I’ll start by explaining a bit of what my game plans are: it will be a platform-adventure (“Metroidvania”) game that will involve the player heavily interacting with his environment (manipulating computers, dodging automated weapons, etc.). Also, the character will have quite a bit of skills of his own, which may or may not work depend on what environment he’s in. This is the first time I’ve tried something like this, and I’m starting to run into a chicken-and-the-egg conflict. I’m trying to design how the player interacts with the level, while also needing to design how the level interacts with the player.
They’re quite intertwined and I was wondering if designing one part was easier in the long run than designing the other. I’d like to find out now before I get in too deep.
You don’t have to be an expert on it – just tell me what you guys think.
Well, probably the best advice I could give you is to design a prototype of this game. Use a capsule for your main character and some cubes or very basic 3d models from your 3d modeling application (untextured unless they HAVE to be textured, the geometry is only there for what is literally required for gameplay.) When you make this simple prototype, focus on doing one thing at a time. For example, first do movement. Then do jumping. Then do a weapon (your game might not have weapons but this is just an example). Then you could do one of the environmentally related skills that your character has, once again only if this is mandatory for the core gameplay. Design a simple level that gets to the heart of your game and finish it to the extent that it is playable, dont worry about visual appeal or anything like that. After you design a prototype, for one you will know what different environmental abilities will work for your game and which wont and scrap the ones that don’t. Then, when you reach production, you should be able to figure out what needs to be expanded, and you should switch between the two as your needs change. One tip is to use prototyping to design levels as well, then almost as a last step, actually model the level. Your underlying prototype can still be used for a large part as collision data and scripting.
If you’re not talking Gameplay, and thinking more along the lines Concept Design then you should design them together.
Having a quick prototype/game level with rough placeholder assets previously to this stage is critical (as Unfathomable says.)
Examples - How big should your character be in relation to a building you move inside of. You need this done before you create/design the real game assets. How high can your character jump? So you can model platforms to fit. I mention the building size because most people first make them to scale which look very claustrophobic with game engine cameras, and 3rd person cameras always collide with walls. They end up being 25%-50% bigger than real life. This is why people make prototypes
Use concept art and storyboarding first. That way you have an easy way of previewing the scenes/story/gameplay from paper.
(all professional studios do that, and for a reason)
Than if you’re happy, use the concept art to make the assets (character, environmental stuff, textures, vehicles, houses, whatnot)
then import them, fitting the concept/storyboards, as a staging, and start adding behaviors (like controls, particles etc etc etc)
part by part… and test much in between…
Thanks guys. Just to be clear, I’m aware of the design process (so to speak). I’ve already drawn 40 props, and I’ve documented a few of the skills that I want the player to be able to do. The problem comes when I’m thinking “what skills should I give to the player that he can use on the props I’ve made” and “what props should I make that the player can use his skills on” interchangeably. It gets confusing after awhile.
What kind of baffled me was which component (characters or levels) I should run through the design process first. I guess I could just try designing them together, and seeing how well they mesh.
I personally design my environment and character together. Concept Sketches are key and so are storyboards but there is no “must do” way. I believe in SCRUM therefore I believe in rapid development. This means a quick couple sketches of the environment followed by hours of reference images. Gather images that are relevant to your goal. This may take a day or just a couple hours. REMEMBER: While originality is everything, you are only as good as your reference library. Which means the more the better. Any experienced level designers here know what I mean.
Play games relevant to your own idea and take notes. Not just visual notes but identify how they achieved (faked in some cases) the end result. Study the mechanics and level structure. Study the level flow, scale approach to design and means of cover. Game Design is more about psychology than entertainment.
When you feel like your ready then dive into your 3d app and create a quick block mesh and drop it into Unity. I recommend a low poly version so that you can play with the lighting. Lighting sets the mood so never do this last. Prototype this as well and play with shadows bouncing off of objects and player visibility. Again DO NOT get married to your initial idea. EXPERIMENT and be prepared to scrap everything. You have a community full of people willing to test ideas, take advantage before making it final.
As for textures don’t be afraid to be different but all the image effects and whistles cant save you from bad textures. When you think of textures think of lights, when you think of lights think of geometry and when you think about geometry think of placement, lights and textures. Sounds like a riddle now but its a must. During all this i am still collecting reference images for props, lights and gathering test textures.
There is a little more too my process but you should get the point from this for starters. Remember the best way of doing things is whatever is most productive and logical.
