5
$\begingroup$

We've had a few questions come up on Meta regarding a specific type of puzzle Joe Z. has dubbed "mass-producible puzzles." These, in essence, are puzzles

  • which can be produced in great quantities
  • for which there are common shared strategies

Two good examples are cryptic crossword clues and Sudoku puzzles. The (potential) problem here is that MPPs (mass producible puzzles) tend to lead to mass-producible questions - that is, we'll end up with dozens of questions regarding specific cryptic crossword clues.

In general, there are two types of MPPs that I see: the first, questions regarding puzzles in the specific ("how do I solve [this specific Sudoku/cryptic clue]"), and questions regarding puzzles in the general ("what strategies move me forward from [certain point in puzzle]", "how does [method] work?", etc.).

We've had a couple meta questions about this topic as a whole already:

These questions, I think, all (generally) come back to this point. Are specific or general MPP questions on-topic?

$\endgroup$
8
$\begingroup$

I think they are on-topic, yes. Similar question styles are acceptable on StackOverflow. Let's take each of your examples at a time:

"how do I solve [this specific Sudoku/cryptic clue]"

These actually seem to be asked more in the format of:

Here's [this specific puzzle I'm trying to solve]
Here's [solving methods I've tried so far, and explanations of why]
What am I doing wrong? What's the right way to find the solution?

A similar acceptable question on StackOverflow would be:

Here's [this specific programming problem I'm trying to solve]
Here's [code I've tried so far, and how I've tried to fix it]
What am I doing wrong? What should I do to fix the code?


Your next example:

"what strategies move me forward from [certain point in puzzle]"

That's an accurate description of some of our questions, but to flesh it out a bit:

Here's [this specific puzzle I'm trying to solve]
Here's [how far I've gotten, and how I did it]
What's the next step to solve the puzzle?

A similar StackOverflow question:

Here's [this specific programming problem I'm trying to solve]
Here's [my code so far, and why I wrote it this way]
What's the next step to achieve the result I want?


Your final example:

"how does [method] work?"

I think what you mean by this is a question like this:

Here's [puzzle]
Here's [posted solution to puzzle]
Why does that solution work? Here's [a bit of my thinking on the matter that doesn't quite reach the conclusion, but shows effort and tells answerers where to start]

A similar StackOverflow question:

Here's [code snippet]
Here's [what the code's supposed to do]
Why does that code do that? [a bit of my thinking on the matter that doesn't quite reach the conclusion, but shows effort and tells answerers where to start]


Obviously there are high and low quality questions in all these categories across all sites (StackOverflow was just the easiest example for comparison purposes). Not every question that follows this type of format is going to be a good one. But I don't think the fact that they do follow this format is enough to make them inherently off-topic. If the asker shows effort, puts forward their attempt to solve, and asks a question about puzzles that's interesting, I don't see why these questions should be disallowed. If it's literally just a puzzle dump, that's like "gimme teh codez", and should be closed/downvoted as no effort/low quality. But that's a separate issue in my opinion.

$\endgroup$
4
$\begingroup$

I believe they are on-topic as a class of puzzle, but not on individual puzzles. The key is to attack the core problem of the puzzle, not one particular instance of it.

The reason a puzzle is mass-producible is precisely the fact that they have a common set of principals that apply to all such puzzles and thus, a canonical answer that covers all such problems should be possible.

This allows us to address the type of puzzle without being overwhelmed by too many of them. It also means that fairly quickly, most of the common questions about such puzzles will be answered and we won't have many non-dupe questions about them after we have a well defined set of questions and answers covering it.

$\endgroup$
  • $\begingroup$ In almost all cases I can think of, such an answer would be too broad. Imagine an answer containing all major Sudoku strategies and methods. There could be (and are dozens) of books on the topic. (I have two myself.) $\endgroup$ – Aza Jun 3 '14 at 14:56
  • $\begingroup$ Asking for solutions isn't too broad, asking for all solutions is. "How to solve Sudoku?" isn't too broad as any one of those possible answers is valid and votes could be done based on how creative or informative a particular answer is. Similarly, further questions could be asked about any of those specific approaches mentioned in the general question. "What are all possible ways to solve sudoku?" would be too broad though. $\endgroup$ – AJ Henderson Jun 3 '14 at 14:59
  • 2
    $\begingroup$ The too broad close reason reads: "There are either too many possible answers, or good answers would be too long for this format." The answers to "How do I solve Sudoku?" gives answers which are both numerous and exceedingly long. Any complete answer would be very long, and even at that not comprehensively answer the question. $\endgroup$ – Aza Jun 3 '14 at 15:09
  • 1
    $\begingroup$ @Emrakul - at that point, any question about efficiently solving any problem would be "too broad". In my experience, too many answers generally refers to where there are too many equally good answers. Maybe I just don't understand the Sudoku situation fully, but I expect that for the most part, there are relatively few completely unrelated solid solutions that don't share a lot of commonality. If I'm wrong about that, then I simply didn't fully understand the complexity of that particular problem. $\endgroup$ – AJ Henderson Jun 3 '14 at 15:18
  • 1
    $\begingroup$ Well... in my opinion, there's a significant difference. Compare it to someone asking on Math.SE asking how to solve algebra problems, versus solving a specific algebra problem. Sudoku is similar: when people are stuck, typically there are only one or two things you can do; a general guide, however, could fill a book. $\endgroup$ – Aza Jun 3 '14 at 15:34
  • $\begingroup$ Sudoku is trivially solvable in the generic case, the complexity only comes from more elaborate solutions. I guess what I would oppose in such a question is partial answers as general tips and tricks to work towards a solution would be too numerous, but actual full blown methods that guarantee a solution efficiently would be fewer in number. The problem with asking about specific sudoku problems is that they are completely undiscoverable (unless you can generically describe the particular problem you are trying to deal with). $\endgroup$ – AJ Henderson Jun 3 '14 at 15:39
  • $\begingroup$ They would be fewer in number, true, but they would all be too long. There are books on just X-wing, Y-wing, and Swordfish/chain patterns. $\endgroup$ – Aza Jun 3 '14 at 15:41
  • $\begingroup$ I'm afraid I have to agree with @Emrakul on this one. I think the wording in that close reason/FAQ somewhere that amounts to "If you can image an entire book answering your question, it's probably not a good fit here" (where here = SE) is pretty spot-on, and applies in this case. "How do you solve Sudoku" is way, way too broad. $\endgroup$ – WendiKidd Jun 4 '14 at 0:18
  • $\begingroup$ @WendiKidd - yeah, it sounds like that example is more broad than I realized. $\endgroup$ – AJ Henderson Jun 4 '14 at 2:17

You must log in to answer this question.

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