7

TL;DR: Please, decide a policy, adhere to it and enforce it.

The main problem here is the lack of cohesion about how to deal with the questions. For that I rant about the following:


Yeah, guys, we reached that point (I'm referring to the title of the question "Should we consider unanswered questions a problem yet?", not the content itself). It is a problem. Seems that some people haven't struck home that questions which are low quality need to be downvoted (some may argue that it is optional, but the very same people shouldn't rant about the low quality questions or that the answer rate is so low). Remember that the quality question blocks only works if you downvote/close questions.

What inspired me to make this meta question was this question, which, at the moment that I saw it, had 3 comments each of them with 2 upvotes, so at least 5 (could be more) persons considered that the question was lacking crucial information without a single downvote. The answer is pretty easy, and if the OP had added the relevant information in the first place, it would be highly probable to be closed as duplicated of one of many questions about "Permission denied" errors.

So, since all my rant won't make you feel at easy while downvoting here is the other method; vote to close:

  • If the question don't have CRITICAL/BASIC information about the problem, vote "Unclear what you are asking" and leave a comment asking for the information.
  • If you're looking at the First post review queue and you are almost sure that there should be another question with the same content, look for it and flag it as duplicate. But, if you didn't find the duplicated or you don't remember, either, ask in [chat] or skip that question.
  • If the question is vague enough to make you think "What is this guy talking about" flag it as "Unclear what you are asking" and add a comment. If you don't want to leave the comment, don't press the "No action is needed"!. The only fact that you doubt, makes you unable to take a definitive action, hence Skip the question. If you hit "No action is needed" you are hurting the community!

Now, I will use my super powers to predict some questions:

Are you out of your mind!? You don't know that the Close queue is a mess just with the normal workload? now you want to add that?!

Yes. I'm out of my mind, because I'm using something for what it is supposed to be used. There are enough >3K reputation (active?) users to take care of all the Close Votes needed.

Why flag to close immediately? The comment should be enough no?

Yeah, it should be enough in a pink world with ponies and world peace. But, in the real and crude world, where the users bail out after leaving all the crap behind we need real men/women to take the crap out to react accordingly. If the user is kind enough to provide the information, the question should pass to the Reopen queue, and be approved almost immediately (in fact, that queue is lazy enough, so it needs a bit of work ;)).

Again, aren't we a bit hasty?

Nope. Out 186 question a day, 160 are made by new users, 140 needs more information, 20 gets more information, the rest is crap (this aren't actual values/statistics). Why no deal with the crap while it's still fresh (sounds dirty, no?) instead leaving it get into our backlog and getting bite when we want to take care of it, or we don't want the bother to take care of it now that the problem is serious.

Lets put it in a nice saying:

What is the difference between being hasty and acting quickly? The answer is incredibly simple. It is decided by whether the outcome is successful or not. I took this from a Light Novel, but seems pretty awesome quote.

We are not doing the work as we are supposed to, what is the outcome? We had a clean-up week, where are the results? We have nagged enough, and the results aren't satisfactory.

Stop ranting and make those flags/votes fly! We know how to be kind but strict. Show the ropes to the users, some of them care enough to use them. And improve the quality of AskUbuntu!

Eliah got a glimpse of what I'm trying to do. But let me state clearly: while I made a link to Should we consider unanswered questions a problem yet? the main problem that I want to asset is to keep the crap out of our systems influx of low quality question to the minimum. Which this, I plan to accomplish several (ambitious) objectives:

  • Attract new professional answerers.
  • Get the ones that stepped down, due the quality of the questions, back.
  • Make of AskUbuntu a site where you don't go to ask how to resolve problems, but to find how you solve them.
  • Make of AskUbuntu a reference site for the global community.
  • That new users will see this as efficient that they will like take part of it, saving the recent view of Ubuntu in general. (yeah, the answer rate help with this)
  • Don't commit the same error as UseNet.

There is an advertisement that I remembered while reading jrg♦ comment below that says:

Help us to help you.

That's the message that our actions should deliver to every AU user, and I believe that it wasn't loud enough.

More rant about Eliah answer:

we are not a factory for minimizing unanswered questions

You are absolutely right! What we are is a factory of high quality questions/answers.

