11

This isn't really a question, but I'm not exactly sure where to put it. I would like to request stack exchange linking rewards.

Everyday many of the same questions are asked. Users that are experienced will link to a Post where the question has already been answered. If the link is put in as a comment, and the the question is recommended for close, the user gets no rewards for that. But if they answer it again (or answer it with a link to duplicate question), they are more likely to get rewards.

Linking common questions to well established answers would seem to be desirable, and should possibly be rewarded more than another new answer.

I'm not saying how this could be implemented, but I wanted to throw this idea out there as something to consider.

Edit: I want to acknowledge that I think it's good for the same question to get asked in different ways by different people. This makes searching to find the answer easier (assuming that they all link to a common answer). All duplicate questions should ideally link to the correct question they are duplicating. Searching to find the best answer can take some time and effort on behalf of the person answering/requesting close, and would seem to warrant some reward.

  • 3
    I agree. If it is accepted that duplicates are undesirable, the responsible thing to do is get them closed. A comment linking to the dupe helps the questioner find the answer quickly, and should be encouraged. Answering obvious dupes and collecting reputation points is currently possible and incentivized, but could be viewed as an abdication of personal responsibility to better this site. I'd like to see a 'best practice' for dealing with dupes spelled out clearly. If you lost extra rep attached to a question that gets closed as a dupe this might be good enough. – Tom Brossman Jul 15 '12 at 20:48
  • 1
    @TomBrossman Questions closed as dupes often have valuable answers. Some have them merged into the master question. When they don't, it's often because the master question's answers already have enough information. That, in turn, is often because they were edited to incorporate content from the dupe's answers, often by the dupe-answerers themselves. It's hard to see seeing how reversing rep on dupes' answers wouldn't punish lots of useful actions. Plus, many such answers may be written by newcomers who we shouldn't expect to search for dupes to a question they can answer, before answering it. – Eliah Kagan Jul 15 '12 at 23:31
  • 1
    @TomBrossman To look at it from another angle: Suppose I answer a question, and subsequently realize it's a duplicate, but (for either good or bad reasons) I don't want to delete my answer. Right now, I can still comment/close-vote, without fearing losing reputation I've gained if it's closed as a duplicate. What you propose, on the other hand, would actually take my reputation away in direct response to me doing the right thing and getting it closed as a duplicate. As a result, it stands to reason that far fewer duplicates would be closed as such, if this policy were implemented. – Eliah Kagan Jul 15 '12 at 23:34
  • 2
    @Eliah I agree that people would be disincentivized to close questions in the cases where they would lose rewards. It's always good to have input from every possible perspective. Hopefully website administrators here can look at the points made on this issue and come to a reasonable solution. – Sepero Jul 16 '12 at 7:11
  • 2
    @EliahKagan Users answering dupes should lose rep when failing to look for similars, same as the asker. This is by design, it is desirable. Best practice when considering a question should be to look for duplicates first, since so many users fail to (they are even pro-actively suggested when asking). Help new users find an answer with a comment, then flag or cast a close vote. Help all users fight fragmentation of good answers by linking to, editing, and developing canonical questions and answers. – Tom Brossman Jul 16 '12 at 10:24
3

Provided that a question asks the same thing as another question on the same StackExchange site, the way we deal with that should not just be to comment with a link, but to close the question as a duplicate.

I think commenting with links should not carry any reputation or badges. Except when accompanied by close votes, it's not particularly more valuable than any other kind of comment.

This leads to the question of whether or not there should be reputation or badges for closing questions, under some circumstances.

Reputation is awarded for authoring content--that is, you get rep for upvoted questions and answers, and a tiny bit for edits that are approved by other users. I don't think it would be appropriate to give reputation for closing questions, under any circumstances.

As for badges, we should be careful. There's nothing inherently good about closing a question--it depends on whether or not the question was rightly closed. We don't want to unbalance people's judgment by offering extrinsic rewards for closing questions. (To conceptualize how this would be bad, think of the inverse: Imagine if we gave reputation or badges for reopening questions.)

However, the system is capable of granting badges based on complex rules, so I think a badge for closing questions well is something we should strongly consider. The simplest way I can think of, that I think would work well, would be to have a badge (or multiple badges) for voting to close some number of questions N that were:

  • each subsequently closed as duplicates, in part as a result of that vote
  • each not reopened, at least for some reasonably long time period T after closing
  • each subsequently visited by some number of users K, who then visited the master question and upvoted it and/or one or more answers

Granting the badge based on the question's closure leading to upvotes on posts in the master is, I think, the critical element necessary, if rewarding close-votes is to be balanced and helpful.

