I wrote this last year during GDC. In essence, scribbling these impressions was the genesis of this site; I wrote this stuff up and had no where to post it. These are from the hip, shotgunned impressions written day-of, so excuse me if they’re a bit wide-eyed. As I prepare for GDC 2006 (next week) I thought it would be fun to post this, raw as it is. My plan for the conference this year is to do essentially the same thing with a daily ‘dump’ of my notes and thoughts. Hopefully, that’ll be useful in some capacity. As usual, I’ll be focusing on insights from master game designers, practical tools for game innovation, and things that seem noteworthy innovation-wise: anything that might help someone design an amazing game.

Without further disclaimer:

Thoughts, feelings, and inspiration (at GDC 2005)

GDC 2004 seemed tinged (again) with quiet, desperate futility. The theme of ‘sequelitus’ was now, by my count, on its third sequel. The industry was farther along the same crash course with homogeneity, driven by budgets and faceless Suits, led inexorably forward by Electronic Arts, the televangelist of the game industry. The Gospel: bigger, shinier, and more ‘to the X-Treme’. And hey, wow, now with crassly hip (hop) urban street flair.

According to the new religion, innovation and creativity are features to creep, a bullet of marketing copy, shelf space and buzzwords. Fewer publishers funding fewer titles made for far more money by fewer developers (most of whom are now owned by said publishers). Same old same old, essentially.

Enter GDC 2005.

This year a hand reached up and tried to jerk the wheel. Whether or not the course has changed is not easily judged, but one thing is certain: hope has returned to the game industry.

The Future, Period.

Or ‘What Will Wright Has Been Saying for Years About What he Learned About Content, Community, and Game Design From The Sims
…And Why Everyone is a Moron for Not Listening
…And How Procedural Methods Will Save the Game Industry’

If the Hand That Jerked the Wheel belonged to anyone, it belonged to Will Wright.

Walking through the conference center after Will’s ‘lecture’, the excited shockwaves rippling through the industry were almost palpable. Warren Spector was, as I understand it, somewhat dejected (having taken a stance on the story/Hollywood/data intensity side of the fence, he was seen hanging and shaking his head saying things like ‘how can I compete with that?’). I was elated, inspired, and full of hope.

Will started with a few words on storytelling in games, with which I whole-heartedly agree but had not previously heard articulated as such. Successful story in games, he said, is about ownership. Ownership in general adds huge amounts of value to player experience, as evidenced by the meteoric and enduring popularity of The Sims. Specifically, he illustrated the simple difference between watching a pre-rendered cut scene and one in which the player can substitute a customized avatar. The example he used was his customized character from Grand Theft Auto: San Andreas, a character who wore nothing but cowboy boots and a diaper with tiny hearts on it. He still spoke gruffly and the other characters in the cut scenes reacted seriously, but the juxtaposition of Will’s farcical avatar altered every scene, creating something completely new.

For my part, I think this ownership extends in many different directions and can surface in any number of different contexts and types of games. ‘Story’ in games can be “remember when I single handedly won that round of Counterstrike by hunting down the remaining six counter terrorists using only the Desert Eagle?” Alternately, it can be things like “I once played a game of Civilization in which I had effectively ‘lost’; I had no chance of completing the game as the foremost civilization but I continued playing, a virtual Switzerland, clinging to life between two warring superpowers. It was fascinating to set a goal not of conquering, winning, but simply of surviving.” Any gamer has a number of these stories, all deeply personal and wholly engaging. For many, these remarkable experiences are the reason to play. Personal stories are, ipso facto, owned.

Also, I find the most successful instances of story in games to be in those games which allow the player a blank emotional canvas. My favorite of these games and the one I consider the most successful is X-Com: UFO Defense. Faces light up at its merest mention, faces like that of Daniel Greenberg. It’s not an experience I can easily share linearly; your best bet is to track down a copy and experience it yourself. Suffice it to say that X-Com is algorithmic, procedural in nature, and provides not only for emotional projection but for character growth, change, and evolution. This evolution exists primarily in the player’s mind, unfolding as certain favored characters survive missions, grow statistically, and are subsequently showered unabashedly with the finest armor and weapons the player can muster. I often find myself imaging how the other, less festooned characters feel about this, and playing out pithy banter between grizzled X-Com commanders and their alternately cocksure and petrified recruits. This is something Wright has said many times in reference to The Sims and its popularity: leaving game characters intentionally blank, emotionally ambiguous, allows a designer to offload emotional processing to the player’s mind, which is far more efficient. Let the computer do what it’s good at, crunching numbers, and let the player do what she’s good at, finding emotion. This is, of course, brilliant (though it’s nothing Will hasn’t said before.)

