I've been thinking about using genetic ideas to create game elements a lot recently.
After all just taking a stream of bits and translating that into interesting game effects without having huge amounts of code is something that would be really cool on mobile devices.
Over the weekend I had an idea which I wanted to get down on paper before I forgot it.
The basic idea is to create a massive range of spells from a few basic genes.
Have a look and see if it's one of my mad ideas, or if it really does have worth.
If we think it's worthwhile I'll do some code for it, though if I do I would like someone else to take my code and generate a devmaster format tutorial from it.
I don't have much spare time, well any spare time, and I'm not a graphic artist.
After reading your idea I started to think about 3D magical mechanics/physics.
Mage have a set of basic 3D magic structures: magic frame lines, magic power lines, motor, delay line, mana limiter/concentrator, fire/water/wind/earth effector, aura detector, logical elements etc. Each magic structure have some set of parameters (size, magic line thickness, type of mana...) and have well defined mathematical model behind it. When mana concentration in structure get over some threshold then that structure activates. Structure can break if concentration is too high or disappear if it's too low.
Mage creates full spell from structures with below threshold mana concentration, then fill some startup structure and put magic physics into action. Hi-level spells consist of many individual magic structures, so mage must build it incrementally, keeping partially built parts between disappearance and activation thresholds, and then charge it with sufficient for full activation amount of mana.
So at least my ramblings have got your brain working
I, personally, don't think more is always better. I would rather have fewer spells that had a strategical difference in the game. I'm pretty old time about rpg's, though. I like turn based, grid type, so my opinion is pretty much meaningless.
This sounds like something you would want to make the centerpiece of a game, and base most of the challenges around finding just the right combination of genes to meet a variety of challenges; almost like Scribblenauts, but with your magic system replacing the dictionary system. Making a series of fairly open-ended challenges which the player can complete any way he can think of (within the limits his avatar level and the system itself allow) would be the best way to showcase the variety and multifunctionality of such a system without it just becoming thirty-thousand different flavors of Fireball.
I think; but maybe I missed something?
EDIT: You know, I've been giving this about five or six stray thoughts here and there, and it seems to me this mechanic would be perfect for a 'Wizard School' kind of game, where almost the whole thing is one big, extended tutorial in a scholastic setting. You could make some pretty neat approximations to what it might really be like to be learning magic, from the most basic firebolt spell to some pretty complicated transmutation or conjuration stuff. It's a cool idea, if you could make it work well enough. In that context, I don't even think the system would really have to be that balanced, even. Heck, Scribblenauts is completely unbalanced, almost broken, and it's still quite fun.
The idea of "genetic spells" is in line with my musings from a few years ago.. http://sol.gfxile.net/magic.html
I had an idea for an MMO where you'd have spells take the form of a sort of computational currency like BitCoin. What I mean by that is, you'd build some multi-staged, difficult puzzle as the spell creation system (not necessarily an abstract puzzle - it should be things like the mana wires, genes, etc systems). When you're done building, you can compress the spell into a 'packet' which then just has some resultant mechanics in-game that are basically invisible. The trick is, you can teach someone your spell, but they just get the packet, not the design or the ability to teach others.
So this creates an in-game marketplace of the best spells to do X or Y or Z, where players can create resources for the marketplace by coming up with better/more efficient solutions to the spell mini-game. This provides some disincentive to posting good spell diagrams on wikis and the like, which means that figuring out the subsystem can remain as a longer term part of the game-play experience.
Whether or not it would be fun depends a lot on the implementation of the puzzle and of what the spells actually do in the world though. And while there's kind of an academic interest in the economics that would result, I'm not sure they'd actually be enjoyable to experience as a player.