26

Yesterday, a question was asked. It wasn't a very good question, but it was a valid, cromulent question. I wouldn't downvote it, or vote to close it in a vacuum, but I also wouldn't upvote it either. It should also be noted that it doesn't have an upvoted or accepted answer.

This morning, another question was asked. This question was, essentially, the same question as the first. However, this new question is objectively better than the earlier question as well. It's more detailed, includes more information to provide context to the reader, better written, and broader in the best way possible. (Which is to say, it is broader by a small amount that allows for answers that are more broadly useful without losing any of the essence of the original question or otherwise losing focus.)

So, I did what one does when one encounters a duplicate question, and I voted to close Question One as a Duplicate of Question Two.

Whereupon, the inevitable happened.

@LessPop_MoreFizz actually since this one was asked first the other is a dupe of this one. – spartacus 14 mins ago

To which I say:

So What.

While it makes sense in the general case to point new questions as dupes to old questions, this is not a hard and fast rule. If the newer question is of higher quality (or, as sometimes happens when doing housekeeping on old dupes, has higher quality answers), there is nothing wrong with pointing the older question to the newer one as a duplicate. Our first concern when closing duplicates should be accuracy. Our second concern should be that the duplicate referred to (and left open) has good, high quality answers and that we aren't wasting peoples time by shuffling them between unanswered questions (a concern that is at least partly addressed by the requirement that dupes have upvoted answers.). Our third should be that the newly 'canonical' question is of high quality. Only after we have considered all of these should we fall back to chronology to determine which question gets redirected.

Join me in ending the tyranny of chronology! Stop insisting on sending dupes to the oldest question purely by virtue of its age!

  • 2
    I agree. The problem with being dependent on time in determining which question is the duplicate is that time has little to no effect on the quality of the questions at hand. Maybe the older question has more time to gather answers or edits, but we shouldn't decide solely on time. – Yuuki Jan 8 '14 at 13:59
  • 1
    It was never a hard rule to always close the newer question as a duplicate, but the occasions where this really matters are rather rare. In practice, this is also something that most of the time only moderators can do effectively as there is typically much more momentum behind the closure in the conventional direction. – Mad Scientist Jan 8 '14 at 14:18
  • @MadScientist agreed with mods usually being required to get this accomplished (unfortunately); if there's a case where a duplication should be reversed I'd recommend flagging it for a mod if it's already closed in the wrong direction. I've had to reverse a few on UX already – Ben Brocka Jan 8 '14 at 14:26
  • 6
    The quality of the 2 questions didn't differ by THAT much. Frankly I don't see how the other question is 'much higher quality' that the first one wasn't. It just needed some cleaning up via a helpful edit. – l I Jan 8 '14 at 14:43
  • 4
    Here's a relevant discussion: What is the policy on older questions rendered obsolete by newer ones?. There, the general consensus on policy appears to be to close a lower quality question as a duplicate of a higher quality question, regardless of time. – Schism Jan 8 '14 at 15:49
  • 2
    Personally, I think that if two questions have similar scope (they ask with the same amount of generality) and are not vastly different in quality (as in this case), then simply editing the older question would be better than asking a reworded version. – Schism Jan 8 '14 at 15:51
  • I feel like we have this discussion annually. As before, I am completely agreed. – Matthew Read Jan 8 '14 at 16:53
  • 8
    Let's move on to a discussion where we argue about how to determine if one question is "objectively better" than another! – Sterno Jan 8 '14 at 20:17
  • 2
    Not to mention at this time, I like the awnser on the closed question alot more then the awnser on the open one. – Lyrion Jan 9 '14 at 8:41
  • The more I'm looking at this, the less my comment about "objectively better" becomes a joke and the more I think the wrong one got closed on the basis of quality alone, age aside. – Sterno Jan 9 '14 at 11:28
  • to be fair, the closed question's answer was edited to be much better after this meta question started. Its also why I haven't edited the closed question to make it look better so it doesn't pollute the point less is trying to make. – l I Jan 9 '14 at 14:54
  • @spartacus Ah, that makes more sense. – Sterno Jan 9 '14 at 21:17
  • 2
    I'll throw out for the sake of argument, that @LessPop_MoreFizz 's argument seems to discourage editing. Those focused on rep (who are many) could simply scan for bad questions and reask them with a better write up. I'm not sure that is actually a bad thing (since we end up with better content), just pointing out that it could be an unexpected consequence of the policy. – EBongo Jan 14 '14 at 16:44
  • @EBongo In general, though, a question that's been around for awhile has some pretty darn good answers. That's what we're after, and just re-asking a question better doesn't get that advantage. So it would be closed as a duplicate of the original. If two question ask the same thing, and neither one has answers, then we might run into this problem, though. – Frank Jan 16 '14 at 16:08
  • @Frank Of course you are correct. I think the phenomenon Less presents most often happens when two questions are asked within a short time period of one another (as was the case in his example). This gives opportunity for the older question to either not have answers, or not have good ones. When the older question has good answers, I think there is general consensus to close the newer one (even if it is asked better). The problem I propose would present itself by users reviewing the question queue and asking new questions (rather than editing) when they see poorly asked questions. – EBongo Jan 17 '14 at 2:21
3

