Adventure Architect: Part Nine

Building a working adventure game world

(Or, everything goes wrong, but I charge ahead anyway)

You know that saying about the best-laid plans of mice and men? Like, how they often go awry? Well, I would hereby like to officially nominate amateur adventure game designers in general, and myself in particular, to our rightful spot next to mice and men in that vast pantheon of those whose plans don't always work out according to, um, plan.

So, it turns out I've made a few missteps as project coordinator for this, my first foray into adventure game design. Not fatal mistakes. Not project-killing flaws, by any means. But the boneheaded, what-the-heck-was-I-thinking (or wasn't I thinking?!) variety. And those oversights only became evident to me after I started to plug all of the game's various design elements--backgrounds, objects, animations, music, sound effects, et cetera--into the Adventure Game Studio engine to try building a functioning game environment.

You see, with the story and plot now complete, the puzzles for Acts I and II finished, background work and layouts underway for all four acts, and a top-notch animator finally on board, it was time. Time to test out the game engine. Time to stretch its legs a bit. Time to really get down to business.

And then all hell broke loose.

Because, it turns out, I'm an idiot. I waited too long to firm up the screen size and corresponding character sprite size, and having never made a game before, I hadn't even considered the issue of exact proportions. Certainly I knew that the character sprites had to be relatively proportional to other objects on the screen, but I somehow skipped the step where I actually measured in pixels what those proportions would be. I was, in retrospect, just relying on the game engine's ability to scale characters as a catch-all for anything that I couldn't simply eyeball at the right proportion.

Bad idea. The kind of idea that probably gets professional game producers fired, in fact. Fortunately, I work for free, and I hired myself, so that's at least one disaster averted.

As you'll see below, the very first screen of Act I--which has been my pride, joy, and inspiration for lo these many months--has two problems that I didn't notice when we turned our layouts into the finished screen that I planned to use in the game. First, the signpost in the middle of the path is far too large. Monstrously so, in fact, when compared to a reasonably sized character sprite. And second, that same signpost in the middle of the path is literally in the middle of the path! Duh, Josh. This makes maneuvering around it infinitely more difficult than necessary.

Figure 1: Jake discovers he's a small man in a very big world.

I ran into similar problems on other screens as well. In one, for example, the character sprite for Jake is required to walk across a railroad bridge looming over a large canyon. Unfortunately, the gaps between the boards were so wide that Jake appeared to "float" over the empty space when, technically, he should have ended up falling tragically to his death. Another screen found him completing a puzzle that required him to build a ladder to reach a ledge that, all things being equal, was pretty much about shoulder high.

In all, these proportion issues affected about 10 screens. Of those 10, half were easily fixed with PaintShop, a few are currently undergoing the kind of massive restoration usually reserved for fading Renaissance paintings, and two are just plain unsalvageable. It could have been a lot worse, though. I was able to immediately put the brakes on all background screen production, work with the artists to get the correct the proportions on each layout sketch, and resume progress without missing a beat.

Figure 2: Gone, but not forgotten. We'll add a more reasonably-sized signpost as an object before this screen is finished.

So what did I learn? Well, for one thing, I should have decided on the screen resolution and character sprite size beforehand. In fact, I probably should have created a sprite or two at the very beginning of the process in order to test each screen before moving it into final production. I also discovered that I have a tendency to squeeze too many things into each screen. If I'd kept those screens as I had initially conceived them, I probably would have had to scale down the character sprite in order to squeeze him into the scene.

Fortunately, I caught my mistake before the project moved too much further along. The end result, then, was a brief moment of panic followed by a comprehensive review and assessment of every game asset. Basically, a speed bump instead of a train wreck.

And production continues.

Next: Anatomy of a hero!