Game design process: Project OTD

[rogerborg]

Get someone to make you some models and sounds .. thats more important than any amount of code.

I could not possibly disagree more. Well... perhaps a bit more.

A game should be fun regardless of the visuals. Pacman is still a fun game, as is Tetris. There's no reason that a game designed in 2008 should aim for a less enjoyable core gameplay experience.

Sounds are important, if they contribute to the visceral experience.

Visuals... eh. Whatever you produce is obsolete as soon as it hits the screen. My prototyping (should I go ahead) will be done with decals; if the game isn't compelling when played with sprites, then I might as well abandon it.

[ebo]

Thats perhaps a bit influence by my nearly total absence of modeling skills

A game should be fun regardless of the visuals. Pacman is still a fun game, as is Tetris. There's no reason that a game designed in 2008 should aim for a less enjoyable core gameplay experience.

Well .. you arent planing to do Pacman and game design has obviously changed a lot since 1980.
As you are not into coding yet: Graphics is a _major_ part of game design. Not only does it attract attention, it also helps you getting a feeling for the story behind your game.
Of course getting your models done is a iterative process, but it really helps if it looks nice.

[bitplane]

I disagree ebo, myself I plan to make the best RPG ever using mostly billboards. I'm playing Daggerfall at the moment and it's an order of magnitude better than Oblivion, I've also been save-scumming my way through Nethack- using the text only client of course. If my game is half as good as those then I'll be one happy man.
This whole interactive movie fad has completely ruined the textual depth of games imo, kids can't be arsed reading or thinking I suppose. I used to imagine the 2000s would be the days of interactive characters and dynamic procedural environments.. instead we got countless hours of voice acting and FMV cut-scenes, and endless boasts about the quality of physics, animation and hand painted normal maps.
One time a friend and I made a chess board from strips of green and white paper towels weaved together and stuck down with sellotape, made in under 3 hours and had 100 hours of play before it wore out.
No style, all substance, that's my kinda game.

[rogerborg]

I'm enjoying this debate, by the way, so please take this as a robust but respectful rebuttal: graphics are largely an irrelevant distraction during design and implementation.

"Attracting attention" is a marketing problem, not a design issue. Since I plan to DIY the entire project, I'm a long, long way from there.

A story is only relevant to design if it has functional consequences. HUD and UI mockups inform design decisions. Modelling of actors does not, unless there is something about that representation (e.g. size) that needs to be accounted for. If a visual triggers ideas ("Hey, that pointy ship looks like it's made for ramming. Let's implement ramming."), that's fair enough, but it's accidental and incidental.

I think we're using different definitions of 'design'. I'm using it in the purely software context, and I'll explain why.

A game can and should be fun when played with simple sprites or blobs or, yes, even text, as in the Nethack example. The yardstick of good gameplay is that it's engaging despite having primitive graphics.

The wonderful thing is that visuals can always be added on top of gameplay. If the core gameplay is sound, then when the sparkle is added, you've got a winner.

But if you start with final models and effects, then much of your effort inevitably gets channelled into showcasing those models. Good models can mask unengaging gameplay, but not indefinitely, and only to the detriment of your final product.

I'm not asking you to believe me, just explaining why I actually prefer working with programmer mockups until code complete. For one thing, having a playable game with awful graphics would seem to be more likely to attract volunteer modelling resources than just pitching a concept, then basing the game on whatever few models I was able to scrounge.

[ebo]

As I said: I'm _not_ talking about a final models... Just better than what i could produce myself.

In retrospect i have to say that being unable or at least feeling uneasy about making models was one of the main reason i stopped working on irrGolf (see sig). Not the only reason, of course, but for me the most important.

As soon as i think, i got a really nice idea for a game, i ask myself: Ok so i'll do all the programing work but it wont work without some nice models... Who will make those...
And that usually ends up in me dumping the idea. The only thing i'm worse in than making models is producing sounds...

Perhaps i'm a bit visually assessed but i like it, if it moves and looks nice. And if it has a nice UI (meaning: I can easily find all the places, i need) a game has a very good chance, that i'll like it.
Of course gameplay is important, but come on .. HL2 wouldnt be half as good, if it didnt look so awesome.

Edit:
Perhaps I went a bit on the software engineer side. If someone gave me 1m $ (nah better €) and sayed: produce whatever game you like. I'd start in making documents about gameplay, ui, concept art, graphics, etc... Programming everything _is_ important, but it's sooo much easier with a good _written_ plan.

[BlindSide]

3) I get bored or distracted by something shiny.

That's my achilles heel.

I voted for StarShips because I like the intertwining themes between space and the open seas.

Namely things like pirates, self contained nature of the vessels, the similarity of planets to islands in an endless sea. SpaceBoats, SpaceKayaks, SpaceAquaBikes, SpaceJetSki, SpaceErrDucks?

Good luck with this, and watch out for that third one.

Get someone to make you some models and sounds .. thats more important than any amount of code.

Ok you've grasped the fact that game design is changing, but did you know that the entire creation procedure has changed significantly throughout the last few years. There is no longer a distinct line between "Artist" and "Programmer". Programmers are being forced more and more into expressing their creative side.

Procedurally generated models, procedural sounds, you name it, nobody actually needs models anymore (OK OK I'm half joking here), it can all be done in code if you want.

Look at things like Julia Sets, procedural terrain and wood, metal textures using perlin and gradient mapping, procedurally generated cities, hell, entire planets are made out of thin air these using a bunch of well aligned vertices and a more sophisticated version of rand().

The code IS the art, it's everything.