It's a bit unclear what the action item is here. There is a proposal in the question and two answers contesting it with none supporting it. Does that mean that no one supports the proposal in the question? Or that everyone does?

The goal of Stack Exchange is to give people the most helpful answers to their questions. Sometimes this is by getting answers to their questions. Sometimes this is by finding answers on another question. But regardless, we want to give them the best answers.

In picking a dupe target (the question to which we send people from the questions closed as duplicates), we should pick the one with the best answer (or answers). If the answers are equally good, we should pick the best question. If the questions and answers are equally good, then we should pick the oldest one.

  • I don't think @Kalina's answer was contesting the proposal in the question. It was actually agreeing with it and provides an objective criteria on which question should be the dupe and which should be the original. Without an objective criteria, and like Sterno commented in the question post: 'Let's move on to a discussion where we argue about how to determine if one question is "objectively better" than another!' – galacticninja Aug 1 at 4:45
1

In my opinion, in the instance where you've got multiple questions all asking a similar thing but none of them have got a definitively upvoted (2+ or more votes, bounties are automatically handed out from having 2 votes so 2 seems a reasonable number to me) - the first question to reach that criteria (either a definitively upvoted answer and/or an accepted answer) should be treated as the 'master' question and all others closed as a duplicate of that question.

There is no point in closing questions as a duplicate of another question that doesn't meet the criteria to be considered answered, as this doesn't help the people asking the questions and that's pretty much what we're here for.

When a question is asked is largely irrelevant from this point of view - it would be perfectly possible to have a month old question asking something, and then a brand new question with an upvoted answer that answers the question. In this instance we should not be closing the new and answered question as a dupe of a potential tumbleweed candidate that has been ignored.

Using your particular instance though, I'm going to make the following observations;

  • The title on the original question seems more helpful from a "finding the question via Google" perspective to me
  • The content in the latter question does seem better written and of a slightly higher quality
  • The answer on the original question actually answers the question "What is the recipe for Fiz", I'm a little disappointed it wasn't upvoted and this entire situation averted
  • The answer on the latter answers the question "What did Edward tell me?"

I'm personally not completely in agreement that these two questions are in fact duplicates of each other: they ask about the same part of the game but from different angles, and the answer to "What did Edward Tell Me?" does not answer "What is the recipe for Fiz?" and vice versa.

  • 2
    "There is no point in closing questions as a duplicate of another question that doesn't meet the criteria to be considered answered, as this doesn't help the people asking the questions and that's pretty much what we're here for." Leaving both open doesn't help anyone, either. All that helps people is getting them an answer... how we manage dupes has very little to do with that. Of course, what you state is the way SE now works, so my arguing about it is kind of pointless. :P – Sterno Jan 8 '14 at 21:24
  • 2
    The point is, one of them will eventually get an answer that meets said criteria, at that point, that question/answer becomes useful to everybody else, and we close all other instances as a dupe – kalina Jan 8 '14 at 21:26
  • 1
    Procedure as per Stack Exchange (from way back when the closing changes rolled out and moderators became required for "unanswered dupes", if you will) is to link the questions together in comments and then later close one, when possible. – user98085 Jan 8 '14 at 23:18
0

I disagree mainly for two reasons:

1) We may get users that post the same question on purpose but slightly better worded, or with better screenshots, even though they don't have the problem themselves, just to drive the traffic from the dupe to his question in an effort to game the system.

2) Deciding that a new question is better than an old question and deserves to be the "root" question is subjective and will create friction in the community. How will we handle this? We don't treat subjectivity well around here.

I still believe we should preserve the chronological priority, if the second question is really much better than the first, then that one can be edited and the missing information added to it.

You must log in to answer this question.

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