What is meta? ×
Meta Stack Exchange is where users like you discuss bugs, features, and support issues that affect the software powering all 150 Stack Exchange communities.

I think it would be cool if a question would eventually open up to moderators to accept answers on the user's behalf, once a time period has passed. The asking user would get the rep score for the accept.

I know I've seen questions successfully answered with lots of upvotes and no accepted answer that sit there for months unaccepted. This would give superusers a little more functionality and would tie up a lot of loose ends.

What do you think?

share|improve this question
17  
I think you misunderstand the role of the moderators. They are solely here to ensure that interactions on the site is civil and handle disputes with users. There are only 2 SO moderators, how long will it take them to read and accept answers on 33000+ questions? –  BinaryMisfit Aug 27 '09 at 22:28
11  
I would consider myself quite unqualified to accept answers on a wide range of programming subjects. Sometimes I can't even decide if a question is programming related or not. –  Bill the Lizard Aug 27 '09 at 23:42
9  
Note that this question has not yet had an answer accepted ;) –  Default May 20 '10 at 13:04
6  
Nor will it ever. Think of it: this question will never, ever be answered. Pretty unsatisfying, huh? –  Chris McCall Jun 3 '10 at 21:06
1  
Mods don't need more work! –  studiohack Jan 19 '12 at 23:47
3  
What if the user left a comment that says thanks, but didn't flag it as an accepted answer because they don't know how to use the site? –  CharityAbbott Dec 28 '13 at 2:31
3  
It doesn't have to be moderators, It can be a review queue for trusted users over 20,000, and you need 5 of those to accept the same answer. or even users with high score in that specific tag like synonyms. –  i3arnon Jan 31 '14 at 14:56

10 Answers 10

Ended up here from a new question being duped to this one.

I welcome the idea to get questions marked as "answered" and have question authors indicate which answer helped them the most.

However it must be via the question author.

As defined in the Help Center, an accepted answer is specifically:

Accepting an answer is not meant to be a definitive and final statement indicating that the question has now been answered perfectly. It simply means that the author received an answer that worked for him or her personally

So an accepted answer:

  1. Does not mean it is the best answer
  2. Does not mean it will help others even with the same or similar question, as users have a vast array of code and approach differences between each other
  3. Is only a feature for the question author and only signifies a specific answer helped them above other answers
  4. Has no bearing at all on how the community in general perceives an accepted or other answers in a question. The accepted answer may have (e.g.) a score of 3, and another answer in the same question upvoted to a score of 20.

So users other than the question author can only mark what they think is the best answer based on the question and answer text and which answer directly answered the question. This would not necessarily be the answer which specifically helped the question author the most.

Only the question author knows which answer was the best answer which helped them. And this is the entire purpose of an accepted answer.

If we allow anyone other than question author to mark as "accepted" then we have literally changed to meaning and purpose of "accepted answer" - and that is a new proposal altogether and not one which should be hidden as an unintentional side-effect of a different proposal (as with this one here).

So no, we cannot let anyone other than the question author accept an answer, not even moderators.


If you want to clean up unanswered questions, you need an idea and feature-request where either:

(1):
You somehow get the question author to come back and accept.

Bearing in mind users may well forget which answer helped them the most and just accept the highest upvoted one.
More "rewards" to entice them back may get answers accepted just to earn the reward.

Neither option necessarily means question author will accept the one which "helped them the most", so all we'd do is "clean up" and not show genuine accepted answers.

(2):
Users other than the question author resolves this without marking an answer as "accepted".
Such as (just an example I'm not suggesting we do this):

Do not label a question as "unanswered" which:

  • Is older than X months
  • And has X total answers
  • And has X answers with a net score greater than Y

So for example a question:

  1. Is 8 months old
  2. Has 3 answers
  3. Has 2 answers with a net score greater than 5

This is not perfect, and has issues, it's just an example idea of a way we can clean up old "unanswered" questions which are arguably answered in some sense (have decent and upvoted answers).

