What is meta? ×
Meta Stack Exchange is where users like you discuss bugs, features, and support issues that affect the software powering all 127 Stack Exchange communities.

Update: Found in the "Highest voted questions" list on the 10k tools "stats" tab today:

both with obvious duplicates of long standing. I've voted to close, but I doubt it will ever happen, not am I entirely sure it would be fair at this point.

But I think is points out very clearly that we're not getting enough policing.


Original:

On any given day, I can open the 10k tools on Stack Overflow, go to the close tab, and burn through all my close votes in an hour or so (hey, I take some time to compare proposed duplicates and search out better ones myself, and try to write politely encouraging comments to users I guess may be to be new or sensitive).

And it doesn't make a dent in the supply of bad questions that need to go.

I put to you a set of linked questions:

  1. Do we need to get more questions closed?
  2. If so, what can we do to encourage more closing?
  3. What do we need to do to maintain quality control while we do it?
  4. Are the cultural norms we bring to closing sufficiently professional? What can we do to create fewer hurt feeling even while we maintain discipline and focus on the site?

Some related questions:

  • Are we closing the right questions right now?
  • Are we closing too many good questions, and if so are they getting consistently re-opened?

Many related resources:

share|improve this question
2  
I remember the days when one could close with one click, no votes, just a personal decision... those were good ol' fun days –  jmfsg Feb 19 '10 at 14:57
    
If everybody would be as kind and encouraging like you dmckee. –  user152723 Nov 6 '10 at 16:19
add comment

8 Answers

Recently it was brought up that perhaps there should be more abilities opened up for users in the 5k-10k range. How about offering a very limited set of moderator tools, such as a simple list of questions that have close votes on them? 3k users already can close questions, so being able to find some of these faster might be beneficial.

Edit: one more idea: perhaps users with higher rep should get more close votes per day? 12 is not very many for a 10k user who can see a list of many many questions that should be closed. This could be even higher still for 20k+ users (since it was also brought up that perhaps there should be another tier beyond 10k).

share|improve this answer
4  
Considering the fact that the number of 10k users is increasing, I don't see why that should be done. –  alex Feb 3 '10 at 17:10
    
@pi: it appears to be the contention of the OP that it is not increasing fast enough. –  Ether Feb 3 '10 at 17:29
    
I thought about asking for 1 more close vote per 10k rep, but tentatively rejected the idea becuase it seemed like it would accentuate imbalance between popular and unpopular tags. –  dmckee Feb 3 '10 at 17:50
3  
@dmckee: something like 5 extra votes per 10k of rep sounds reasonable to me, actually. –  Ether Feb 3 '10 at 18:08
2  
@pi I think that there are enough close votes in the system right now, if more people were diligent about using them, but that is hard and (as shog says) unglamorous. –  dmckee Feb 3 '10 at 18:33
1  
@dmckee That's actually a good point; the problem isn't an insufficient number of people with the ability to close, it's an insufficient number of people willing to actually do so. I wonder how many 10kers actually check the close tab periodically and use it to close questions. Extra votes for people that regularly run out and have largely uncontested closes (i.e. the closed question wasn't reopened) might not be a bad idea –  Michael Mrozek Aug 9 '10 at 21:44
add comment

I stopped going over the "Close" list because of how quickly I exhausted my close votes. Give me more close votes for that list, and I'll start cleaning up again.

share|improve this answer
add comment

Are we closing too many good questions, and if so are they getting consistently re-opened?

I think there is a non-trivial set of questions which occur in variants. That is to say, the set of answers to question Q is not sufficient to answer Q', while Q' may be considered nearly identical to Q.

I would suggest allowing people to merge similar questions (the mechanics of which I am not certain of). I think this would eliminate the issue of hurting people's feelings or discouraging them and also the possibility of closing good questions.

share|improve this answer
1  
Merging is not a trivial action from the stories of mods. –  random Feb 3 '10 at 6:51
3  
merging is not hard, per se, it's that two EXACT EXACT duplicate questions are not as common as you might think. There are often subtle differences. –  Jeff Atwood Feb 3 '10 at 7:00
2  
These are the hard ones when considering a close due to "Exact Duplicate". It is rarely sufficient to read the titles and sometimes necessary to compare the answers. Which is a problem for me because I favor closing duplicates quickly: before people invest time and energy in writing duplicate responses. –  dmckee Feb 3 '10 at 7:01
2  
just becouse there was not a good anser last year when the question was asked, does not mean there is not a good answer now. I don't know how to cope with this... –  Ian Ringrose Feb 3 '10 at 9:58
    
What if we opened another close condition: "Child of other" and made it the responsibility of the OP to rewrite the question? –  Mechko Feb 3 '10 at 15:29
    