Will next launched into an explanation about the balance of data and process in games, which he distilled to simply ‘code and data’. The crux of the concept is the idea that while every piece of software uses a mixture of code and data, programs that emphasize code (process to use Crawford’s wording) are fundamentally superior. In games, code defines the verbs, the dynamics, the processes, what the player does, and is represented by algorithms, equations, and branches. Data is things like images, sound, and text, stored by the computer as bits. Process, of course, is the strength of the medium of games and of interactivity in general. Because the computer has the ability to crunch numbers, to process at unprecedented speeds, it can respond seemingly instantly to user input. In the same way a series of drawings played past a certain speed crosses a threshold and becomes a dog jumping, a man running, or a woman softly weeping, a computer returning positional data in the form of a constantly refreshing display becomes an avatar, an extension of will and a form of personal expression. The key difference is ownership. So, more dynamics and interactivity provide more ownership and expression, the fundamental strength of the medium of computing. To use a computer to display data is to ignore its fundamental strength, processing. That’s the gist of it.

In the first computer games, data was kept to a minimum because all the available resources were being used by code. As the years, games, and expectations have progressed, data (content, to use the eponymous) has swelled to biblical (epic) proportions, while code has remained relatively static. Bit of an odd role reversal, but there it is. This swell in the demand for content is what’s led to the much-lamented rising budgets, risk adversity in publishing, and painful lack of innovation. So, Will has decided to step up to the plate. “I’d like to offer an alternative.” he says, then drops a bomb on the game industry. On conventional wisdom, anyway.

Spore

Moving on, he arrived at the ‘and what I’m going to do with procedural methods’ part of his speech. Basically, he’s solved, with blistering elegance, most of the problems of developing content for next generation hardware. For an encore, his carefully assembled team of ‘algorithm ninjas’ are well into creating what is very likely the most ambitious game design of all time. Not only that but it appears to me as though they, in approaching this most monumental of tasks, decided to tackle it head-on. Based on what they showed, it seems as though they made a list of all the technology they would need to create the game, prioritized it by what didn’t exist yet, what would be the most difficult hurdles to overcome, and set to work. And, boy, is the work impressive.

Technology

In terms of technology there were a few things that bent my mind sideways. Now, I don’t consider myself a technophile; I think that tools are tools and are base and inert without creativity to wield them. The thing that wowed me was the overwhelming evidence that creativity was indeed being given primacy and though the technologies being shown were totally new and impressive, they were created with a very specific purpose and were being wielded with the proper spirit. Technology was, for once, not wagging the dog.

The various creature editors, especially for the land-based creatures, are phenomenal. As Will promised, the creature editor allows for intimately personal expression. Ownership on an unprecedented level. One major interaction was with the spine, the ability to add or subtract vertebrae and to bend, pull, or resize the spine to taste. Of course, this is the best way to create a customized creature, to start with the skeleton (especially in light of the way that the most effective way to animate a 3d mesh cheaply is to deform it based on bones.) All of this seemed extremely fluid and intuitive as Will was pulling and bending the spine around from straight horizontal to a curled up kind of C-shape. An interesting note, as you’re pulling and moving the spine, the mesh was perfectly interpolated around it, compensating for tearing and other undesired effects. Also, as Will noted, the mesh retains knowledge of its topology as it’s being deformed and modified. So, for example, no matter how you contort the spine the texture on the ‘belly’, wherever it ends up, will always be a different flesh texture and fade nicely into the rest of the body. And, of course, all the textures are procedural, so they don’t really stretch or distort.

