10

So I clicked "review". Then, my eyes popped out of a mixed feeling of surprise, shame and anger. There were 11 posts in the close vote queue! (And I had checked the queue a bit earlier)

First, just to be clear:

Close voting is essential for an SE, or any site that doesn't want to get flooded with crappy questions. It's simple:

If you feed help vampires, you'll attract them.

We need close votes if we want to attract experts and ultimately build a source of knowledge.

But . . . 

I ended up agreeing with only 3 of the 11 closures; and I think I acted rather harshly.

For instance, this question received a downvote and a close vote, presumably from the same person. So,

  • Is this question "primarily opinion-based"? Nope.
  • Is this question "unclear" or "too broad"? No.
  • Is this question "General Reference"? No way. They have proven they've looked up the dictionary.
  • Does the asker need to do more research or provide more context? Not at all. In my very honest opinion, the question above is more interesting than a lot of questions I've seen.

In fact, if memory helps, 9 out of those 11 questions had the "Needz moar contekst or research" close reason. Obviously I can't speak for the voter here, but let's note:

  • With enough research, any question is answerable.
  • A lot of times the inquirer didn't have the correct keywords to do some or any research.
  • We're here to help learners. Let's not forget our mission.
  • Overdoing meta or being too strict with quality standards backfires.

Of course there's a lot of learning to do to be a decent close voter, but there's an option: "Don't close the question."

Now, I'd like to see everyone

  1. Comment when they sense there is a problem with the question. Sure, explaining downvotes or close votes isn't required, just as same goes with upvotes. 1 But let's face it: Telling "your question sucks" without explaining why sucks. 2
  2. Not vote to close questions they're not sure are gonna get closed.
  3. Not vote to close questions they don't know the answer to.

Stay tuned in; we'll probably strike with more meta posts about the how and the why of closing questions, and will try to refine the system. Clearly (well it's clear to regular reviewers), we get unjustified close votes almost everyday; and for a major part lack of teaching is to blame.


1: Some concerned ELLers in fact think there's a problem with upvoting wrong answers on ELL. That's another issue.
2: A lot of people in SO downvote without telling why because a lot of revenge downvoting or cursing happens afterwards; but I'm rather sure that's not the case about ELL. At least, not yet.

  • 1
    As with other meta posts on close voting issues recently, I think it is important to keep in mind that there is a reason why 5 close votes are needed to close a question. And even then, 5 votes could reopen it. My point is, each of us is entitled to their interpretation of the question content. As long as we try to be diligent, and excluding serial close voting (hopefully moderators can check for and address this), I think the existing system is OK. – user3169 Aug 30 '15 at 17:58
  • The current system is fine, but it needs a lot of polishing. ELL hasn't grown big enough to have close vote robo-reviewers, but when it does, we're gonna be in so much trouble. Furthermore, FWIW, it's better not to have the first close vote than to waste the time of the close voter and the reviewer. And I don't have statistics, but I think I can safely assume ELL's unjustified close votes are way more than they should be, comparing with the size of the sites. – M.A.R. Aug 30 '15 at 18:01
  • 2
    I don't think we should discourage folks from voting to close questions that they aren't certain will actually be closed or voting to close questions that they don't know how to write and answer to. I think that everyone should vote according to what they believe is best for the site. If five other folks don't agree, it won't get closed. Maybe I shouldn't take those points literally, because I think what you're saying is to be careful that you're evaluating the right qualities of the question when you vote to close, but if I do take them literally I don't think it's good advice. – ColleenV parted ways Aug 31 '15 at 18:03
  • @Colleen there's always an alternative: Comment. They shouldn't take actions based on something they're only 50% sure about. – M.A.R. Aug 31 '15 at 18:06
  • I think there are too many questions that could be answered by one look at OALD or a basic grammar. But I've got the feeling that a lot of posters don't know what Oald or Longman DCE is and that they have never seen a real grammar and would not be able to use it. – rogermue Sep 8 '15 at 6:13
3

I'm editing my answer to bump this discussion, because I think we need to talk about it again. We just closed two questions as duplicates of each other. The first close votes happened around the same time, but the ones after that should have looked at the question that was proposed as a duplicate, and seen the comment that the other question was proposed as a duplicate of the question the reviewer was currently looking at.

We need to be more thoughtful about reviewing. Closing a question or allowing someone else to edit someone's post is a big deal and shouldn't be taken lightly. It causes newer users that aren't familiar with the SE model a lot of heartache because it seems harsher than it is at first. Don't just go along with the crowd - any one of us can miss details or misinterpret something a learner has written once in a while. It's much better if we bring our own eyes and understanding to the review queue than to assume that the folks that voted before use didn't make a mistake.

I think that close-voting is an issue for all sites with an active community. There are many questions to be reviewed, many folks pitching in to help ease the moderation load, and many diverse views on which questions are OK, which questions are salvageable, and which can't be fixed.

