Site Policy: Judging When a Question is a Duplicate

Question

For anyone who knows the kind of question I post, I put a great deal of effort into making sure a question is unique. I check the site for similar questions to see if those answer the one I would otherwise ask and, if they don't, I carefully word my question to be clear how my question differs.

So, when my recent question was flagged as a duplicate, I checked the proposal, confirmed it was one I had reviewed before asking, and edited my question to indicate why mine was different. After that, I noted that I lost a close vote, then gained three more to close the question.

There was no further discussion on the subject to indicate why those continuing to vote to close disagreed with my analysis of the two posts. It seems appropriate to link to my usual reminder to close-voters, but I'd also like to bring the issue to the community's attention so I can do better next time (if there is something I can do better).

Answer 1

I'm going to leave my original answer, which was partially in response to Frostfyre's query about the closure of his question and partially an introduction to the issues of judging duplicates. This answer is focuses solely on the issue of judging duplicates.

Duplicity

There might be a better word out there, but my head's chasing all the ideas for this answer and can't think of it. "Duplicity" usually identifies something deceitful or fraudulent. A person with a duplicitous nature identifies a person who, for example, has ulterior motives.

However, I'm going to wantonly use "duplicity" to mean "the state of being a duplicate." If you can think of a better word, lemme know.

Why judge at all? Why not just answer the @#&! question?

Worldbuilding.SE is the most creative, imaginative, and therefore the most subjective stack in the Stack Exchange galaxy. That's actually a problem, because Stack Exchange designed its service to accommodate very objective, clear, precise, concise, specific, even obvious questions. In the world of programming, for example, there may be a dozen different ways to approach programming for a specific effect or need — but the effect or need is very specific and it's obvious that anyone asking about that specific effect or need is asking a duplicate question.

Judging duplicates serves two important needs on all stacks.

1. It brings the many different ways of looking at the same problem into focus. This brings value to the overall stack by gathering all the best answers into one place with the several (sometimes many) duplicates all pointing to that one source. Even knowing that your perspective is a duplicate of other perspectives helps programmers stack users to see better how to solve their problem.

2. It reduces clutter on the stacks. If you don't mark duplicate questions as duplicates, (a) you get the same answers repeated over and over and over and over and over and (b) good answers are not all in the same place, meaning greater research is needed to find them, meaning the quality and usefulness of the stack is reduced.

A substantial component of how we do things here at worldbuilding.SE is based on Stack Exchange's expectation of how their service (not our service) is to be used.

Unfortunately, we deal with stories. I've been a micropublisher, and I can honestly tell you there are only the proverbial nine unique stories. Period. Boy meets girl, battles robot, saves the girl. I can't tell you how many tens of thousands of times (I mean that, we processed 3,000-5,000 manuscript submissions each year — as a micropublisher) an author would tell us "there's nothing like my story out there!" or "This is a completely new and unique story!" No, it isn't. There's no such thing.

But there are an infinite number of ways to tell those stories. And that's where publishers look for uniqueness. The flavor, the ambience, the worldview....

And that's the problem we have with judging duplicates. Where 99% of the users of (e.g.) Stack Overflow will likely agree that two questions are duplicates of one another, maybe only 70% will here (or less) because stories are far more interpretive than, e.g., programming — and the answers of question #1 might not actually address the specific problem of question #2. Or maybe they do and the OP of question #2 is just being stubborn and short-sighted. Or maybe they don't because it's the judges who are being stubborn and short-sighted. See the problem?

You stupid engineers...

Finally, before actually getting to the meat of this answer — I'm an engineer. I've also been a publisher and am a writer (both technical and artistic). You might not believe it, but this is a useful perspective. I've had to deal with authors who don't see the difference between artists and engineers. If you'll forgive the simplification:

Artists live in a world full of rules, but don't want to believe it and often pay little or no regard to those rules. They're trained that way. They need to see the whimsy of life and expression of imagination to be good at what they do. Consequently, they often have difficulty being published because publishers are businesses (rule-driven hierarchical organizations) that must pay their employees.

On the flip side, all engineers see is rules. We're trained that way. We literally can't do our jobs without both understanding the rules and regularly using processes that create new rules. We can't afford to give up those rules or bridges collapse and people die. We also have difficulty being published because consumers are emotional and are looking for an escape from their rule-burdened lives.