Quoting all the "unanswered rate" phrases and sentences

Oddly enough I didn't mention the word "unanswered" in my first revision, and "rate" just once in allusion to the people that keeps nagging about rating but never propose any solution. The only thing I could use the rating for is to do not make the same mistakes of the past. In my sociology class, I learned that: "it has the same steps as the scientific method, except experimentation, and the laboratory for the sociologies is the history." What I learned from the history? What are my solutions so mistakes of the past aren't going to hunt us? The ones I wrote above.

1. We're not providing enough good answers to enough of the good, answerable questions being asked.

The first problem can be solved by more people writing answers, more appropriate canonical questions, and better ways of figuring out when questions are duplicates.

I try to navigate in all the crap that get to our seas:

I produce about 2-4 answers per day. My unscored rate, since we need numbers here, is about 23.8 percent. Only 10 unscored answers of ~165 (darn query don't say how many answers is the base) which is a pretty fantastic number *eats a cookie*. Then I leave 38 comments (yesterday) for about 20 odd questions. I normally use my 40-30 up/down votes/day I have, but I couldn't use all my flags (I think I flag 10-15 answers/questions day). I don't know others, but I give closure (makes the question answerable, flag, answer) to an average of 10 questions/day *pats his own head*. And I still do things in RL. The questions I normally answer are the very same I get from the review queue (any review queue).

I wonder how others manage to come across questions that they might have a godly answer between all the noise, when I still rant about this problem, and seems to be efficient. I would love to write more canonical questions/answers, but since Unity + USC + Gwibber + other oddities + my system = problem, I have to keep myself away of Ubuntu. BTW, remember that great questions makes great answers.

Users often post badly formulated or unanswerable questions, then expect they will get good answers.

The second problem can be solved by constructive reviewing, commenting, upvoting and downvoting, editing, closing, and reopening.

Lets teach them that's not our philosophy with the tools we have at hand. While I can edit bad grammar of some users, if it still lacks something that I have to comment, I put edit aside, favorite the question (or check my own revision history), leave the comment then if he comes by and fix his question, then I fix his grammar, if it hasn't be fixed yet (since I'm taking 2 reputation of another user, I might leave them). Of course, this doesn't mean that I don't follow the question... I take care of each of the ~15 questions I review a day. I also keep myself updated of each answer that other users make to this questions, and this is what takes most of my time here.

Closing more questions is valuable, but it would only contribute slightly to solving the first problem. We should close more questions, but not to improve some made-up number. Our goal is to do better, not to look like we're doing better.

Totally agree.

3. When a question has a good answer, that often means we should not close it as unclear. When considering whether or not to close a question, how many answers it has and what they say really is sometimes relevant.

I have some reserves about this.

Let's take extra care to make sure we're not misreading hard questions as vague.

Actually those are the ones I like.

7
  • 1
    I think you are preaching to the choir. :) Aug 26, 2013 at 16:57
  • 1
    "this aren't actual values/statistics" - this is stackexchange, we demand science! Aug 26, 2013 at 17:05
  • 1
    Excellent, excellent post. I think you're helped his the issue on the head. People aren't using their super powers enough, or are using them wrongly. I think half the battle is just education. Users either don't know about the tools they have, or they don't know how to use them properly. The system only works when all the features are used properly.
    – Seth
    Aug 26, 2013 at 17:40
  • 1
    There was this sometime ago: meta.askubuntu.com/questions/6833/… and I liked one answer there because it was "different": meta.askubuntu.com/a/6849/25656. With Ask Ubuntu being pushed at people, there is going to be a decline in quality relative to other SE sites which retain their quality standards.
    – user25656
    Aug 26, 2013 at 17:49
  • @vasa1 I clicked the link before you edited your comment, so I read the entire question/comments/answers and was about to rant that Luis Alvarado's answer is too lenient, that it doesn't work anymore, etc... then to give you another rant about why I like Tanel Mae's answer... seems that you got safe of my full rant/preaching mode ;)
    – Braiam
    Aug 26, 2013 at 18:39
  • @JorgeCastro 1) I hope in a good way; 2) I'm fighting with the darn query that says "invalid syntax near 'where'" and my reaction is "which where? there are 20!", so that have to wait :(.
    – Braiam
    Aug 26, 2013 at 18:41
  • " I think half the battle is just education" quoting Seth above. As a short time user I've begun attempting to answer questions as a method of keeping my technical skills sharp. I don't always find the policy answers I need in the help. I take my best shot at it and when I'm wrong or make a mistake I appreciate being corrected as all knowledge is power IMHO. Using it properly takes time and practice
    – Elder Geek
    May 5, 2014 at 16:53

1 Answer 1

6

Questions that are not clear enough to plausibly be answered should be closed. And there is no reason to wait to close them. After all, closed questions that are edited are automatically added to the reopen queue, and these days the system even allows us to take back our close votes.

