25
$\begingroup$

TL;DR: What does it mean when a post is "very low quality"?


The site allows us to flag a post as "very low quality", in which case that flag is reviewed by several users which can either dismiss it or recommend deletion/closure of the post.

But what does it mean "very low quality"? And I'm not talking just about what SE suggests that it would mean. I am asking what does it mean for this community.

Is "very low quality" just a matter of something which has bad formatting, or consists of something which is effectively a comment/different question? Or is "very low quality" something which should include posts which have negative mathematical impact?

In the past moderators would review these flags, and then there was a very clear issue that moderators are not referees, and unless the user really goes out of their way, the moderators rarely intervene in mathematical content. But this is no longer the case. Moderators don't review the VLQ flags, unless the original flag was disputed.

So should the users of the community review VLQ with respect to mathematical content as well, or should we still only care about outer appearance?

$\endgroup$
  • 3
    $\begingroup$ I do not mean to detract from this question, but I think we first need to ask "What is the purpose of MSE?". Any answer to your question will be taking a stance on this "meta-question". Is MSE a place to simply ask and answer homework questions, or is it a place to gather great questions and answers together? I still do not know. One stance will favour well-posed questions, another won't care. And so on. $\endgroup$ – user1729 Jul 8 '14 at 18:23
  • 2
    $\begingroup$ @user1729 The site could be all of those and much more if only SE would provide adequate tools so that each user can efficiently filter out the questions that they desire to see. $\endgroup$ – Bill Dubuque Jul 8 '14 at 18:34
  • 1
    $\begingroup$ I think the title does not ask the question you wanted to ask. "VLQ" is a term used for flags, it is not used within the review queue handling those flags. Reviewers are not asked "is this Very Low Quality [y/n]?". They are asked to click one of these buttons. Perhaps we need a flowchart for arriving at the correct button. $\endgroup$ – user147263 Jul 8 '14 at 18:38
  • $\begingroup$ @This: While not "Very", the review queue is called "Low Quality Posts". I am asking, what constitutes a low quality post. Is it just syntactical structure (or rather lack thereof), or do bad mathematical answers fit this description as well? $\endgroup$ – Asaf Karagila Jul 8 '14 at 20:26
  • $\begingroup$ Yes. My point is, your question appears to ask "when to flag as VLQ" rather than "how to review posts in the LQ queue". $\endgroup$ – user147263 Jul 8 '14 at 20:28
  • $\begingroup$ @This: I don't see the difference. If flagging is arbitrary, but reviews are specific then we overload the queue; if reviews should be arbitrary and flagging is specific then we are missing a lot of things which need reviewing. $\endgroup$ – Asaf Karagila Jul 8 '14 at 21:18
  • 3
    $\begingroup$ @Asaf Is this not a subjective question? One's definition of low-quality is subjective, and that is why to close a post, 5 close votes are required. If 5 users agree with you that the post must be closed, then it's closed, since that shows that the others in the community agree with your initial judgement of the post. $\endgroup$ – user122283 Jul 8 '14 at 23:55
  • 2
    $\begingroup$ @Sanath: This is a feature that different people use in a different way, and that may cause discrepancy in expectations. I am trying to bring this up so there can at least be a discussion, and that people will at least be aware of the topic. Also, unlike closing a question, the VLQ review queue is not quite the community but rather "first people on the scene". If I raised a flag that 99% of users would agree, but the first ones to review it were the last percent, then the flag is dismissed and gone. This is not quite like closing a question. $\endgroup$ – Asaf Karagila Jul 9 '14 at 3:41
  • $\begingroup$ @Asaf Thanks. However, isnt it very unlikely that the first people to review it were the last 1%? :) $\endgroup$ – user122283 Jul 9 '14 at 4:05
  • 1
    $\begingroup$ @Sanath: Since you don't really need that all the reviews would come from the 1%, just part of them, this is not as unlikely as you'd think. Moreover while questions get closed and sent to the close votes queue, answers get dismissed and forgotten (and if someone else raises the flag, it goes to the moderators, which are not referees and may dismiss it on that grounds). $\endgroup$ – Asaf Karagila Jul 9 '14 at 4:31
  • $\begingroup$ @Asaf Thanks again! $\endgroup$ – user122283 Jul 9 '14 at 4:32
  • 4
    $\begingroup$ @Sanath Suppose the rebellious 1% all haunt the VLQ queue (because they are militant), and further suppose that only 2% of users look at the VLQ review queue (I think this scenario is not too unlikely). Then it is a 50/50 chance if the action taken is actually the community consensus. $\endgroup$ – user1729 Jul 9 '14 at 7:00
  • 2
    $\begingroup$ en.wikiquote.org/wiki/Potter_Stewart $\endgroup$ – Will Jagy Jul 9 '14 at 21:36
10
$\begingroup$

IM(ns)HO, "very low quality" flags, if used at all, should be reserved for posts that are beyond redemption: those posts consisting of indecipherable nonsense that cannot be salvaged (yes, this is pretty much the SE definition). Similarly "not an answer" flags should not be used for simply incorrect answers, but for "answers" that make no attempt whatsoever to answer the posed question: commenting on another post, asking a new question, digressing on the eating habits of one's pets, etc.

Judging the mathematical content (or "helpfulness") of answers is what your votes are for.

