Since I'm now a designer on a successful Kickstarter project, developer on a very popular set of wargame rules and currently starting on a new project for release it looks like I can legitimately claim to have a design "process". As I'm starting Moonflight I thought I'd use this as a chance to examine my personal process in some detail.
Just to say first of all, stages is a very apt description because by developing a range of projects by stages you allow yourself one of the most important creative tools, settling. It is useful in all creative fields but particularly in writing to let an idea settle. Write it down and walk away, forget about it, give yourself time to fall out of love with it then return to it. It's easy to believe that piling in with drive and passion and smashing your project through by sheer self belief is a good idea but its generally not. Let it settle and allow it to win you over when you return to it in just the way it would win over a new user.
Before you start designing make an effort to educate yourself in what makes good games good, play and study games. Frequent your local charity shops and buy any game you don't recognise. If you've never played a deck builder or dice builder, worker placement or area control, push your luck or take that game make a special effort to change that. As a note, the difficulty of piecing together what makes a game work is why I always worry when people say they wrote a game because they played something and "thought that they could do better". When I'm playing a good game I don't examine it, if something misfires or clunks everybody notices it. Improvements are easy to make to bad games and so are poor motivation. Personally, I feel most inspired to write a game just after playing something I can't improve.
Then get yourself a note book or piece of paper or something and start a list of your ideas. I've heard it said that "real" designers have a rich and deep supply of constant ideas. I just want to say, if you've altered a rule you're a "real" designer, just as you are a real writer after you've completed your first short story. However, just as a writer grinding out copy to pay the rent or one that has honed their craft for years to gain the respect of their peers will rankle when you then group yourself in their ranks so will a games designer who has spent years polishing a game they pitched around a hundred publishers react harshly when their craft is opened to the hobbyist. There is nothing wrong with games design as hobby but it is distinct from games design as profession.
Having access to a solid bank of ideas is part of what defines the full time designer or professional from the hobbyist. Ultimately, if you have only one mechanic available to resolve a particular issue your project will suffer. To believe that every one of my first ideas would be good enough to make it unchanged to my final project would be an act of terrible arrogance. If pitching to publishers, a range of ideas will help you to fit into their output and if publishing yourself you can't honestly select your best idea if its your only idea.
Thinking of ideas breeds ideas. Personally, game design is the thing I think of when I'm not thinking of anything, but that's a frame of mind that comes on over time. Until then, list every idea that might be useful, but be sure to curate that list, edit and examine it. My list tends to consist of a single line description of an idea, a guess at its format and a few notes. But how do you decide what goes on the list, or at least what makes the edit?
My ideas tend to fall into one of two general categories, narrative or mechanical. I'm not really a world builder so my narrative ideas tend to be interpretations of TV shows, movies and novels. To assess the worth of such an idea you must examine the story's structure, the nature of the protagonists and the "feel" of the story. Narratives and games tend to conform to a three act structure but narratives generally have an introduction, set back and overcoming, its participants will tend to be at their weakest in the middle of the narrative. Conversely, games usually have a continual progression of power. So to assess a story's suitability you'll need to assess how well its story beats fit into a game's pattern of progression. Secondly, narratives tend to be strongly tied to their protagonists; unless your game is an RPG you'll be unlikely to represent or direct its participants in any but the most general surface fashion. If your game ever allows a player to gain advantage by acting in a cynical, cowardly or brutal fashion and that would destroy the story you have a serious problem. Finally, most good narratives have a sense or "feel", those directed at younger audiences may even have a specific moral. It is extremely difficult to service such a sense but to fail to do so effectively is a terrible disservice to your source material. It is entirely possible to make a best selling game about horror that is not at all tense or scary, one about comedy that is not funny or one about sharing and sacrifice which rewards cruel and selfish behavior. A sense of fear or humour is generally based upon novelty, to create it repeatedly in a game is punishingly difficult, to provide a satisfying sense of guilt or regret is potentially dangerous.
Usually my most fully crystallized ideas at the list stage are those attached to specific game mechanics. They should be as original as possible, have "meta" options and sit within a clear rules framework. As far as originality goes, not every minor rule needs to be totally original, but a mechanic that you note and hope to build a game around should be a unique attraction. Meta options are difficult to judge at such an early stage but mechanics worthy of note should suggest some form of emergent narrative or game play, emotional response or broad rules interaction at least. Finally, you should already have a rough idea of the framework that the new mechanic will sit into and confidence that it will not collapse that framework around it. Being able to picture a new mechanic in a rough framework is generally a result of studying the mechanics of the widest range of games possible.
A deep and well managed list of ideas is not a necessity but it is a hugely beneficial resource and any serious designer should set about building and maintaining one sooner rather than later.
So, for the sake of example, a couple of years ago the idea occurred to me that a basic deck builder's tendency to score an entire deck seemed to be throwing away the design space afforded by scoring only a player's hand. How could a player control their hand's contents for scoring, how would game end scoring be finally triggered? It was a game mechanic idea that seemed, to my current understanding, to be largely unique. The available interactions of cards to both build and trash decks, the order in which the engine would need to be built and dismantled to a player's best advantage, all provide a range of options for both meta game building and emergent game play. Clearly the mechanic would fit easily into an existing deck building framework. So the root idea for Moonflight went on the list as "deck wrecker".