66

Consider the following theoretical situation:

New user asks a question:

Why does my program process a sorted array faster than an unsorted array?

I see this question, I can:

  1. Vote to close as duplicate
  2. Answer the question with a link to the duplicate, with possibly a little bit of extra explanation to their exact scenario

In the first case, I get rewarded with nothing.

In the second case, I will probably get an accepted checkmark and 15 rep (if I am FGITW), and the asker will get two rep for accepting my answer. We are both encouraged to leave this duplicate open!

Therefore, I propose that we make the rewards for these two scenarios equal.

Now that we have a new user interface that allows people to close their own questions as duplicate, we should have the user verifying that their question is a duplicate provide the exact same rep as if they had accepted an answer that said the same thing.

Some caveats:

  1. To prevent people from posting duplicate questions on purpose to gain rep, we would not award the +2 for an accept, just the +15 accepted answer rep to the duplicate finder.
    • It seems the consensus in the answers is that giving this +2 rep is not only okay, but preferred. FWIW I am okay with this.
  2. We would only award the +15 rep if the asker themselves clicks the "that solved my problem!" button. If it's closed as duplicate from 5 CVs, then no rep would be assigned.
    • This basically means that the person marking it as a duplicate is now encouraged to explain why the duplicate question is a duplicate in the comments, because they must get the asker to believe their question is now answered. In my opinion, this is a good thing.
    • This is a core part of this proposal. If this were not part of it, I wouldn't support the idea myself.
    • This piece of the proposal is also why this question is not, itself, a duplicate of this or this or this.
  • 1
    Related: meta.stackexchange.com/questions/260898/… Also undownvoted. We really need more information. – Deer Hunter Jul 17 '15 at 17:00
  • 14
    I gotta say, I like this idea. Needs a bit of debate, but seems far less awkward than most of the previous suggestions. – Shog9 Jul 17 '15 at 17:23
  • 1
    How would this tie in with gold badge closures? For the tags I have a gold badge in, my choices are to instaclose or leave a comment saying that I think it is a duplicate. For newer users that don't understand the rules, they'll often visit the linked question and copy/paste it as an answer to their own question. – cimmanon Jul 17 '15 at 18:08
  • 1
    @cimmanon per my reading award goes only to those who voted (and flagged?), commenters get nothing (otherwise people would start abusing it by posting dozens comments pointing to different questions in hope to guess one right). I think this means dupehammers are outta this game. Can't tell if this is fair or not, although given that gold badge indicates expectation that finding dupes is easier than for other users (based on profound knowledge of the tag) I am rather inclined to consider this okay – gnat Jul 17 '15 at 18:47
  • It isn't fully clear if you want the "duplicate solves my problem" flag replaces the accepted answer flag. I'm assuming so, since this (usually) moves the incentive for the answerer from "answer duplicate - gain rep" to "mark as duplicate, likely gain more rep". – psr Jul 17 '15 at 19:41
  • 2
    consider editing to account for the case if question gets reopened. For duplicates closed by asker this is likely a minor corner case but still... – gnat Jul 17 '15 at 20:43
  • @gnat Consider proposing what should be done in that situation in your own answer – durron597 Jul 17 '15 at 20:48
  • I would do but couldn't figure it. You see, this is rather weird case - asker says it's a dupe, it solved my problem - and... five 3K users suddenly pop up outta thin air and say no it's not ("it doesn't really solve your problem"???). Having difficulty imagining how to handle this :) – gnat Jul 17 '15 at 20:53
  • @gnat Choosing community duplicate could lock the question in the same way migrated questions get locked. But I'm just spitballing I agree with you it's a pretty weird case – durron597 Jul 17 '15 at 20:55
  • hm locking could solve this, yes - but it could add a different issue - question would become harder to maintain. No edits, no retags? not good - given how unlikely is the case we try to address, it feels like killing gnats with a cannon. Lock would apply to all self-closed dupes, while probability of reopening these may be around 0.01% – gnat Jul 17 '15 at 21:00
  • 3
    We don't allow anyone to override the OP's choice of normal accepted answers, @gnat; just let the dupe finder keep the rep unless and until the asker accepts another answer/duplicate. – Josh Caswell Jul 18 '15 at 15:40
  • @JoshCaswell that's fantastic, thanks! this reasoning really makes it clear how system should behave (implementing it technically may be not quite as straightforward as we handle regular accepts but that's secondary). Since you did it so well, :) mind trying to crack yet another nut - how could dupehammers integrate in this picture? – gnat Jul 18 '15 at 15:46
  • ...oh and I think I figured "straightforward" technical implementation for this corner case. System should just automatically raise mod attention flag when such a reopen happens ("DisputedCommunityDuplicate"). Since we expect this to be rare, exceptional case, resolving it sounds just the kind of job for moderators. @durron597 consider editing this in if this reasoning makes sense – gnat Jul 18 '15 at 15:54
  • 1
    That's a harder one, @gnat, but if the "stub answer" is included in the way this is implemented, then that would probably be the way for the OP to approve the hammer action. – Josh Caswell Jul 18 '15 at 15:55
  • 1
    @DeerHunter This question is beyond the scope of this MSE question. If you'd like to chat about it, you can find me in the SO Close Vote Reviewers chatroom – durron597 Jul 22 '15 at 14:43
25

