guys like to annihilate left and right radicals, and to collapse large cardinals too. $\endgroup$ – Asaf Karagila Oct 14 '13 at 14:04

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .