Beginnings are always difficult, as the fellow said in Trouble in Paradise, and boy do I know it. I've spent two-plus hours banging my head against the wall, writing and then discarding one intro paragraph after another in seemingly fruitless attempts to get to the part where I talk about what I want to talk about, which is kinda apt and maybe even ironic, because what I want to talk about is starting a thing - an area where, the occasional stubborn blog-thing aside, I usually don't have many problems.
I recognize of course that this isn't the case for everybody. More than once an aspiring designer will ask folks on the internet, "How do you actually do it? Where do you start? How does one design a game?" They've read the theory, they know techniques and mechanisms; what they want to know is that next (or rather, first) step that takes it from an idea to a game. That part remains mysterious, even esoteric, as secret as the ancient knowledge that allows Lamont Cranston to cloud men's minds.
I've been on the receiving end of that question before, and it's about as impossible to answer as that old chestnut, where do your ideas come from? Heck if I know. It's not something that's really in my control; I can't will an idea into existence. It's not outside of my control, either - I'm not some passive receptacle waiting for some ethereal muse to gently plop an idea into my brain. I work at it, sometimes working very hard, and I certainly have something to do with it, but there's no recipe, no step-by-step method that reliably results in inspiration.
It's like trying to have a romantic evening with your partner. You can put on mood music, dim the lights, smile a certain way and speak with a playful softness; you can create an atmosphere that's conducive to romance, but that doesn't mean it's gonna take. Likewise, you do things that make it easier to come up with ideas. Read a lot, play a lot, think a lot, be interested in things, be engaged in the world around you, be open. But you can easily do all of these things all day long and always end up empty-handed.
Really, that's why folks ask the question over and again, even though the answer is almost always the same no matter who they're asking, even though the answer always leaves them dissatisfied, will always leave them dissatisfied. They don't want techniques and tips and "read a lot" and "be open"; they've tried that; they want the answer. But there isn't one. There never has been. There never will be.
The same is true of the question we started with: how do you design a game? Recently and against my better judgment I poked my head in on a forum post where someone was decrying the uselessness of all the design how-to books, podcasts, and blog-things, because while they gave you techniques, broke down theory, and dissected mechanisms, they never actually answered the question. They never let you in on the secret recipe, never actually instructed you, step-by-step, how to design a game, what the right decisions are, and how to make them.
Every game is its own thing, its own miraculous and impossible creation. It's a magic trick. It gets easier with time and experience, and you pick up techniques and short-cuts - the craft you can fall back on - but it's always a process of discovery and iteration and trying things that don't work until they do. There's no secret handshake, no special or arcane knowledge. (Or if there is, nobody's ever let me in on it.)
How do you design a game? I design it. It's not a helpful answer, but it's the only one.
How do you start? I write some rules, I make some pieces, I get it on the table. I just start. I just do it.
Sometimes it doesn't work, and that's frustrating. Heck, early on, most of the time it didn't work. If you ask some gamers who are not particularly enamored of my output, it still doesn't work, but that's neither here nor there. The only way to learn how to do it is to do it.