That said, we do risk losing out on users potentially going through "unanswered" questions and providing an even better answer than the current ones.
The above algorithms could be tweaked to fix that, but it might just be the case that we cannot do anything about questions being "unanswered" for a long time or indefinitely.


Either way, it needs a new angle than allowing moderators (or anyone) to accept answers on the question author's behalf.

share|improve this answer

Why not add a "recommended" option?

Leave the accepted answer option alone

Add the ability for "experts" to recommend answers

Only somebody with a gold badge in one of the question tags would be considered an expert and could recommend, and their name would need to be shown

Obviously they cannot recommend their own answers

More than one expert can recommend answers and more than one answer can be recommended including the accepted one

The most recent recommendation would be listed first

A question with a recommended answer could come off the unaccepted answer queue

share|improve this answer
2  
We already have a 'recommended' option, it is called upvoting. –  Chenmunka Jul 10 at 12:54
    
@Chenmunka not exactly the same thing, and doesn't necessarily help with the issue that the OP chose a questionable solution for –  Glenn Teitelbaum Jul 10 at 13:46
1  
I think the core problem here is the artificial sort order that places accepted answers ahead of higher-voted ones. I think that should change, and would resolve most of the issues with lower-quality or wrong answers being accepted by the asker. The community would be able to override them with pure votes. As for highly-voted answers that are wrong or harmful, Tim's going to propose something about that: meta.stackoverflow.com/questions/298905/… –  Brad Larson Jul 10 at 16:15
    
@BradLarson basically Tim's considering the negative of my idea –  Glenn Teitelbaum Jul 10 at 19:16

What happens on SE sites:

  • Users don't accept clearly correct answers themselves, even when reminded to do so (causes: inactivity, don't know how to use site, not enough reward for them)

  • There already is an accepted answer, but now it's obsolete or not working. The now correct answer will (should) get more upvotes, but it will be listed after the accepted one (causes: better answer made by someone, the subject of answer get's updated/changed - usually software - making old answer obsolete)

People that visit such question will be visually presented with wrong answer or no answer at all, which IS a problem that can/needs-to be solved.

Solution:

  • Give high enough reputation (should be high - it's up to discussion) users the privilege to accept answers. Also voting could be the mechanism so the power is not on an individual.

  • The change of answer should be possible only after some time and/or only to the answer with more upvotes.

Counter arguments for some other answers here:

  • User oriented site argument - So let's leave everything up to users and remove privileges and mods altogether.. well no, it would be a mess. That's why we have mods and why high reputation users have privileges. Accepted answers which are an important part of every question is solely in hands of ordinary non-trusted users. This is not right, trusted users should have privilege over this to keep the site clean. The site will stay user oriented.

  • It won't put more work on mods. It will get them better tools to keep the site clean and organized. That's what mods like to do, why would they do it otherwise?

  • High reputation users are experts, when couple of them votes that an answer or another answer is better - than it is better. Same like they can vote that another question has duplicate answer that answers the question. It's the same thing. And if I am not sure about the subject I would not act - same as I would not vote on a duplicate question when I don't understand it. Very simple. An erroneous accepted answer would not be introduced but removed/corrected.

And now the big paradox:

What I can do already: Edit every letter in an accepted answer to totally rewrite it when I have edit privileges.

What no one other than OP can do: Touch which answer is marked accepted.

I would rather mark another answer as correct than to change/update answer, to leave it there. I don't know what the dilemma is and why we don't have this functionality long time ago..

share|improve this answer

No, this is a user-oriented site.