I'm in agreement with Shog9 that this seems like a reasonable suggestion. Duplicate questions bug me... a lot, so much that during my interviews with SE it was one of the things I raised as something I'd like to see fixed or improved. I'm putting some thoughts into an answer to get some feedback and maybe flesh this idea out a bit more.

Your proposal is very close to something I have been kicking around for a little while now. The things I like about it are:

  • it rewards users to find duplicates
  • we can potentially still give something the user who asked the duplicate question (with the +2 accept), many times these are new users who are trying to contribute to the site and gain some rep - if they accept the other question is the solution, then I think rewarding them with something is good
  • we'll have more signposts to the best question and answers

I also think there is an additional side effect that may result from this, in that, it might alleviate some of the frustration users feel when people answer duplicates. Minimizing even a little frustration is a good thing... right?

I do see a few potential problems that need to be hammered out on this.

  • What happens if the question gets re-opened? Does the user who benefited from the original duping of the question lose the rep? I'd say yes to minimize gaming on something like this.
  • Does the OP who asked the question also lose the rep awarded for validating the dup if it gets reopened?
  • How will this work with gold-badge holders? We still want them to close obvious duplicates but there will not be a UI that is created for the OP to accept it due to the single vote. Excluding a large group of users with significant knowledge in a tag seems detrimental and might result in gold-badge holders answering questions instead of closing. Maybe, they get some reputation if the question doesn't get reopened after a certain number of days... not sure yet though.

I'm worried about potential abuse that I haven't yet thought of but I'm going to chew on this a bit more. I'd really like to see something implemented to better handle the duplicate question problem. I am going to this because it's high time we do something about this.

  • 2
    Thanks for your excellent input! My stub answer idea addresses the "what if it gets reopened" problem, because then everything gets treated within the regular system. As far as gold badge holders, I've been thinking about this all day... I don't think it matters too much if their incentives are wrong because they already have so much rep to have gotten a gold badge (I have 13k in basically one tag and I still don't have a gold badge) they probably won't care. And there just aren't that many of them. – durron597 Jul 17 '15 at 22:32
  • 2
    @durron597 I also like the stub answer idea. I really like this proposal it's really similar to what I've been trying to work out in my head. There are still a few kinks that need to be figured out and I need to think a bit more on how to prevent gaming but I'm hoping this would be turned into something. – Taryn Jul 17 '15 at 22:40
  • As part of the review, it might be useful to have stats as to how often an asker actually participates in a duplicate closure. My gut tells me this will only be a small fraction of duplicate closures, as these closures either happen by the community before the asker has come back (particularly with gold badge closures) or the asker tends to be upset with having their question be closed. Even if this proposal is put in place, I wonder how many questions will be impacted by it. – Brad Larson Jul 18 '15 at 1:00
  • 2
    @BradLarson you need to account for that additional incentives will likely increase amount of dupe votes / flags. Users currently motivated to only answer will get an option to gain from proposing dupes. Who knows how many jump that train. Current stats can't tell us – gnat Jul 18 '15 at 1:05
  • 7
    Roughly 15% of the duplicates closed in the past 30 days (1186 posts) were triggered by asker confirmation, @Brad. Not gonna change the world, but not peanuts. Worth making a prominent part of a slightly more elaborate plan, I'd say. – Shog9 Jul 18 '15 at 1:19
  • 2
    @Shog9 - That does seem worth it, then, particularly if incentives promote more of this behavior. Thought the numbers would be much lower than that. – Brad Larson Jul 18 '15 at 1:36
  • @Shog9 Are those 1186 posts network-wide or just Stack Overflow? – durron597 Jul 18 '15 at 3:28
  • That's just SO @durron - I'm much too lazy to add this up network-wide. – Shog9 Jul 18 '15 at 3:58
  • Is the "Shog9" link supposed to go to some super secret room that the rest of us can't view? I'm also curious how this would affect gold-hammer holders, as it's not like they stop caring about reputation after they get their hammer. – Kevin Brown Jul 18 '15 at 14:48
  • No idea where that link came from, @Kevin - I've replaced it with perhaps a more likely target. As for gold-badge closers, I believe this could be made to work for them as well. – Shog9 Jul 18 '15 at 16:22
12

