Game Design is the process of deciding the rules and mechanics of a game, and solving balancing problems to achieve the intended play experience. For questions about design of software code, use the Architecture or Algorithm tags instead. Likewise, questions about visual design should use Art tags.
A game's design is the set of rules, mechanics, and player feedback that determine how it plays and feels. This design is what makes, for example, the gameplay of Tetris so recognizable, even when comparing any of its dozens of incarnations with different code, art, and music.
Game design is the process of deciding what these mechanics should be, in order to support a particular kind of play. As Katie Salen & Eric Zimmerman define it in Rules of Play, "The game designer only indirectly designs the player's experience, by directly designing the rules"
For examples of the kinds of game decisions that fall under game design, see Liz England's The Door Problem, which contrasts common design considerations with those of other specialties in game development.
An effective Gamedev.StackExchange question about designing a game mechanic should be sure to hit on these four points:
Establish the context of your game - what is the genre, platform, intended audience, etc. A suitable design for Peggle is not necessarily a suitable design for Dark Souls, and vice versa.
Explain the feature whose design you're working on. How does it work currently? What is the player experience like?
Define a specific desired outcome - what should this feature accomplish when the rules, balance, and feedback are turned right?
- This might include demonstrating a problem with your current scheme that you want to solve: an unfinished gap, edge case, exploit, or a player experience that strays from what you want to create.
Ask for a strategy to adjust your game's rules, balance, and feedback to reach your desired outcome.
To keep game design questions from straying too broad or opinion-based, it's important to be very clear about your constraints and end goals. "Creating something without any outcome in mind is not design but experimentation" as Vili Lehdonvirta and Edward Castronova put it. A solid problem definition helps your peers identify which possible answers correctly solve it.