It is intended to answer people's questions. They know their problem the best and know what solves it for them. We need to leave it up to them to accept whatever answer they want. People will know what the community agrees with by the upvotes. The power to accept an answer should stay in the user/asker's hand. That way you have both opinions/views in a question.

share|improve this answer
28  
Let's check back in on this discussion when drive-by askers build up a huge library of unanswered questions with no mechanism for acceptance. At present, there is no way to reduce the number of unanswered questions. Eventually there will need to be some kind of cleanup process. –  Chris McCall Aug 27 '09 at 22:00
    
Maybe eventualy, but SO is the longest running of the trio and that has only been up for a little over a year. At this time, we shouldnt be forcing anyone to answer questions. The questions are still useful to people because of the up and down votes indicating what people agree with even if there is no "accepted answer". This may change when the site has a another year or two behind it, but for now, we should let it be. –  Troggy Aug 27 '09 at 22:11
3  
@Chris: The statistic of "number of questions without an accepted answer" is not a problem in itself; it's just an arbitrary metric. If there really is a problem of askers not finding acceptable answers to their questions, simply finding ways to reduce the metric won't fix the actual problem. –  ShreevatsaR Feb 13 '11 at 19:05
3  
It's not arbitrary, it's incredibly annoying to review my answers and see several that I spent a lot of time on that have not been accepted. There's no way to get in touch with the asker and no way for the answer to ever be accepted. It discourages me from answering questions, or at least spending a lot of time on my answers. This problem will eventually be solved, and the solution will be even more arbitrary and unsatisfying than having the answers accepted by a superuser. Mark my words. –  Chris McCall Feb 14 '11 at 17:27
1  
@ChrisMcCall: Why does it matter so much to you that your question is accepted or not? A question with unaccepted answers is still useful to the people who read it. The only time this might matter is if an answer had a lot of competing higher-voted answers for reasons that don't have to do with quality (those answers were first and not enough people have looked at the question to upvote the better answer). –  Nicol Bolas Dec 8 '11 at 9:06
3  
@Nicol because I crave acceptance due to family problems –  Chris McCall Dec 8 '11 at 18:14
    
How about this question, which clearly states the only given answer was correct, yet it was not accepted? –  Gaffi Mar 31 '12 at 3:57
2  
@ChrisMcCall, three years later and the prophecy has come true. –  Adnan Dec 7 '12 at 17:00
2  
I'd be interested to know whether there is any steer on this a few years later. Currently there are 179,500 unanswered questions and I imagine a large number will be ones just left. I find it frustrating when I put in effort to answer a question, the community believes it's right but I miss out on some rep/a badge because a user doesn't care about SO and just wants to solve their problem ASAP. –  Ian Aug 2 '13 at 19:10
    
@Adnan it's not a prophecy if I specifically will it to happen :) –  Chris McCall Jan 20 '14 at 18:09
1  
I disagree with the first sentence: On SO, the emphasis is on a post, not a user. –  Bleeding Fingers Mar 15 '14 at 17:58
    
Chris, I have learned to be content with the upvotes. Drive-by questioners are in the cards, and SE is specifically not organized to exclude them. The acceptance checkmark still provides useful information, that shouldn't be diluted by a moderator actin in loco questioner. –  Walter Mitty Jul 10 at 11:29

I don't agree

Even though there is not an accepted answer, the most voted one will float to the top.

There were a few features added to encourage accepting answers, so I don't think something like this will be implemented (people are going to want to raise that accepted ratio eventually)

share|improve this answer

There are corner cases where this may be a good idea. Maybe the number of times this is true is too small to worry about, but I'll put it out there for others to consider.

For example, this answer:

Android: Failure delivering result ResultInfo{who=null, request=1, result=-1, data=Intent......

Solution:

Draw the bitmap during SurfaceChanged, using a flag whether to load image.

As of the time of my post, the OP answered his own question and clearly identified that it was the desired solution in text, but they neglected (forgot or due to lack of knowledge) to mark it as the accepted answer. It also could have been someone else's answer and the OP could have indicated its the desired answer in the comments.

[[ In this particular case, the question is new - the OP is still around and may remember to accept the answer at some point, but its not inconceivable that the OP could never be seen again. ]]

I think it would be nice for the community to flag situations like this in older questions, so a mod (or diamond mod) could review these "hanging chads" case-by-case and reflect the user's intent by accepting the answer on the user's behalf. The mods are simply converting a non-standard acceptance of an answer (by the OP stating in text that a given answer was the correct one) to the standard accepted answer format.