  • I understand that closing questions is the current way it's done. I do think that reputation should be rewarded on recommending to close a question (as much as a correctly authored answer), on the stipulation that the person is correct. Of course, the problem is that only the person asking the question knows for certain if is it a duplicate of their question. Perhaps if a question is closed, the asker could have the opportunity to reopen it, so they can clarify how it is not like, or not answered by the solution being redirected to. – Sepero Jul 15 '12 at 18:51
  • 3
    As long as recommending to close questions carries no rewards, the greater incentive will remain to simply answer questions with links to previously answered questions. – Sepero Jul 15 '12 at 18:58
  • @Sepero Then other members of the community can vote to close as a duplicate, and even downvote those answers. (An answer that's just a link to something else is not usually a very good answer, so in this situation downvoting would be reasonable and would provide disincentive to answering rather than commenting and/or close-voting.) Also, please note that answering with a link is much more likely to get a question close-voted than to get a lot of upvotes, and while not a particular good approach, it might be considered helpful, especially compared to doing nothing. (New users can't comment.) – Eliah Kagan Jul 15 '12 at 19:00
  • You are correct, and I agree with everything you're saying here. Hopefully though we can agree that this does not resolve the issue. You are talking about "disincentivization", and that's not what this website is founded on. The problem currently is that there is no reputation incentive for users to do the correct thing in the context of duplicate questions. – Sepero Jul 15 '12 at 19:12
  • @Sepero There's no reputation incentive for doing edits that don't require approval, for reviewing, for voting on other people's posts, for posting on meta, for posting useful comments, for voting or running in a moderator election, and so forth. Many useful actions don't have a reputation incentive, and yet people still do them. I don't think it's obvious that this is really a problem at all. You might want to edit your question to make a more compelling case that it is. (You could try to provide examples, and perhaps even statistical information, about duplicate questions, for instance.) – Eliah Kagan Jul 15 '12 at 19:24
  • 1
    I apologize if I made it seem like people won't do something without reputation incentive. That's not what I meant to imply, and yes people do a lot of things without reputation incentive. My only point is that when people are given two options: reputation incentive(for repeat answer) vs no reputation incentive(correctly requesting a close), the average person will be inclined to take the obvious choice. The result can be established through a simple matter of logical deduction. – Sepero Jul 15 '12 at 20:58
0

Since we are part of the larger Stack Exchange family of websites, I knew duplicates would be an issue well covered already. Duplicates about duplicates on two Metas, oh, the irony.

Required Meta Stack Overflow reading here and here and here and here.

My preference going in to this was to see users asking and answering duplicates lose reputation when the duplicate was closed. I don't care as much now, though I still wish more would spend time improving existing Q&As. Merging can only be done by a select few, but we can all edit and improve questions. Do so, it lessens the burden for moderators.

I propose a 'De-duplicator' badge, or some other way of rewarding those that take the time to seek out and link to dupes. This takes a lot of effort, many of us do it religiously, and there is no recognition nor reward granted.

Rewards need to be based on a very high percentage of your dupe flags being accepted (maybe 90-95%?), to discourage reckless flagging.

  • 20 Dupes closed in one day with your 40 flags? Not bad Tiger. Here's a bronze.
  • 200 in one month? Is that a silver badge in your pocket or are you just happy to see me?
  • One thousand duplicates flagged correctly? B-I-Double-G-I-E De-Double-duplicator come on over here and get your gold medal.
  • I think you make great points there, Tom! And I really like the idea of rewarding people who tie duplicates together. The great thing about dup-linking is that it generally requires less skill/knowledge than fully answering questions. So it's open for opportunity to more people, but it still takes time and effort though! Dup-linking can serve to get value from people who may not know many answers, but are willing to take some time to find the answers for a little reward. – Sepero Jul 17 '12 at 15:59
  • 3
    (continued) Getting dups closed and linked to a consolidated answer takes effort, ultimately benefits everyone, and makes the whole website better. There is no reason it should be a thankless job. Stack Exchange is largely based around incentivization, and this seems to be something very worthy of being incentivized. – Sepero Jul 17 '12 at 16:01

You must log in to answer this question.

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