It's the people who can merge the artist and the engineer within who are often most successful.

Regrettably, what you're about to experience is 98% the engineer. Fair warning. Let's begin.

Titles, Questions & Answers

I suspect a common habit is to look at the titles of the two posts and judge their duplicity based solely on those few words. I was guilty of that early in my experience with WB.SE. Unfortunately, while a valid point of judgment, titles are also the weakest aspect of two posts to compare.

A thorough reading of the body of both posts (the question) is mandatory for a comparison. Someone who skims over the two without really considering the many aspects of a question (conditions & restrictions, background or premise, and the question itself) is providing a superficial judgement.

But I wholly agree with TimB II that the answers must also be taken into consideration. But this needs a grain of salt because a post may not have been answered sufficiently (even though it was answered adequately) which affects the judgement.

So, to begin, good judgement requires evaluating the titles of the two questions, the body of the two questions, and the answers to the first question. If you're not willing to take the time to do that, please consider not voting.

2+2 = 1+3 = 1.8+2.2 = 421-317... or do they?

I'd like to introduce my evolving set of rules for judging duplicates. This is, of course, my own opinion. Don't let my use of the word "axiom" fool you.

Axiom 1: If two questions can only be answered with exactly the same answers, then they are duplicates.

This first axiom is completely true and, after considering what Mithrandir24601 had to say, almost completely useless. Not completely, but darn close. Unlike our more objective sister-stacks, questions asked here are rarely perfect duplicates, which underscores the need to evaluate answers along with the questions when judging duplicity.

Corollary 1.1: If two answers can only be the result of exactly the same questions, then the questions must be duplicates.

This corollary is almost as useless as its axiom in that the world of writing is almost never this neat and tidy. But I've found it useful because comparing how respondents have interpreted questions has proven valuable to my judging process. And this deserves a quick tangent.

OP Intent: The original poster of any one question posted that question with a specific, if not well defined, intent or expectation.

Respondent Interpretation: Respondents have no choice but to interpret the question to ascertain the OP's intent because there's no such thing on Worldbuilding.SE as a "perfect question." There will always be a disparity between the OP's intent and the respondent's interpretation — especially among our more creative respondents (bless them all!) who will throw a curve ball at the OP just to stir things up.¹

Therefore...

Axiom 2: If there is a variance between the OP's intent and the respondents' interpretations, in keeping with Stack Exchange's policy of preserving OP intent (...to clarify the meaning of a post without changing that meaning...), it is the OP's intent that must be preserved and judged when considering answers as a basis for duplicity.

Case: more vs. less

Axiom 3: If the two questions are asking about the same thing and differ only in conditions, limitations, or restrictions, then they might be duplicates.

Corollary 3.1: If the conditions, limitations, or restrictions of Question #2 narrows the question as found in Question #1, then Question #2 is a duplicate of Question #1 because the answer to Question #2 could be found associated with Question #1.

Corollary 3.2: If the conditions, limitations, or restrictions of Question #2 expand the question as found in Question #1, then Question #2 is NOT a duplicate of Question #1 because there can be valid answers to Question #2 that cannot be found associated with Question #1.

That was ugly to read, but what it boils down to is that narrow questions are always duplicates of their broad counterparts. Sometimes...

Corollary 3.3: If the conditions, limitations, or restrictions of Question #2 narrow the question as found in Question #1 but demonstrate a critical understanding or unique perspective unlikely to be casually found in an answer to Question #1, then Question #2 is NOT a duplicate.

And there's that fuzzy border. Without the OP of Question #2 making a darn good case for uniqueness, identifying that border is very subjective. Remember, the people who use this site range from untrained children to highly trained adults that don't have the training other highly trained adults have. In other words, where one person may see a very clear distinction that makes this an easy judgement for them, another may not and yet feel like they have a clear and justifiable perspective. Ugh.

Lemma 3.3.1: While it's true that Stack Exchange does not require anyone to justify any vote or flag, considering how subjective the subject matter on this site is, you should err on the side of leaving an explanation in comments — if only so people can contact you for clarification.

