6

I've noticed this a couple of times recently. There have been a number of steak questions, specifically Cooking steak on the grill, where I feel like the information has all been well trodden on the site, but I couldn't find an exact duplicate to close as. Closing with a "Read all the stuff with the Steak tag" doesn't seem like a great solution either.

A recent question, Getting better in the kitchen, about beginning to cook, was just closed by Aaronut. Again, I do not think this is an exact duplicate. The new question is really about getting to a point where one has the prerequisite knowledge and skills to be able to improvise in the kitchen, where as the question it duplicates is really about starting from square one. It's not a huge distinction, but it is one. There are also a whole bunch of questions linked in Aaronut's comment that also cover similar ground. This seems very similar to the steak example. I don't think it's an exact duplicate, but it covers similar ground and a lot of the information that could go in an answer has been covered somewhere else on the site. But is a question really a duplicate if the close reason includes a link to half a dozen questions?

So when does a question become a duplicate? Have we covered steak well enough that any new question should probably be closed as a duplicate and involve a link to 3 or 4 different questions? That doesn't seem particularly satisfying to me, but on the other hand neither does rehashing the same conversations over and over.

Some thoughts from the SO blog: Joel argues for closing (although he argues against it if there's a ton of extraneous info or a RTFM approach) and Jeff argues that it's probably ok, unless I'm missing something in their arguments after an admittedly light read.

Thoughts?

3 Answers 3

3

That doesn't seem particularly satisfying to me, but on the other hand neither does rehashing the same conversations over and over.

This is the crux of the matter; if the regulars are bored with providing repetitive answers, there is a duplicate problem.

The textbook definition of a duplicate question, to me, is a question where all the answers on question A are exactly correct as-is on question B. Do note the caveat, however: if the answers on question A have some specific, subtle ties to the wording or contents of question A, they can't always work on question B even if they are duplicates in spirit. This isn't necessarily a problem, but it is a big reason why I advocate "just enough" duplicates.

But a key sign that you have too many duplicates is answerer fatigue, for sure.

1

What Joel really argues for is editing, which not a lot of people here do. In other words, if there are 10 mostly-insignificant variations on the same question, choose one to be the "canonical" question and edit it (both question and answers) so that it answers all the variants as well. This is the conversation we were having around the endless stream of food safety questions - they're all slightly different but they all could be answered by one well-written Q&A thread, if we had one.

The question I closed, I would have closed as Not A Real Question if I hadn't closed it as a duplicate. Really, all it's asking is, "How do I learn to cook?", and that is practically what our entire site is about. But, it did duplicate other questions, and close-as-duplicate seems to create less angst than close-as-NARQ, so I chose that reason instead.

There's a particularly insidious form of duplicate on Stack Exchange sites which I like to call the frankendupe. These take the form of extremely broad and usually very basic questions that look like they've been constructed from the bits and pieces of a dozen other, older questions. It's hard to point to a specific dupe because there are so many questions being asked at once, and in fact the question could potentially cover a significant percentage of all subject matter on the site.

In fact this is one major reason why we have the NARQ ("too broad") close reason - when a question consumes an entire subject (or several), it tends to cannibalize the expansive and valuable array of existing and/or potential smaller questions on the subject, and limit their depth. Like plopping a Wal-mart into the middle of a busy strip plaza; you're not going to see any more Mom & Pop stores, and there's a good chance that the ones there now won't survive long.

My opinion, and anyone reading this is welcome to convince me otherwise: We only need one general "Getting Started with Cooking" question on this site, and we already have way more than that. Unless a new question provides a truly unique or exciting new angle, it should be closed, either as a duplicate or as NARQ.

4
  • I don't particularly care about the "Getting Started with cooking" question. I actually think the steak question is a better example, as it's much more specific and clearly on topic.
    – yossarian
    Jul 21, 2011 at 21:34
  • By which I mean, it sounds like you're responding more to one example than the other. If it were just the steak question, would your answer change?
    – yossarian
    Jul 21, 2011 at 21:37
  • 1
    @yossarian: The steak question, to me, is actually a different ball game. Both the question you linked to, and the one linked as related/duplicate, are questions that are, IMO, too general/broad to begin with. "How do I cook X" (in this case X = steak, but we've also seen chicken, fish, etc.) is perilously close to a recipe request. "How do I cook X using appliance Y" is better, but still poll-ish. A much more interesting and focused question is this one: Rinse the salt off a steak before cooking?. That is (IMO) the right scope for a question.
    – Aaronut
    Jul 21, 2011 at 21:55
  • 1
    In other words, @yossarian, in the specific case of the steak questions we're analyzing the symptom rather than the cause. While canonical questions are great, many of the original "How to cook X?" questions were basically pre-emptive "frankendupes", which make a lot of future questions look like dupes because it's impossible to craft a new question that's not already covered by the mega-poll.
    – Aaronut
    Jul 21, 2011 at 21:58
0

Hmm ... I guess I missed this when it was first posted.

I'm actually going to go for an even stronger requirement than Jeff's ...

To be a true duplicate:

  1. All valid answers for A are valid answers for B.
  2. All valid answers for B are valid answers for A.
  3. The answers for each are valid for the same reason. (the equivalent of 'correlation isn't the same as causation' ... just because they have the same set of answers does not mean they're a duplicate ... otherwise, every question of 'how long can I keep (x) in the fridge?' could be considered answered, even if x1 and x2 aren't related)

I've just voted to re-open the question that Yossarian referenced, as I don't think it's a duplicate. Although there are valid answers to the new question in those previously asked questions, there might be additional answers that aren't necessarily valid answers in any of those others. And there are some answers to those other questions that aren't valid for this question,

I also think that tips for someone who's been cooking for some time and doesn't think they're progressing are not the same as a beginner. Yes, some cookbooks and cooking shows to watch might be the same, but I don't think the two lists are identical. And the question ould've been better if it had been more focused (I'm having trouble with (preparation) or (type of food)), but then it should've been flagged as too broad, not a duplicate.

4
  • This really is in an inappropriate set of criteria, as it literally permits any question to be re-asked with a single arbitrary restriction; that's entirely against the spirit of the duplicate system. Even so, the question that was closed seems to fit all of your criteria here; your personal list might be different, but there is no objective means by which to declare any conceivable answer for one as invalid for the other. And, sure, we could also close it as Not Constructive, but then people would be complaining because we didn't close the other nearly-identical one.
    – Aaronut
    Aug 11, 2011 at 0:47
  • @Aaronut : I'll go back to my policy when I was a moderator -- when in doubt, wait to see how the community reacts. If they start marking it as a duplicate, then I'd close it. If not, let it ride. I'd say that some of the moderators here are a little too heavy-handed on closing answers. The goal might be to become an authoritative knowledgebase for all things cooking ... but you're never going to get there without building a community first. If duplicates are really a problem, find a way to head them off before they're asked.
    – Joe
    Aug 11, 2011 at 3:31
  • 1
    @joe, I kind of agree with you about the question being ok, but not strongly enough to fight for it. Re: heavy handed mods, a lot of the close 'votes' we get are flags that the regular population can't see as you need 10k rep to view them (although you can, so should know).
    – yossarian
    Aug 11, 2011 at 13:36
  • 1
    And I don't think there is a way to fend them off before they're asked. You'll often see comments on a question "yeah, thought it might be off topic, but I asked anyway", duplicates that any search would have uncovered, or questions clearly against FAQ guidelines. I think it gets back to the RTFM mentality it's easier to ask for forgiveness than permission, so people just ask and apologize.
    – yossarian
    Aug 11, 2011 at 13:37

You must log in to answer this question.

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