@Ian and dmckee Would it be fair to say that simply because an answer has been 'accepted' in a question, that it is not necessarily a good question? I have certainly accepted answers that aren't good because after a week they are the only answers that have come up, but it's not important enough to start a bounty. I feel that a lot of the questions marked as duplicates are actually duplicates of such questions. Perhaps we need another way to mark such answers. –  Mechko Feb 3 '10 at 16:55
1  
@mechko: don't mark them at all. Consider deleting your question and asking a different one, or editing it to add more information / clarify what you're asking. –  Shog9 Feb 3 '10 at 17:05
2  
For my self, I pay no attention the presence of absence of accepted answer when evaluating questions. Acceptance only means "this is the solution I used". –  dmckee Feb 3 '10 at 18:29
    
@mechko, I think it is more then that, the best tool for a even job can change as new versions come out etc, also has hardware gets faster, there are more options to how software is written. I think when a question was asked is part of the question. –  Ian Ringrose Feb 3 '10 at 23:02
1  
@Ian, sometimes new versions of software can change the answers, and over time the increasing power and standard hardware will change the balancing point for some trade-offs, but the vast majority of duplicates seem to be people who either didn't search first (most, I think), or failed to find what they were looking for (relatively few). Without a concrete event to change things, I consider time a very weak element of the question. –  dmckee Feb 3 '10 at 23:39
    
@dmckee Even if that is the case, wouldn't it be good to have a 'suggested to merge with' option when voting to close, which would then allow a user to either expand on existing questions or question existing answers? I know people claim that you should simply leave a comment, but even 2 weeks after a question has been closed, you rarely get people who aren't ASKING that question looking at it. Even a 'vote to expand' option for answered questions... –  Mechko Feb 4 '10 at 6:20
    
meta.stackexchange.com/questions/38210/… here's a coarse example of what I was talking about. The question asked is asking for slightly more information than the 'exact duplicate' that caused it to be closed. I don't think this warrants a new question, but it certainly doesn't help the poster-boy learn what he is after(pardon the pun). –  Mechko Feb 4 '10 at 6:25
add comment
  1. There are still well over 1000 open, "not-programming-related" questions. So either more closing is needed, or more editing.

  2. Closing is... effectively janitorial work. It's not glamorous, and shouldn't be. It's going to appeal to certain people who appreciate the site and see it as a way they can pitch in, and those are the people who should be doing it, not those looking for some reward. I think the best way to encourage them is simply to avoid discouraging them. Which goes hand in hand with...

  3. ...Maintaining quality control. Those of us who joined during or shortly after the Beta had an example to follow: there weren't nearly as many questions posted, and so when Atwood closed one of them it was easy to notice. It's much harder now: there are far more questions posted, and only the most egregious violations are closed while still visible on the first Questions page. So where's the example for new users to follow? I think making the list of recently-closed questions visible to users with 3K rep points could go a long way toward helping here, allowing new users to learn from example before jumping in themselves.

  4. We can leave comments explaining, politely, why a question is being closed. However, some people are still going to take this badly. I try to leave comments on questions posted by new users, but tend to consider those who've been around for a while to be willfully abusing the site and therefore usually avoid commenting. Sadly, there are people on SO who so hate the very idea of closing that they will act badly regardless.


FWIW... There was a time when I would take about an hour each evening and go through posts that I'd missed during the day looking for things that needed to be edited or closed. Like you, I found it easy to use up my daily allotment of close-votes. I've cut back on this practice lately, primarily because:

  1. The timezone I'm in causes end-of-day voting to use up my votes for the next day - meaning I have no close votes left if I see something during the day that should be closed quickly. (See also: Please let me know when I’m running low on close votes!)

  2. It doesn't feel useful anymore. I can't really explain this one... Donno, maybe the incessant whining got to me finally. Or maybe winter made me lazy.

share|improve this answer
6  
"It doesn't feel useful anymore." I hear you. That is the very frustration that motivated me in opening this topic. –  dmckee Feb 3 '10 at 18:32
    
I think a sufficient number of downvotes should be enough... We tend to ignore downvoted questions anyways –  Mechko Feb 4 '10 at 6:27
add comment

My thoughts, starting with my answers:

  1. Yes.
  2. Not sure, I'll say more later
  3. Quality here means closing most of the bad questions and few good ones, and getting most inappropriately closed questions reopened. To make that happen we need to have something like unto a consensus about what the close reasons mean and how they should be applied.
  4. Mostly. I only occasionally see signs of excessive emotional commitment to closing, which is good. One thing I have been seeing more of lately is multiple votes cast against a question without a single comment; often the question really is that bad, but I don't think this helps either the poster or newly eligible closers.

I have started trying to comment on my reasoning on more questions to educate both the perpetrator-of-the-bad-questions/victim-of-my-humorless-gestopo-like-rule-mongering and any close qualified users who haven't been exercising the power. But this is a slow, even glacial, way to make progress.

I am a bit wary of offering a badge, for fear of encouraging closing for its own sake. We, or at least I, want to close questions because they detract from the mission of the site, not just because those are the rules.