Another example:

[SOLVED BY AWN] Mysql group_contact with conditions

The OP retitled the question to: [SOLVED BY AWN] Mysql group_contact with conditions by did not accept the answer (as of the time of this edit). Again, the intent is clear.

share|improve this answer
    
Is the number of times really to small to worry? I vote your question up to suggest that it is not that small. I sometimes search a given tag with hasaccepted:0 to see what problems are not yet solved. And I find the situations you described quite frequent among these "unanswered" questions. If the questions were closed with a reason "self-answered" instead of being accepted it would satisfy me too. –  jarekczek Sep 20 '12 at 8:06

The moderators are in no way product experts. We were chosen to oversee the community and ensure that the users have someone to turn to for assistance. We are only here to deal with disputes and ensure the sites remain civil.

Although most of the moderators know the subject core of their respective sites well, none of them will necessarily know what the right answer is, even if it is voted. Furthermore it adds more work for the moderators to do.

For example. There is currently 33000+ questions without accepted answers on SO, how long will it take the 2 SO moderators to read and accept everyone of these questions?

I don't think this will work.

share|improve this answer
    
I know your post is almost 6y old, but I just don't agree with "The moderators are in no way product experts.". If the author's comment below answer says "It worked. Thanks", you don't need to be software expert to know this answer was answered. You just need to know english a little bit. And there could be a comment flag - "This answer was accepted", so moderators will not receive millions of accept request. Very few I'd say. –  voytek Jun 22 at 17:41

The only thing that annoys me more than no documentation is incorrect documentation.

Since these are Q&A sites we're talking about, if I go searching for an issue and don't find anything then I'll just keep looking or ask the question myself. But if I were to search and find a question on StackOverflow with an accepted answer and plenty of upvotes, and that answer doesn't actually resolve the problem, I wouldn't be pleased. Rather than moving on, most people would now be wondering why a known solution is failing to resolve their problem.

The best way around this would probably be to leave a note saying that the question was forcibly accepted by a moderator, and may or may not work. In which case, what is the point of even marking it as accepted?

As far as your concerns about there being a huge collection of unanswered questions, I'd rather the questions be left unresolved than misleadingly marked as resolved.

The focus should be on encouraging users to resolve their own questions than finding ways to do it for them.

share|improve this answer
    
"In which case, what is the point of even marking it as accepted?" To get it out of the queue of Unanswered Questions, so that actual unanswered questions get the attention they deserve. –  Chris McCall Aug 27 '09 at 22:06
    
If thats the case, then I'd prefer a new classification be added to the effect of Abandoned questions. Then they won't be incorrectly marked as resolved and they'll clear the way for active questions. –  Brandon Aug 27 '09 at 22:09
    
*and of course, abandoned questions wouldn't be shown in the unanswered section. –  Brandon Aug 27 '09 at 22:11
3  
I agree. No information is better than bad information. –  Troggy Aug 27 '09 at 22:48

The discussion of someone other than the OP has come up many times, and each time it's been met with general disdain. The prevailing thought is that only the OP should be the one to decide whether or not the answer met the expectations / qualifications and as such should be the sole judge in awarding the acceptance check. The community gets to vote and that is enough.

share|improve this answer

What assurance do we have that the moderator knows the correct answer? If they're just going to select the most upvoted answer, we could automate that answer. That would be just as inaccurate, and save the moderators the work.

Moderators should be there to handle the exceptional, not something so mundane. Think of the work load you'll add.

share|improve this answer
    
Moderators enjoy more trust than the community at large, I feel they'd be qualified to choose the correct answer. As for work load, it's optional, not a responsibility. –  Chris McCall Aug 27 '09 at 21:49
5  
Unless the moderator has the exact same problem as the user who asked the question, they aren't qualified to choose the correct answer. They're just qualified to choose an answer that may be correct. –  Brandon Aug 27 '09 at 21:53
    
Couldn't we leave it up to their judgment? Wouldn't that be better than the eventual alternative: a very large number of unanswered questions that only increases? –  Chris McCall Aug 27 '09 at 21:58

You must log in to answer this question.

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