Might this mean that incorrect (or misleading) answers remain on the site? Of course! But this can be viewed as a feature and not a bug. Correct (or helpful) answers should "rise to the top" in terms of score, and readers should treat highly downvoted (or simply negatively scored) answers suspiciously. In an ideal situation, there should be some commentary to the answer indicating what is wrong with it, which then provides additional useful information to the reader: why that particular line of reasoning is flawed or unfruitful. (Additionally, users who consistently post "unhelpful" answers can be rather quickly blocked from posting further answers.)

$\endgroup$
  • $\begingroup$ Post Scriptum. I will say that when "disputed low quality review" flags come through — and these are not uncommon — I try very hard to follow the SE definitions to decide what to do with the post in question. $\endgroup$ – user642796 Jul 9 '14 at 6:07
  • 2
    $\begingroup$ What would you say is the course of action when a well-known crank posts nonsense in answers, and asks soapboxing questions? I mean, besides the obvious downvotes and closure/deletion votes... $\endgroup$ – Asaf Karagila Jul 9 '14 at 6:43
  • 1
    $\begingroup$ @Asaf: Downvote, post your objections to the answer in comments, and don't get dragged into prolonged discussions which may make finding the objections more difficult. In particularly egregious situations, flagging for the mods to keep an eye on the situation isn't a bad route: we can annotate the account and keep an eye out for sockpuppets if/when the original account becomes blocked. (Incidentally, if I have the correct user in mind — and I am pretty sure I do — the crank that spurred this thread is currently blocked from posting answers, so... the system works!) $\endgroup$ – user642796 Jul 9 '14 at 8:35
  • 1
    $\begingroup$ Well, that's good for now. But I still think there should be zero tolerance for the likes of him, which have proved time and time again that they only use this website to "correct the mistakes of modern mathematics". But we're getting off topic here... $\endgroup$ – Asaf Karagila Jul 9 '14 at 13:38
  • 1
    $\begingroup$ I disagree that users who consistently post unhelpful answers are quickly blocked. I've seen a lot of them post many wrong answers and even indirectly admit in some places that they don't know how to justify their answers. Worse still, many of their blatantly wrong answers have positive votes. Also, many of them do not listen when others including myself tell them that their answers are wrong. I don't want to name them but I see no way of dealing with their wrong answers except by downvoting, but they feel that it is unfair. Some of them have rather high reputation as well (~2000)! What to do? $\endgroup$ – user21820 Dec 16 '14 at 16:03
8
$\begingroup$

My flowchart for clicking buttons in the LQ queue (which processes both Very Low Quality and Not an Answer flags). Nowadays most of the posts there are answers, so I begin with those:

For Answers

  1. The post is a misplaced chunk of text (a comment, another question, or some nonsense) $\implies$ Delete, leaving a comment when appropriate
  2. The post has no information other than a URL. The URL might be helpful, but SE is not a link farm, so $\implies$ Delete, leaving a comment.
  3. The post can be interpreted as an attempt to answer the question, and is property formatted $\implies$ Looks OK
  4. The post can be interpreted as an attempt to answer the question, but has formatting problems $\implies $ Edit

For Questions

  1. The post is Homework HALP PLZ, or fits another close reason $\implies $ Close
  2. The post is not 1, but has formatting problems, or has important information added in a comment $\implies$ Edit
  3. Otherwise $\implies $ Looks OK

Remark 1: If I can't figure out which of the above applies, there is Skip. But in this queue I rarely skip.

Remark 2: Some actions are not available from within the queue, e.g., voting or flagging. Also, the comments disappear while editing. For these reasons, I frequently open the post in another tab.

$\endgroup$
1
$\begingroup$

As a data point, I surveyed all the posts that I had flagged as VLQ in the past 850-odd days.

Most (2,3,6,8,9,10) were posts that were in Arthur Fischer's phrasing “indecipherable nonsense that cannot be salvaged”; I have marked these with (x).

Item 5 was a post that the author wanted to delete. (Possibly item 1 was in this class also; I don't know why I flagged it.)

Item 4 might have been a good answer, but was so badly formatted that I could not bear to look at it. (And still can't!) It is the only item on the list that has not been deleted.

Item 7 was an obnoxious non-answer to a troll question. I first flagged it as offensive; that flag was declined.

  1. https://math.stackexchange.com/q/144823
  2. https://math.stackexchange.com/q/150811 (x)
  3. https://math.stackexchange.com/a/153996 (x)
  4. https://math.stackexchange.com/a/177816
  5. https://math.stackexchange.com/q/211784
  6. https://math.stackexchange.com/a/214075 (x)
  7. https://math.stackexchange.com/a/340077
  8. https://math.stackexchange.com/a/590392 (x)
  9. https://math.stackexchange.com/q/618316 (x)
  10. https://math.stackexchange.com/a/630216 (x)
$\endgroup$
  • 1
    $\begingroup$ I'm tempted to flag this as Not an Answer, as it does not attempt to define proper use of VLQ... $\endgroup$ – user147263 Jul 9 '14 at 17:46
  • 4
    $\begingroup$ I suppose it's too late to follow your usual MO of downvoting the question so that the Community moderator will delete the whole thread. $\endgroup$ – MJD Jul 9 '14 at 17:58
1
$\begingroup$

To me, a very low quality question or answer is one that is incoherent, basically devoid of mathematical content, and has no redeeming features. Usually it is "short" (one or two lines), although it is possible to have a whole paragraph, or even page, of junk.

There are posts that are clearly "wrong," but where the user has obviously "tried," and contains mathematical content (or at least a reasonable facsimile of such). Those are not really "very low quality" and should be edited or commented on, in hopes of salvaging them.

$\endgroup$

You must log in to answer this question.

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