Case: Near misses

Axiom 4: If Question #2 is similar to Question #1 in ways that cannot be attributed to conditions, limitations, or restrictions then they might be duplicates.

Corollary 4.1: If from the perspective of someone skilled in the art or science necessary to answer the question the two questions are substantially the same, then they are duplicates.

Corollary 4.2: If If you are NOT skilled in the art or science necessary to answer the question, you should err on the side of posting a comment asking for clarification before voting to close the question as a duplicate.

If you thought the last axiom and its corollaries were fuzzy, this one is really fuzzy. It's basically asking you, the judge, to put yourself into the shoes of people who know more than you do. Said another way, it's demanding that you, the judge, accept the fact that you don't know everything. That's a lot easier said than done! But it must be done. If the comparison requires knowledge you, the judge, have not been trained to, then you really should proceed with caution. Yes, it's just a website and just a hobby (at least it should be just a hobby!), but you'd be surprised how upset people can get over hobbies. After all, what you're doing is another form of calling someone's baby ugly. Since the OP of Question #2 will at some level take it personally that their question was closed, please be sure you're right.

Case: When all else fails...

Axiom 5: Protecting the integrity and intent of Stack Exchange should not take precedence over helping an OP solve their problem.

I have no doubt that there's a lot of people who use or have used this site who think I don't adhere to that axiom. They might be right. I'm not perfect. But I am an engineer, so it shouldn't surprise anyone that I default to following rules.

However, it's important to realize there's a balance between meeting Stack Exchange's expectations and the OP's needs. You can't simply ignore one in favor of the other. The OP needs and answer, but it isn't our site. Many people consider places like this (or Facebook, or Reddit, or any other web site) to be "public property" where the rules of free speech reign. In reality, all web sites are the equivalent of private property. Each and every one of us is a guest — a guest who's expected to follow the rules of the host.

As I explained earlier, there's good reasons why Stack Exchange wants questions marked as duplicates. Those reasons should be respected, and so we should have an eye open for the purpose of identifying duplicates when they happen. But we also need to be on the lookout for those gems, those questions that expose an amazing perspective that deserve to be left open. That can be hard to do because one of the most valuable aspects of Stack Exchange is its belief that the answers should help everyone, not just the person who posts the question. In the end, that might be a useful perspective, so let's add it as a corollary.

Corollary 5.1: If the answer to Question #2 is useful to no one other than the OP who asked Question #2, and Question #2 is a near-miss duplicate of Question #1, then Question #2 is a duplicate of Question #1.

OK, my question's a duplicate, but the answers to Q#1 don't help. What do I do?

I think we often forget that there are things we can do — and should do — to solve the OP's problem when said OP's question is marked a duplicate.

• We can set a bounty on Question #1 and ask for more answers. We can link to Question #2 in the notes to that bounty so people see why the additional answers are sought. If a low-rep OP whose question has been closed feels this should happen, use comments and the @user tag to contact one of the higher-rep users in the comment chain (or visit The Factory Floor chat room) and ask said person if they'd be willing to sponsor the bounty. We should all be willing to do that from time to time, just as lawyers are supposed to do pro bono work from time to time. Occasionally we need to be reminded of that. 😀

• We can post a new answer to Question #1 ourselves. Indeed, maybe it should be a habit of worldbuilding.SE users to look a newly-closed-as-duplicate questions, hop back to the first question, and see if there's an answer we can post. If I remember correctly, there are even badges awarded to people who successfully answer old questions.

• Judiciously,² we can edit Question #1 to make the needs of Question #2 more clear in Question #1. This causes Question #1 to bounce to the top of the "Active" list so that it's visible again and can be more thoroughly answered.

---

_{¹ If you want an example of this, go read the answer Willk gives to almost any question. He's a pro at throwing curve balls that open the OP's eyes to possibilities they might never have thought of. That's an incredibly valuable skill on this site. It also muddies the waters when judging duplicates something awful. I wish there were more users like him on the site.}

_{² I say "judiciously" because I'm not a fan of editing questions on this site. It's far too easy to change the OP's intent here. That's a consequence of this site being so naturally subjective. Therefore, while I promote editing Question #1 as a solution, I beg people to be cautious with their question edits. Please don't invalidate Question #1 or its answers in your effort to get answers useful to Question #2.}

Answer 2

I was one of the close voters and, regrettably, also the OP for the "original question." I voted as a duplicate because I believe that if any two questions, no matter how they're worded, can be answered with all the same answers, then they are duplicates by definition.

