next »

[jlauener]

Hey, my idea is to create a game similar to http://totallytinyarcade.com/. To make it short the goal is to complete as much mini games as possible in a given time frame. Some games , aside from the win condition, might give you time bonus (e.g. +10 sec). The twist, here, is that the mini games are randomly generated! To do so I'm planning to randomly pick a set of goals, mechanics and rules to randomly generate a game.

My first question is, do anybody know some examples/essays/literature on the subject of random generation of rules/mechanics/goals to create a game? Using google I mostly ended up on random game idea generator which is not what I want... I currently have no idea how feasible this is.

Now about what I'm thinking right now:
I do not want the scope to be too wide and I the project will be done in one month (#1gam). So I decided to always use a top down 2D view with a character that can move using arrow keys and (sometime) shoot. So basically no gravity which means no jump, no platormer-like games. Is it too restrictive?

Using this top-down 2D gamespace here are the stuffs I brainstormed.

Goals:
-kill X enemies
-collect X items

Moving mechanics:
-can move in 4 direction
-can move left/right
-can move up/down
-always move, can change direction using arrow key

Shooting mechanics:
-shoot in the direction the entity is facing
-always shoot in a given direction
-shoot in all 4 direction
-shoot a projectile
-shoot a beam
-shoot a bouncing laser
-drop a bomb
-can "eat" others

note: shooting mechanics can be applied to player and enemies

Enemy behaviors:
-seek the player
-avoid the player
-always move in a given direction
-move left/right
-move randomly
-static

Misc:
-is blocked by world border / can wrap
-immune to projectiles

The presentation will be mostly abstract with basic shapes and colors.

Any other ideas?

Any feedback is welcome, of course

[Graham.]

I'm experienced with generation problems.

Ok, your design isn't restrictive. It is okay. I recommend thinking about what you want to generate. Lack of clarity is the cancer of all gen projects. Here is a good example: build a level by hand, then decide 2 things you want to generate, say the enemy spawn patterns, and maybe the way colors change on the screen - such as bullets, explosions, enemy sprites etc.

You want to pick only 1 or 2 things to generate at a time. You can pick level layouts, momentum of the PC, whatever. Just pick 1 or 2. Then test it out, then ask if you want to generate something different or better.

[Gimym TILBERT]

Look for "angelina" in the devlog
seek for "rom check failed"
Hunt for joris dormans thesis
take inspiration with togelius work
Roguebasin might help a little too

There is obscure academic thesis too, but right now I need to check further.

I'm starting this too, but I start with game idea generator and then swap after for actual game (but not too soon, the goal is usable quest generation first, story second, world third, everything later).
https://dl.dropbox.com/u/24530447/flash%20build/GameIdea.html