I've been working on putting some game design documents together since I came away to India. At first in Manali, a week after I arrived; then here in Goa, towards the end of my trip, I picked up my pen again and opened the exercise book I'd been using to make my notes.
Working on game development isn't new to me. I've worked on two titles, one at a small startup and one at a medium sized company with a respectable back catalog. Both titles made it to release and I was pleased to have worked on a variety of game development programming tasks, from AI to multiplayer, to low level script compilation and execution, to 3D animation and 2D GUI systems. I believe I'm capable of programming a game by myself.
What I didn't realise, when I decided I wanted to make a Role Playing Game (RPG) for the Android Platform, was the true scope of what I was attempting. I've never worked on game design before, so when I started writing my design documents it was new, exciting territory. I enjoyed it immensely. I got to writing it up onto a private wiki. Then my partner, who is also in IT, suggested I should write a functional specification. I've done this kind of thing before too, in enterprise, and thought it was a great idea: from a functional spec, architecture is clearer, and so is the true size of your project. For those not in IT, a functional spec basically describes everything the application (the game) needs to do, in exhausting detail. It doesn't describe how - that's in a technical spec if you write one, or up to the programmer. But it shows you everything you need to do.
So I started on my functional spec. I'm making an RPG. It started off with the basics: how gameplay flowed between the title screen, the main "roaming the world" mode, cut scenes, combat. Then it exploded.
I don't mean it exploded as in, what happened when Luke fired his photon torpedoes into the Death Star's exhaust vent. I mean it got bigger; like a hungry octopus, stretching out its tentacles in every direction; like cancer. Before I knew it, I had multiple documents describing different functional parts of the game: scripting, combat, skills.
A few days into this, however, I was pleased with my progress: I'd gotten huge parts of the functional spec finished, and was even precisely specifying how each episode of the game (there are nine in total) progressed. Then one day I thought it was a good time to start putting feelers out to some friends who were artists and seeing if any of them were interested in doing some 3D art for me. I'm making a 3D game, but I'm not an artist. Even if I learned how to use 3D Studio or Maya, I don't have time to make the creative stuff too. I've already realised how full my hands will be just with programming. But to get any kind of accurate quote for creative assets, I need to figure out how many I need!
I start counting: first how many different "character classes" are in the game. (This doesn't include "extras" that might populate the world but not be directly involved in the story). "Kingdom Warrior", "Queen", "King". "Enemy Kingdom Captain". "Tiger". "Noble". The list expands before my eyes. I gulp.
I write down which character classes could re-use the same models with a different texture. The list still ain't small: 14 models, 51 textures. Oh shit. I have difficulty sleeping that night: I really, passionately want to develop this game, more than any other game I've worked on in my life. Yet I've badly underestimated its scope. The reason there seems to be a gap in the market for RPGs on Android becomes clearer: RPGs are BIG.
How many people worked on Final Fantasy 6, my standard for a good RPG? The game I'm most inspired by? Almost 50. I won't be making a game with anywhere near as many gameplay hours as FF6, but still. That's scary.
I still want to do this, but I've learned a valuable lesson: spec as early as you can. Derive some roadmaps estimating your project's lifespan as early as you can. Figure out a ballpark figure of how many man hours that adds up to. As early as you can.
Then do it anyway.