I would suggest that the asker still be given +2 rep for saying that finding a duplicate solved their problem. Though there would be an incentive to ask duplicate questions for the sake of 2 rep, this is only slightly worse than the current situation, where you can intentionally ask a duplicate and hope to get at least one answer before it's closed. Either way a user risks downvotes and/or question bans by doing this.

If you don't award the +2 rep, the asker has an incentive to accept an answer rather than agreeing that a helpful duplicate has been found.

Giving the +2 seems like a good tradeoff, since it creates a minor and hazardous way of gaming the system but helps close an effective and pernicious way of gaming the system - answering known duplicates before they get closed. It also helps duplicates get found faster and cuts down on the burden of cleaning up sites.

  • 5
    Come to think of it, all that would have to happen is one person notice the duplicate gaming the system and one downvote would negate the +2. – durron597 Jul 17 '15 at 20:02
  • Maybe make the +2 a +1 if its closed as duplicate to encourage non dupe question asking? – Jessica Brown Jul 22 '15 at 19:51
5

This is a follow-up idea that is not an essential portion of the first idea. Therefore, I am adding it as an answer so that it can be voted on separately.

When an asker selects "That solves my problem!" it could create a stub answer with a checkmark, exactly as if the person who found the duplicate had written it as an answer.

This would accomplish two things:

  1. It would be even more clear to future readers that this duplicate solved the asker's problem.
  2. It would allow the asker to remove the checkmark if they change their mind and decide that their question is not a duplicate.
    • This would not reopen the question, they would need to get 5 reopen votes to do this.

Unfortunately, a side effect of doing this is that it would be awkward not to assign the +2 rep for accepting an answer. The only solution I can think of is flagging "creating duplicates on purpose" for moderator attention; I think it would not be that hard to if people began to create duplicates on purpose just for the +2 rep. This rep could, potentially, also be capped in the same way that rep is capped for suggesting edits.

  • 1
    possible duplicate of Create CW answer for every “Vote to close as duplicate” vote :) – gnat Jul 17 '15 at 18:39
  • @gnat It's not a duplicate of that because this answer wouldn't be CW (because you CAN get 15 rep from it). I suppose you could get more than 15 rep actually, if people upvote the stub. – durron597 Jul 17 '15 at 18:39
  • meta.stackexchange.com/a/194495/165773 – gnat Jul 17 '15 at 18:40
  • @gnat stackoverflow.com/q/28504747/1768232 – durron597 Jul 17 '15 at 18:42
  • the one at SO matches Section 1; one here seems to be more like covered in section 2 "only looks similar..." :) – gnat Jul 17 '15 at 18:43
  • 4
    I like this a lot. It goes some way toward addressing the primary concern over the proposal that gnat cited, which is a link-only answer that encourages other link-only answers - once the question is closed, no more answers can be posted. I have no real concerns about awarding +2 to the asker; if you exclusively post duplicates, you'll be q-banned before you can garner any real reputation that way. – Shog9 Jul 17 '15 at 22:36
  • 2
    Often, more than one user casts a duplicate vote before the OP accepts it's a duplicate. Who owns the answer? Who gets the rep? I guess only the first one, who did the hard work of finding the duplicate. Makes sense... – user259867 Jul 17 '15 at 22:58
  • 4
    This really gets fun when you have multiple voters suggesting different dup-targets... I still like it though; life would get boring without edge cases like that. – Shog9 Jul 18 '15 at 1:21
  • 1
    @Shog9 there couldn't be more than 5 voters, but come think of flaggers. If, say, 10 flaggers suggest 10 different dupes and asker accepts, how would system handle that – gnat Jul 18 '15 at 8:33
  • 2
    Can this "stub" be visible only to the OP, @Shog9? My big worry with this is users who can't VtC posting link-only "answers" that mimic this on other, open, questions, because they don't know that the dupe-answers are auto-generated and think that it's just the way things are done. – Josh Caswell Jul 18 '15 at 15:54
  • @JoshCaswell how about voter/flagger who proposed a dupe? how about 10Kers, would it make sense to show it to them? I feel like not but am not sure... it seems natural for visibility of such an answer to be the same as of accept duplicate dialog, that is shown only to asker? – gnat Jul 18 '15 at 16:01
  • Yeah, those would make sense, @gnat; really just so long as it's not shown to people who might not understand that it's a special, system-generated non-answer. – Josh Caswell Jul 19 '15 at 0:34
  • 2
    @durron597 it is a problem. To outsiders, it would look like a rewarded link-only answer - we better abstain of widely "advertising" stuff like that (wrt green circle it's an interesting twist, really worth thinking about) – gnat Jul 19 '15 at 0:46
  • 1
    @durron597 I think you underestimate "creative copying" ability of inexperienced visitors. Following what is shown to them, expect them to write their answers like that: "##This is my answer for a question that was an exact duplicate.<br/><br/>You can find the answer here: <Link>" (I bet you'll see stuff like that:) – gnat Jul 19 '15 at 0:51
  • 2
    Looks reasonable to me, @gnat, but then I'm not the one implementing it. The other (possibly easier) option would be a post with a completely different appearance than an answer, but still some kind of pseudo-checkmark. – Josh Caswell Jul 19 '15 at 18:28
4