Small suggestions: we could, perhaps, benefit from a blog post on why the sites should sick close to their mandates and how closing contributed.

share|improve this answer
    
I agree with #4 –  Mechko Feb 4 '10 at 6:29
add comment

I think part of the problem is the separation of community users, and transient users. When communities user came to Stack Overflow et al. from reading Joel and Jeff's blogs, Reddit, etc. to form a community, and the transient user who comes typically from a search engine with a question they need / want the answer to right now, and for their problem.

The issue is (IMHO) that you don't know a priori whether a transient user will care about the community, and as such, its quality of a Q & A database, or whether they are purely interested in solving their own problem, with no plans to revisit the site(s) unless they have another question.

I think transforms the question into, how can we discourage purely selfish usage, while being useful to users who just discover the sites. I think the closing of questions from users who are willing to be community members, is a matter of learning, they may need to be corrected once (or twice), but do intend to improve and be good community citizens is a manageable amount.

I think the open question is how to encourage and/or manage better usage of the site by new (transient) users, who come seeking their own gratification (i.e. resolution of their problem) by may still be willing to become community members?

I don't know if the threshold or website work flow of a new user asking a question can be improved to encourage better usage by new users, and simultaneously discourage poor / lazy usage.

Perhaps the user must click through the top n number of "Possibly Related Questions" (as from the Related sidebar) before posting as a entry bar for users with rep of 1 asking their first question. Or maybe all users? That would explicitly target duplicate questions from lazy / new users hopefully.

The other option is to perhaps develop a grey-list of keywords that suggest the question is trying to be posted to most suitable site. I'd let the user manually override if they still feel their question is on the correct site, but if framed positively i.e. "You could expect more/quicker/better answers to this question on site XXX (or "elsewhere") as this site does not focus on keyword-1, keyword-2, keyword-n" from the matching keywords of grey-list.

As far as I can see it, encouraging users to feel like they have become members of an online community before asking questions, and possibly improving work-flow to stress the importance of prior research and suitability to improve the question asking process I think should be the number of questions suitable for closing.

share|improve this answer
    
Go ahead. Turn the problem back to fron and then analyze it. I dare ya! I **double** ***dog*** dare ya! –  dmckee Apr 11 '10 at 1:16
    
More seriously, I think this suggestion deserves it's own question: it looks like a good idea at first blush, but there are several issues that need to be thrashed out. (Rep limits, how to make it work with the Users don't read principle and so on.) –  dmckee Apr 11 '10 at 1:18
add comment

Ok, my opinions inline:

Do we need to get more questions closed?

Yes and I think sometimes we close questions incorrectly. Also, it needs to be easier to find relevant questions and some thought needs to be put into the long term common question approach - should we have a more wiki-style site that takes the output from SO so that users can "go there first?". I appreciate this would take a lot of effort to store effectively - I'm just saying perhaps some thought needs to be put in. It's kind of an unintended (very positive) consequence of SO. An example of this is a user who created a GBD primer after I posted something. Can we have something in the open-directory style too?

If so, what can we do to encourage more closing?

We don't need more closing - we do need more votes to close for higher rep users as you suggest. I propose an extra two votes to close for every 1K above 3K, so 3K=12 votes, 4K=14, 6K=16 etc. In addition, a vote to move to SF/SU should not count as a close vote and perhaps migration votes should be enabled at 2K rep?

What do we need to do to maintain quality control while we do it?

Can moderators veto / stop users voting to close/open? If so, that'd be good (so that a question can be either frozen open or closed - votes frozen too so users can't vent their frustration) so that continually fluctuating questions stop happening. I agree the suggestion of access to limited mod tools at 5K. I might even spread the mod tools (I don't know what they are) out between 5K and 25K to make these reputation marks more meaningful

Are the cultural norms we bring to closing sufficiently professional? What can we do to create fewer hurt feeling even while we maintain discipline and focus on the site?

Not really. I think we should have some form of system that informs the closer that a user is recently registered without having to look at their profile and a suggestion: please explain why you're doing this.

Are we closing the right questions right now?

I think so, mostly.

Are we closing too many good questions, and if so are they getting consistently re-opened?

Not sure.

share|improve this answer
add comment

I think the majority of people would agree that SO doesn't need to condone any more censorship.

share|improve this answer
19  
Closing questions is decidedly not censorship. Nor is deleting them. It's a simple rule: use your own resources to get your message out, and use Jeff and Joel's resources in accordance with the rules they set. –  dmckee Feb 3 '10 at 7:08
27  
I think the majority of people would agree that the majority of people don't agree with you. –  GEOCHET Feb 3 '10 at 17:06
1  
I agree with Evan and I'm clearly in the minority. I think I am a diehard inclusionist. –  P a u l Apr 10 '10 at 23:35
5  
@ P a u l: Even if you think stuff shouldn't be closed you don't agree with Evan. Trust me on this. No one agrees with Evan. –  dmckee Apr 12 '10 at 1:23
    
-- Hey I worked on the Mark II detector there, long time ago. Good times. –  P a u l Apr 12 '10 at 3:32
add comment

You must log in to answer this question.

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