A contributing factor is lack of time - if I don't have time to edit a question to bring it on-topic, or to comment with some guidance on how to get the question re-opened, I don't touch the review queue. If I come across a question in the queue that is going to take more time than I have allotted, I skip it. It's better to let someone else with more time handle it than make a snap judgement.

I also have access to the tools that show recently closed question and recent reopen votes because of my reputation level. I go there before the review queue and see if there's any questions that I think could be reopened with some help.

I understand that this takes much more time than clicking through and choosing close or leave open (this is part of the reason I didn't run for a moderator position!). The standard of moderation I try to achieve takes a lot of time, so I wouldn't hold someone else to it, but I do think for new users especially, more effort should be made to help bring the question on-topic.

Related discussions

Wait a few minutes before closing new users' questions?
Less Closevoting, more Editing!
Closing questions from new users
Voting to close questions as off-topic: are we overusing it?
DO NOT FEED THE BEARS
Robo-reviewers on ELL

  • Totally agree. Some SEs have all-in-one FAQs for new users so that they won't need to link to different meta questions on important community consensus on a matter. The fact that the community will or won't like this idea, though, is another discussion. – M.A.R. Aug 31 '15 at 17:12
  • Technically, it shouldn't be possible to create circular dupes... SE should have protection measures in place to prevent them... meta.stackexchange.com/questions/263331/… – Catija May 17 '16 at 14:40
  • @Catija It seems like a tough problem to sort out if they manage to get into the review queue. It's not clear to me how the system would automatically handle it - how could you tell which would be a better "master" if both have the same vote count? – ColleenV parted ways May 17 '16 at 16:02
  • Unfortunately, it sounds like the "recommended path" is "whichever gets to 5 first, loses"... then the close votes on the other question would instantly age out... but it's better that one question is open than that neither are... also, this is something that can easily be flagged for a mod and reversed to make the dupe closure the other direction. – Catija May 17 '16 at 16:08
  • @Catija I think that I may have caused the edge case by voting to close, but not as a duplicate (they don't call me CrashMaster for nothing). Regardless, it's been reopened now. – ColleenV parted ways May 17 '16 at 16:28
3

My comment made me curious: How many closures are unjustified?

So I asked rene to write a SEDE query (Thanks rene!) to find out. The query gives you the links to all of the questions that received close votes but didn't get closed.

So I ran the query for 8 other sites than ELL; which, according to the SE list of sites and their info have some of their stats close to ELL: (Syntax: [name of the site]: [what stat it has close to ELL] ||| Sorting: Alphabetical)

Let's take a look at the results:


Sorted by x
*: Deleted questions and answers not included
Data accumulated from Stack Exchange Inc. list of sites; all stats are up to August 31st, 2015.
Traffic & QPD stats are measured as a median of the respective stat entry from Aug 17 to Aug 31.

Where x is the number of close votes that have aged away 1; and QPD means "Questions Per Day" (i.e. the number of questions the site gets everyday on average).

Observations: We do get lots of random close votes on questions that shouldn't be closed. ELL is the second youngest SE among those, yet it has the third place on aged away close votes. And first and second place are sites that are 5 years old and have very strict question standards.

We either have

  • A lack of teaching; or
  • A lack of unity or meta consensus; or
  • A lack of close voters.

It's definitely not the third one, as ELL currently has 391 users with more than 500 reputation (And 65 users with more than 3k reputation, in case the privilege threshold increases). Whether it's the first or the second, it is a problem we have to find and solve; because as the site grows 2 this particular lack will result in real closures as the votes will accumulate.


1: In big sites like SO, or in small or abandoned communities, this number couldn't be a representation of unjustified close votes as neglection of correct votes leads to aging away. While on sites like ELL (Where the question usually gets the attention of enough users)
# of aged away close votes ~ # of unjustified close votes

2: As the list of SE sites sorted by age suggests, ELL already has done a monumental job, and is the second to Magento's stats in sites with an age of 2y < age < 3y.

  • Great data. A couple of comments: 1. Please explain what "aged away" means. I think I know, but better to define it to be clear. and 2. The 500 reputation level for close voting might be too low. For reference, is this threshold similar for the other groups you listed? – user3169 Aug 30 '15 at 23:58
  • 1
    @User close votes "disappear" after 14 days if they take no effect. And, the threshold is the same in blender (because it's in beta) but it's 3000 reps in the other ones. This might be contributing to the "a lack of teaching" bullet point but I have to investigate it further, as I haven't seen people with less than 1 or even 2k rep actively participate in closing questions. Thanks; I didn't take that into account. – M.A.R. Aug 31 '15 at 0:05
  • I wonder if being a younger site actually makes it more likely that we will have a lot of close votes that don't come to a conclusion because our community is still growing. I think too that a lot of our posters are receptive to the guidance that we give them, so one or two initial close votes stop making sense after the question has been improved, but they never get retracted. Editing only triggers a re-open vote if the question was actually closed. – ColleenV parted ways Sep 4 '15 at 1:45

You must log in to answer this question.

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