So the core thesis behind this proposal is a good one. It is simply that we should use the review and closure systems as they are intended to be used.

However, I think there are some potential problems with the details of this proposal.

First of all, we are not a factory for minimizing unanswered questions. The rate of answered questions is defined as the number of open questions with at least one answer of strictly positive score or marked as accepted, divided by the number of open questions. That is one of many possible measures of how well we're doing. A low answer rate indicates a problem, but it is not a problem in and of itself.

Our lowering answer rate points to two interrelated actual problems:

  1. We're not providing enough good answers to enough of the good, answerable questions being asked.
  2. Users often post badly formulated or unanswerable questions, then expect they will get good answers.

The first problem can be solved by more people writing answers, more appropriate canonical questions, and better ways of figuring out when questions are duplicates.

The second problem can be solved by constructive reviewing, commenting, upvoting and downvoting, editing, closing, and reopening. If a question cannot be answered, we should not string the asker along; we should instead help them improve it. This includes, but is by no means limited to, closing questions pending improvement and efficiently reopening those that have been fixed.

Closing more questions is valuable, but it would only contribute slightly to solving the first problem. We should close more questions, but not to improve some made-up number. Our goal is to do better, not to look like we're doing better.

So:

  1. Let's take extra care to make sure we're not misreading hard questions as vague.
  2. When a question would benefit from being more detailed or concrete but appears answerable, it should generally not be closed.
  3. When a question has a good answer, that often means we should not close it as unclear. When considering whether or not to close a question, how many answers it has and what they say really is sometimes relevant.
  4. It is critically essential that we not pretend an increase in answer rate from closing more questions somehow means we're getting better at helping people who ask answerable questions. If we fall into that trap, then any gain we get by closing more questions will be more than made up for by our own complacency.

I have a bit more to say about situation 2.

Whether or not even a well-asked question makes sense often depends in part on the technical proficiency of the reviewer. Furthermore, a question's meaning may be somewhat vague but clear enough that it could generate answers even without editing or more info.

If a question has elements of vagueness but a reviewer is reasonably confident that the question is answerable in its present form, the reviewer should often click "No action needed." A reviewer is not hurting the community in that situation.

In conclusion, the central thesis of this meta question is totally correct. But it is important:

  • that we not take things too far, and close answerable questions that are merely bad or ugly. (This already sometimes happens, so we must be on our guard.) And
  • that we not assume closing more questions will solve the bulk of our problems.
6
  • 1
    We are in the business of trying to get people to give us enough info to let us help them.
    – jrg
    Aug 26, 2013 at 19:02
  • I never mentioned the Answered rate... none of the ratings actually, but a general problem in the policy about the quality of the questions.
    – Braiam
    Aug 26, 2013 at 19:21
  • @Braiam The very beginning of your post cites this post, which is almost entirely about answer rate as the problem that you are trying to solve. Aug 26, 2013 at 19:43
  • I think most of your answer belongs in jokerdino's question... @Braiam wasn't really talking about the answer rate (joker was), he is talking about the lack of action taken or the wrong action taken. It's more of a PSA.
    – Seth
    Aug 27, 2013 at 2:54
  • @Seth The bulk of this answer is specifically about closure and responding directly to what Braiam has said. My answer belongs here, makes sense here, and only a small part of it would make sense on jokerdino's question. Both the beginning paragraph and the second two thirds of this answer are exclusively directed toward this question. The extensive new material in Braiam's question to address what I've said here would also make it strange for me to move this elsewhere. Aug 27, 2013 at 10:44
  • Well if that's what you wrote it for OK.. However you seem to be mostly worried/discussing an answer rate which Braiam never mentions, so I don't really see where your answer fits (probably just me..)
    – Seth
    Aug 27, 2013 at 14:46

You must log in to answer this question.

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