But I've been wondering about that belief, and this is as good a time as any to bring it up. Basically the question boils down to this:

• 1 + 3 = 4
• 2 + 2 = 4

Are those "duplicate questions" since they share the identical answer despite having ostensibly unique question conditions?

I'm happy to change my voting habits to conform to community consensus.

Answer 3

I think that JBH's answer has a lot of merit and is a good rationale of the YES vote. To be blunt, I'm swayed by it but given the lack of a NO argument I'd like to put my original proposition from when I first read this question...

In the case of the question you propose, there is a similar intent, and therefore the propensity for similar approaches. Your argument for uniqueness is one of scope. Ultimately therefore I think we can look at this from two different perspectives.

Restriction of original scope
The second question is asking about achieving a similar effect or using a similar approach, but does so by limiting resources, timeframes, or applying new constraints on the effort.

Expansion of original scope
The second question is giving 'free rein' on the original question by removing constraints, timeframes, resource limitations, et al.

These are in effect optimisation issues. Of the two, I prefer the idea of the restriction of original scope as a unique question because it's easier to see how (from an engineering perspective at least) that can lead to the need for a new or refined approach. The problem is that if you have a really elegant answer to the original question, the limitations may not actually restrict the optimal elegant answer from applying in your case.

On the other hand, while expansion of the original scope by definition means that the original answer is still valid, you're actually more likely to come up with a better optimised solution through removal of some of the engineering constraint.

Therefore, while I actually agree with JBH and have upvoted him accordingly, I'm going to argue that the answers provided to the original question must be taken into account when considering if 'scope change' questions are in fact duplicates or not. I'm further going to argue that this should be done by the questioner in the first instance.

In other words, when you become aware that there is a duplicate challenge on your question, it's insufficient to read the original question in isolation to the answers before deciding whether your question is unique or not.

Scope change questions can be unique, but IMHO only if the answers that have been provided fail an elegance test that would make them valid across multiple scopes.

This is a subtle argument, and possibly wrong, but at least the NO camp is represented.

Answer 4

I'm afraid I have to agree with the close voters.

The answer would be the same, it's not possible to remove all life on Earth without basically melting the planet. While your constraints are slightly different the fundamental outcome is the same.

To address the general case rather than this specific one:

Thresholds of difference

There's a threshold in the answer to the existing question and a threshold of change of constraint in the question to be balanced. "Given an extra 1000 years of technology" is a significant change of constraint, even though the answer may be the same it could be a different question. "But make the bombs blue" is not a significant change of constraint.

In this case the threshold in the answer, the requirement to melt the planet, puts the change of constraint in the question, leave the buildings standing, below the threshold required for it to be a different question.

Quality of answers

A good answer will answer the question and any very similar question. A really good answer will of its own right increase the breadth of questions that become duplicate.

In this case, the answer to the original wasn't just "you can't do it like that", but "here's the standard you have to achieve to do that". That significantly increases the range of questions that are now duplicates as any question that doesn't allow you to achieve the melting of the planet for the effect is now a duplicate.

Answer 5

Let's take an extreme example.

Question 1: What grows in my garden? Answer: Tomatoes, cucumbers, carrots, peppers

Question 2: What goes into a salad? Answer: Tomatoes, cucumbers, carrots, peppers

Do you VTC Q2 because Q1 has answers that fit? I think we can see that the question itself is important in relation to the answers, not just the answers themselves.

Q1 can also have these answers: daffodils, spruce trees, cacti; Q2 can be answered with olive oil, salt, vinegar. So not all answers from Q1 fit Q2 and not all answers that fit Q2 would fit Q1. Some basic answers might fit both, but a comprehensive answer would only fit one or the other.

I think the important point is whether the OP gets the answer to the question that he's looking for, not whether the question can somehow roughly sort of be matched with another answer where some of the answers might possibly fit.

