I recently posted a self-answered question concerning the strategies I use to beat Minesweeper games consistently.

Doorknob justifiably pointed out that the question could, perhaps, be a bit too broad. While I disagree with him, I do recognize that he has a valid point. This question is, well, questionable. It pushes the line between 'too broad' and 'not quite too broad' for "how do I solve this puzzle/logic game?"

So, I turn to the community! How broad is too broad? For instance, can I ask:

  • About how to solve a Rubik's cube?
    • About how to solve the last layer of a Rubik's cube?
    • About how to change the cube into a snake pattern algorithmically?
  • About how to solve the Petaminx?
  • About how to solve Picross?
  • About how to solve Minesweeper?
  • About how to solve 2048?
  • About how to solve Freecell?

Some of these questions require more broad solutions, and some of them could be summarized in a few sentences. These are the major questions:

  • Where do we draw the line for 'too broad' with these questions?
  • Assuming these questions are modified to be more specific, how specific to they have to be to be acceptable?
  • $\begingroup$ I would say the 2048 one is also a bit too broad, although I feel it's not so much "too broad" as it is "not detailed enough". In general, a question about "how do I solve *" could give you a million different answers in terms of a puzzle-related question simply by virtue of its sweeping generality. You can make it more specific simply by rephrasing it and asking for specific strategies. $\endgroup$ – user88 May 22 '14 at 1:56
  • $\begingroup$ @JoeZ. Hmm, I see what you're saying, but that's tricky to do for some puzzles. Generally speaking, though, this question is more so about where the threshold is for how specific a question can be. I'll edit it to clarify. $\endgroup$ – user20 May 22 '14 at 1:59
  • 1
    $\begingroup$ It's times like these that I miss the "not constructive" close reason. $\endgroup$ – user88 May 25 '14 at 20:19
  • $\begingroup$ @Joe Z, having personally studied the patterns, and replicating them for a college assignment, I can assure you that you only need to follow one simple pattern. Pick one side and move constantly to that side. If not, try the side to its immediate surroundings (If you picked top, pick right/left). If there's no moves, move to the opposite side of the first pick. Follow that pattern and the puzzle is solved. The Puzzle can be solved rather easily, however in a long process (which I described) $\endgroup$ – Oak Feb 23 '16 at 8:20

I'm of the opinion that, if the solution varies significantly, it is off-topic, but if the solution is unique, it is on-topic.

Let's take a look at the examples I've provided:

  • About how to solve a Rubik's cube? This question has too much variance. The solution is non-unique, as dozens of solutions exists with thousands of algorithms and their variants. As a result, there is no way for there to be a singular, correct answer to this question.
    • About how to solve the last layer of a Rubik's cube? This is closer to on-topic, as it does have only a handful of answers - only a small portion of algorithms solve the last layer this way, and they're done in very similar ways. This question would still benefit from refinement, though.
    • About how to change the cube into a snake pattern algorithmically? This question is on-topic, as it has a defined, non-subjectively correct answer.
  • About how to solve the Petaminx? Take a look at the picture. Clearly too broad.
  • About how to solve Picross? This is another one of those questionable areas. There's really only a set number of methods one can use to solve Picross before guessing, so it would be okay by me, but this one is for the judgment of the reader.
  • About how to solve Minesweeper? There really aren't multiple ways to solve a minesweeper puzzle by logic. There are a set number of convenient patterns, and the rules of the game mean there really isn't any deviation.
  • About how to solve 2048? This puzzle is very generic, but one could argue that there's not much of a singularly correct strategy to this problem. I'm not sure I'd personally take action on it: this is also left up to the reader.
  • About how to solve Freecell? As there's not a single unifying method for approaching this problem, it is too broad.

In essence, what it comes down to is whether there's subjectivity, variance, etc. in the solutions to a puzzle. If there are too many possible answers, then the question is too broad; if the answer involves the combination of too many stages or steps, it is also too broad.

In response to this feedback, I have split my Minesweeper question into two parts: What frequent patterns can be used to overcome Minesweeper obstacles? and How do I approach a "stuck" game of minesweeper?


You must log in to answer this question.