Kung Fu Legends
As Yahtzee pointed out in his review of Dragon Age: Origins, we’re in a weird situation if there exists (without irony) the concept of a “generic fantasy game”. Even moreso when that phrase is basically equivalent to “action- and statistics-oriented role-playing game”. Seems like it’s time for people to try something a bit different, and that’s the point of my project Kung Fu Legends.
Historically this project began many, many years ago. I used to take a train from Brisbane to see my folks during university semester breaks. It was a good time for me to do some reading or thinking. One of the things I worked on was an unnamed kung fu game. I was then currently addicted to Jackie Chan films like Drunken Master, The Fearless Hyena and Young Master. For about a day I brainstormed all the things I would have liked in the game and potential gameplay ideas. To be honest, at that stage I had very little programming ability1 and moreover, probably little ability to create a game. All my ideas went into a weird notebook with a pink exterior. Somehow that book travelled with me through half a dozen house moves and the requisite purging of items.
Fast-forward to about last year. In the years inbetween I had learnt a lot about programming. A lot. I felt I was pretty dangerous with C++ and knew a lot about graphics systems. I had made initial steps on a game temporarily called “Survival Horror Engine” (SHE). I even documented some of my technical ideas on my website. But the project was well beyond my ability. Probably even beyond the ability of a team of ten programmers. I had envisioned a 3D, procedurally-generated city where you could wander an apocalypse and, I dunno, try to survive 2. I needed to ditch SHE and try something a bit more within my scope. I remembered my kung fu game and managed to find the old, battered, pink notebook. In amongst cheat codes for Wiz n Liz, notes on Measure Theory and some disjointed scribbles about a short story involving a bouncer from Jerry Springer, I found my old notes for Kung Fu Masters. Most of it was me just jazzing on the theme and building up a mental image, but I had some neat ideas. I decided that this would replace my noble failure of SHE.
Kung Fu Legends3 is an RPG with roots in kung fu and wuxia films, old Zen stories and The Journey to the West. I don’t want a graphics replacement of some “generic fantasy” game. I want to avoid the “Gaming” part of GNS Theory (or equivalently, the “Gamism” of The Big Model) and push towards the Narrativist/Simulationist approaches, but stay within the confines of a computer game.
This is a fancy-pants way of stating my first philosophy of the game: “I want an RPG, but different.” I’m fascinated by the mainstream push of having players tell or experience stories in games like The Sims series. I’d like people to be able to share interesting chronicles from a kung fu world.
The second philosophy I’m using is that ”If a computer can do it, then it should.” This philosophy embraces aspects at both ends of the spectrum: obviously desirable and potentially controversial. On the obvious side, we have powerful CPUs now. It should take care of trivial tasks like inventory management because we can easily spare the CPU cycles. There is a lot of traction in the gaming world for our computer personas to inhabit our own personal spaces in the form of achievements, statistics and recorded stories. There is scope for things less explored but easily implemented like RSS feeds of your progress, or the collaborative use of save games and the stories that follow from that.
On the more controversial side of things, I’d like to explore two things: AI companions and computational creativity. AI companions have been done before a thousand times over, but they typically fall into a subservient role like pets or secondary characters that obey your every order. I’d like to explore the idea of characters that work with you or guide you in a way unlike a tutorial. Specifically I’d like to mess with the whole Master/Student relationship that kung fu stories have. Originally I wanted to be very indie and put you into the role of a dead Kung Fu Master who could guide a student in a long-term, strategy-not-tactics, implicit way. You could guide them but not force the AI student to do anything particular. This defeats the idea of micromanagement which permeates modern RPGs and takes a step away from the Gamism approach. But maybe that isn’t as fun as actually being in the world, so perhaps you could have the choice of being the student instead, and get guidance from your Master. This is possibly very difficult to get right, but you could at least imagine this working at a basic level.
Creativity is, in my mind, something a computer can attempt. They’ll never have the genius of a human plotting things out, but I think with the right system, you can get interesting enough results out. That is, it’d be “good enough” but I doubt it’d be great. It’d be something like procedural content creation, but a lot more guided than random level generation and the like. The computer will be build to develop plots and push the game world in hopefully interesting ways. This would all fall over if its results were given to the user via an array of canned responses. So to this end I am programming a system that will create written English from events in the game world. And not madlib-style English. Actual English with concerns about aesthetics and phrasing.
It’s all a bit ambitious, but I think it’s more doable than SHE. And, hey, if I fail, what have I lost?
I might leave it there for tonight and talk more about my ideas on Kung Fu Legends another time. I hope to have a bunch of articles on specific aspects of the game as I develop it.