Post-GDC: Spore: Pre-Production Through Prototyping

Maxis Senior Development Director Eric Todd shifted foot to foot as Namco's Keita Takahashi slowly gathered up his notes and folders, grin plastered to his face, slowed by the occasional autograph hunter. It seemed like every time Takahashi thought he was ready, he realized he had failed to retrieve something else. Eventually he cleared off the podium and exited stage left. Just as Eric Todd stepped forward, to belatedly start his lecture, Takahashi swooped by again to collect one last article before dashing to the hall doors, seeming suddenly preoccupied. Todd blinked at the audience and introduced himself.

Heart of Virtue, Brain of Steel

"Prototyping", Todd declared, "is the heart of a virtuous pre-production cycle". He explained the premise of the lecture – that he would be discussing the value of experimental models before dedicating one's self to any one approach to a software problem. He then explained that the following would be an "advanced" talk, that would assume you already knew what he was talking about – so he wouldn't hold back in his explanations or references. Todd rattled off a list of books that the audience might do well reading, to better understand what he was about to say – none of which, it turned out, were altogether necessary.

"What is pre-production?" Todd's slide asked. It answered itself: "what you are going to build" and "how you are going to do it". To put it most simply, pre-production is a working sketch of the design concepts you have buzzing in your head, that you can throw together to see if they really work and to better illustrate the ideas to the rest of your design team – much as with thumbnail sketches, storyboards, or the "pre-viz" animatics used in Peter Jackson's Lord of the Rings trilogy. Pre-production is a form of planning before actually putting expensive pencil to paper, as it were.

The audience for pre-production includes studio engineers, designers, and other team members. Most objectively, a prototype is used to convince others that your concept is "worth the risk of a full production".

Todd explained that the benefit of having a working mechanical model at hand is that, as far as communicating about design concepts is concerned, "words are fundementally a terrible way of communicating interactivity." Unless you are assiduously careful, and even if you are, it is easy to get "trapped in rhetoric". As the old axiom goes: show; don't tell.

Kinesthesia Thesis

Demonstrating the concept, Todd showed a utility that allowed an amorphous worm creature to be prodded, deformed, and manipulated – he demonstrated how it felt to mess around with the utility, which seemed pretty tactile. As Todd said, having a tool like this "short-circuits" an inane conversation. Another benefit is that down the road, with a good deal of refinement, that particular prototype became a real editing tool for the final game.

It's important to know from the outset, Todd said, what problem a prototype is intended to solve. He showed a four-pronged chart: two lines in a skewed cross shape, each of four ends labeled – "Kinesthetics", "Aesthetics", "Technology", "Game Mechanics". For this particular prototype, as it existed to allow team members to get a feel for mooshing around an in-game object, there was a thin spike radiating out from the center of the cross toward the "Kinesthetics" end, with a tiny bump also protruding in the opposing "Technology" direction. Todd defined Kinesthtics as the "feel" portion of the "look and feel" of a game.

Look Polished

Todd then went off on a little tangent about the importance of polish. He showed a little utility for keeping track of the statistics of various in-game entities, explaining how before so-and-so scrabbled up this program, everyone just kept track of things on a spreadsheet and how that had never worked very well. In contrast, the prototype – though humble – had a degree of user design put into it. Objects were depicted by clean, round-edged boxes that became highlighted when clicked. There was some minimal work put into a color scheme and to get the whole package to feel smooth and friendly to use. Though you don't want to obsess, Todd said, "a little aesthetics goes a long way" toward making your prototype something that people will actually use. He called this minimal kind of design a kind of "coating on the aspirin"

The Model Ain't the Jet

Along similar lines, prototypes should be as focused as possible upon just the few things you need to demonstrate to someone; Todd warned strongly to integrate different prototypes or different elements to be prototyped only when absolutely necessary, as each time you start to combine things the prototype takes on a new dimension of complexity. He showed a sequence of prototypes where the Spore team effectively built the game before they built the game, to see how it would feel. They actually put some art design into the project. It was huge, it took forever to build and to compile, it was expensive, and ultimately was wasteful as none of the material that went into the prototype could actually be used in the final game.

That is, of course, part of the point of a prototype: "When you are in pre-production, you're not making the game." If you forget that, and try to save effort by making sure everything you create is to a level that you can just dump it into the game later on, that creates a high-friction development environment.

For the sake of simplicity, the team later moved to an incredibly simple 2D prototype that looked sort of like a colorized Vectrex game. This new model took one person a month to build, and in some senses it gave a better sense for how the game might work. Yet still, all is not rosy. "It was great, you know – but it still gives bogus results." They put a great deal of work into implementing and tweaking a fight system that simply could not be refined at the level of prototype they had; that energy would have been better used elsewhere.

Indeed, when someone finally threw together a prototype to show off the battle system and its animation style, he hot-wired the utility with PlayStation controllers and unveiled it during a staff meeting so as to delight the team and get them fired up about the project. Of course, there was no practical need for the Dual Shocks; there are no plans to place Spore on any console, and anyway users aren't really meant to interact with the game that way. That didn't matter; turning the battle system into a free-for-all brawl created energy. It led to people feeling more ownership over the project and to have more fun with it. The prototype also effectively educated the team what that game component was like, giving them a further impression for the way the game as a whole might come together.

Ideas in Traction

Going back to the early model-prodding utility, Todd explained how it had come about through an idea that a team member simply could not express in public. Every time that a team member tried to explain the process that the prototype demonstrates, each team member in turn looked at him as if his head were on upside-down. So to prove his point, that team member put together a rough 2D demonstration of his concept. Though imperfect, when he brought that prototype to the next meeting, everyone started to get a vague feel for what he was talking about. They suggested he refine the model a little more, resulting in the 3D editor which now a big portion of the game is built around.

As long as team members are kept reasonably reined in, and aren't spending all their time chasing windmills, Todd figures they should be encouraged to illustrate their ideas however possible – artists and development staff alike.

_____________________________________________________