Next, Will started dragging limbs onto the creature. Now, this is huge. The mesh was perfectly interpolating and recreating itself as he moved it around. I could see the tri-stripping and polygon positioning because the body of the creature remained flatshaded in the editor, and it looked excellent as well as economical. This, however, was not the really impressive part. The really impressive part was when he brought the creature into the game and it walked. Perfectly. A perfectly interpolated walk cycle with no skating, no jittering, no nothing. Perfect weighting, perfect timing (slightly offset and ragged gait because the creature had only three legs), and because it was a purely interpolative walk cycle it interacted perfectly with the ground, meaning it walked perfectly up and down any rough terrain without clipping or any otherwise noticeable artifacts of having just been interpolated from an on-the-fly skeleton. Also, the interpolation appears to take mass into account. For example, a huge-headed creature with tiny legs leaned and titled precariously as it scuttled around. He then showed a creature that, despite being a weirdly tangled mass of legs, had produced an extremely fast gait. Presumably, this improves the creature’s ecological fitness significantly enough to have serious impact on the gameplay. I can see myself playing with the creature editor for hours, trying to stumble on the ultimate creature. What was so striking was that the technology, while obviously not ready for mass consumption, was complete. As far as I know, this technology did not exist before Spore. It seems to simultaneously be an animator’s anathema and her wildest dreams come to life in the most spectacular way imaginable.

As a random aside, I have it on good authority that the character editor is very robust. Apparently, an intern at Maxis (not the authority in question) was given the task of trying to ‘break’ the editor by creating the craziest creatures imaginable, taking things to their most ridiculous extremes. The character editor is, by way of these tests, able to make even a creature comprised of nothing but 50 spiraling legs and a pair of eyes walk and move convincingly. Yeah.

Game Design

I feel like Spore gives us the right to dream again. Will said that the only barrier to creating Spore was his imagination: “Once I believed that this game was buildable, it was quite easy to convince my staff that this game was buildable.” I can understand why this would be so, seeing as it is quite possibly the most ambitious game design ever. It defies conventional knowledge. Specifically, the idea that you can’t include tons of different game types in one game. I don’t know that I’ve ever really agreed with that idea, but I think there is great wisdom in not shouldering such an undertaking lightly. Many have tried, most have failed (to be blunt about it). Then again, games like X-Com show such alluring promise, and ‘most’ are not Will Wright (plus the resources he commands, of course.)

Right, so, the important questions are these: can they pull it off? Will the game be a success? And, perhaps most importantly, can anyone but Will Wright use these methods to make games?

I do not envy them the task of balancing, but it’s possible. The estimated release date, despite what they showed, is 2007. Looking at what they have so far, that sounds plausible, if optimistic. I’m told The Sims’ gameplay wasn’t balanced properly until just before its release. Right, so, what they showed included each phase, each piece of the proposed game, but did not have the ‘glue’. As a developer, I know that it’s impossible to really understand the magnitude of what is left to do on a project unless you’re in the trench, staring at the code. So, I don’t really know what’s left to do on the project. What I do know is that there were no transitions between most of the various phases. It’s unclear to me how the game transfers between the 2d and 3d phases, for example. Will would ‘skip ahead’, loading new phases. Lots of work to be done there. Also, the ‘SimCity’ phase and ‘Civ’ phase seemed functionally inchoate. In particular, I could discern little pattern to the movements of the creatures in the cities and no particular reaction to the buildings Will was creating or indeed anything: they were clipping through the buildings as they wandered around. I noticed that Chaim’s bending roads made it in, though . Also, though the vehicles driving around obviously had some behavior – they moved to attack Will’s city – their attacks had no effect and the citizens of Will’s city seemed not to notice them.

All that said, it looked to me as though they’d made a list of all the technologies they’d need to create the game, prioritized them in terms of the most difficult or unproven, and set to work tackling the hardest first. Bravo, guys. All the pieces are there, and the technologies that did not previously exist have been created. Now all that remains is a shit-ton of work. Nitty gritty game design-y work. As I said, I do not envy them this task. But, hey, if you’re going to design the most ambitious game ever, you have to be prepared to roll your sleeves up.

To speak to my last question (can anyone but Will Wright can make games this way?) my answer is ‘of course they can.’ On the scale of Spore, perhaps not, but procedural methods for content creation make sense economically and that will cause them to bubble up to the surface. I mean, hey, EA’s using them.

Convergence

So, what’s really amazing about this whole Spore thing is the convergence of all these wonderful ideas into one vision. There are so many pieces at work here that it takes quite a bit of brain-space to really put them all together and share that vision, but once you do it’s quite a thing to behold. The fusion of evolution in content, community, and game Design puts Spore so far into the avant garde it’s practically a window to the future. The games I’m working on look like scribblings on cave walls in comparison. If they pull this thing off, it will be the future. Period.

…I can’t wait to see what Will has to say this year after another year of production on Spore .

-Swink

1 Comment