Iterative Design
“Game design is 1% inspiration and 99% iteration” - Chris Swain
The process of iterative design is the key to good game design.
Analysis / Conceptualize
In design, analysis is the process of understanding where you are and what you are trying to do. In software development we often talk of the “problem to be solved”, where the “problem” is how to develop the software we want to produce. It also involves understand the resources (including time) you have available to you. Analysis isn’t about solving the problem, it is about identifying and understanding it.
Initially this is about developing the game’s concept, in later iterations it is about analysing what you have developed and the feedback from the playtesters to improve your game.
Design
Analysis helps us understand what we are trying to do, design is the process of finding a way of do it. Specifically it is the process of coming up with a way to do it it is not about actually doing this. Design begins with idea generation and ends with a plan for implementation.
Implementation
Implementation is the process of putting the plan in to action and coming up with a solution to the “problem”. In game development implementation is about getting from a game design idea to a playable prototype as quickly as possible. Early implementations will be basic, perhaps only a character moving around the screen perhaps not even that. Paper prototyping, creating boardgame or toy prototypes of your game can be useful in the early stages, you don’t need to dive straight in to code.
Testing
Testing, play-testing is absolutely essential. You might perfectly design and implement your vision and it might be no fun to anyone except yourself. Get people playing your game, get them playing it as early and as often as possible. Get them playing the basic game play before you move on to anything else. If you are making a platformer make sure people like the jumping mechanic works and the way it feels. If you are making an FPS find out if players like the movement and shooting.
Test as often as possible, don’t make lots of changes between test sessions, you want to understand what the effect of each change is. Too many changes between tests and it is hard to judge what change prompt which response.