There are essentially two kinds of questions in Worldbuilding.SE

This SE is kind of special because it can and regularly does have questions that can have multiple valid answers, but the OP may or may not accept any of them as suitable. Worldbuilding is a creative process and many people come here to get ideas or inspiration, so there may not be a right or wrong answer and many answers are desirable, as they make it more likely that not only the OP but anyone else interested in the question gets ideas that are just right for them.

What I propose should be done with these is this:

1. If a possible duplicate is found, it should be shown to the OP, so he could look through the answers there.
2. If he finds a good answer to his question, his question can be closed as duplicate.
3. If he doesn't think any of the answers suit him, find out why he thinks that.
4. Does he (a) think they answer his question, but he doesn't find any of them acceptable and hopes someone will come up with something else? Or does he (b) think the answers are off-topic and don't quite fit his question?

In the case that it's (a), especially if the original question only had a couple questions and one was quickly chosen as correct, that question could be reopened, a note added to it to explain that more answers are desired, perhaps a bounty placed. Q2 would be closed as duplicate and linked to the original where hopefully more answers will come.

In the case that it's (b), the OP should be permitted to reopen the question after specifying a constraint to get answers necessarily different from Q1. I.e., What goes into a salad (and doesn't grow in my garden)? Or just making his question more specific.

Questions can look similar on the surface and the answers might seem suitable to others, but not to OP because the intent is very different. For example, two questions about body augmentation. The original seemingly has plenty of answers that would fit, but the intent is different. In the original all the answers are about implants/cybernetics, but OP is actually looking for genetic augmentation. Or the answers are at the wrong technological level. Adding more details to his question will allow him to get the answers he wants.

Some questions are very specific to the world being built by the OP, they tend to start out simple, but would get lots of comments, many clarifications of the question and many corrections to the answer(s) to get very specific details for that world that wouldn't be relevant for the Q2 world at all, even if on the surface they look very similar. Planetary arrangement questions, for example. They might both be trying to figure out a particular arrangement of planets, moons, stars, that seems the same at first glance but gets very specific with sizes, distances, periods that make it useful for that particular story. It wouldn't make sense to lump these together at all. Although these aren't directly as useful to the community, they still offer great ideas due to the detail.

Sometimes the original question is already huge and old, with a dozen or more answers with plenty of votes and comments. Even if they are similar, trying to get more answers wouldn't really work, any additional new requests would just get lost in the sea of existing content. Maybe in such cases it may be allowed to open a new question and specify that you are looking for answers that aren't in the Q1. What goes into a salad(excluding everything mentioned in Q1)?

The above doesn't apply to the other kind of questions. Questions that can only have one right answer. Science-based or reality check type questions, for example. Here the OP might not realize that the other answer applies.

The fundamental nature of these kinds of questions is different. Here he's asking something he doesn't know from people who are more knowledgeable than him and the right answer is already known to them even if OP doesn't pick it. He might not realize that some difference he sees between the original question and his might not matter and the answer is still the same.

There isn't really a need for multiple answers saying the same thing or multiple questions asking the same thing in a slightly different way. Ideally the original question should have one or several high quality answers that generalize to many situations and all related questions should be marked as duplicates and point to it. As it happens in most other SEs.

If the OP disagrees that it's a duplicate, he can be sent to chat or sandbox to get it clarified so he can get the question closed or altered and reopened.

---

In the case of the question in the OP, I think that it's the first kind of question, its intention is to get ideas and inspiration. With the right constraints emphasized I think it's possible to only get completely different answers that wouldn't fit the original question at all. Even if on the surface they look similar, their intent is completely different, they are building very different worlds and stories.

It's still true that many answers can overlap and fit both, though. The original has 21 answers and many of them actually fit your question better than the original question (the time frame and tech level needed). And to me the answers look pretty exhaustive, it would be hard to come up with more ideas for either question that wouldn't be variations of some answer from the original.

For some reason you didn't find what you were looking for in the original answers, though. Or perhaps you didn't read all of them well. You explained that your constraints are different and the answers don't fit the question, but some actually do.

So you should clarify why exactly the answers in the original don't work for you and if further constrains can be placed on your question to only get answers different from the original. If it's possible to do that and still get useful answers then I think it should be permitted to reopen.