I totally agree with your proposal. I do have some worries that were somewhat discussed already. I want to make a post out of them because it really worries me...

From the proposal:

We would only award the +15 rep if the asker themselves clicks the "that solved my problem!" button. If it's closed as duplicate from 5 CVs, then no rep would be assigned.

From this comment:

As far as gold badge holders, I've been thinking about this all day... I don't think it matters too much if their incentives are wrong because they already have so much rep to have gotten a gold badge (I have 13k in basically one tag and I still don't have a gold badge) they probably won't care. And there just aren't that many of them.

Based on the post Penalty for answering help-vampire questions? Or reward closure? on MSO, I am worried that gold tag badge holders are excluded from the current proposal to be rewarded. I think they should definitely be the ones that should be rewarded, for a few reasons:

  • If you have a gold tag badge, you've been around quite some time. They probably know to find a lot of duplicates, so we should really encourage them to close if appropriate.

  • They can be seen as the users capable of being the Fastest Gun since they can answer 90% of the questions without the need to test or fact check anything. If they aren't rewarded, they might not vote to close at all, and just answer.

I am also a little worried on another part of the proposal:

  • The user has to take some action himself. As Shog9 mentioned, 15% of the users 'accepted' the duplicate post. That still means that 85% didn't have any participation in the closing process. Would you take the risk vote to close as a duplicate, while you can answer and be sure from some reputation? 15 reputation (as per the proposal) can be achieved by two upvotes, and that doesn't seem so much. Why not just let the closer keep his reputation if the closure wasn't disputed?

A question that came up in my mind was: does the reputation earned from this 'accepted close' feature fall in the reputation cap? If not, it could lead to a flood of close votes. If it does, one could get less motivated to vote to close, since reputation from upvotes are usually spread over days.

And can we please prevent answering a question after it was closed? It is possible to answer now when altering the web page a little. (Reference: How was this answer posted after this question was closed?)

  • Surely you see the problem of giving people free rep for closing another's question? Currently there is almost no way to GIVE YOURSELF rep... One possibility is to create the stub answer when a question is dupehammered, then allow the asker to have the option of accepting it but it wouldn't be placed automatically. – durron597 Jul 18 '15 at 13:27
  • One possibility to prevent abusive duplicate closure is simply to leave it in the hands of moderators and user flags and site analytic secret sauce... With very stiff penalties for repeated closures that aren't really duplicates. – durron597 Jul 18 '15 at 13:30
  • 1
    I am not talking about the problem of rep farming. I am talking about leaving an important group out of the proposal. Also, there must be a way to make sure this works. Not just for 15%, for 100%. I am afraid it else might not work out well. – Patrick Hofman Jul 18 '15 at 13:53
  • I know you're not talking about rep farming; you raise some very valid points here. I am simply discussing the flip side of these points. – durron597 Jul 18 '15 at 14:20
  • 2
    I see no reason not to generate stub answers for all dup closures, including those triggered by gold badge holders. Auto-accept (and thus reputation) would need to be reserved for those cases where the asker was directly involved, but nothing stops the asker from returning later to find their question closed and a helpful answer they can accept... Indeed, this happens today albeit in an ad-hoc fashion. – Shog9 Jul 18 '15 at 16:25

You must log in to answer this question.

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