By Steerpike
April 2004

Thanks, Bill

The above title is not a Steerpike original, much as I'd have liked it to be. It's an homage ("homage" is the word we use to escape the phrase "blatant plagiarism") to a column on the same subject penned years ago by William Trotter for PC Gamer magazine, and since it is so accurate, it deserves use again.

The annual Game Developers Conference was held in San Jose for the last time this year, from March 22–26. With up to 12,000 attendees, the San Jose Convention Center was stuffed to the gills, and as such, Conference host CMP Media will be moving it to larger digs in San Francisco starting next year.

The GDC is not a Gen Con or an E3; it's really not for the general public nor would the general public—even the gaming public—find it terribly interesting. Anyone is welcome, but the prohibitive cost of passes and intense nature of the conference itself keeps most casual attendees at bay. It is five nine-hour days of lectures, tutorials, roundtable discussions, business cards, alcohol, keynote speeches, and development tool exhibitions. It's also the best place in the world to get news on upcoming trends and shake hands with people in the industry.

I write a monthly column for the International Game Developers Association and consult with various game companies, which is why I went. Anyone interested in attending next year's GDC in San Fran should check out www.gdconf.com for the details. Part of the point of the GDC is that you have the opportunity to meet literally everyone in the business, so if you're looking for a job in games, it's worth it to visit and hand out some cards.

It may be intense, but it's also a lot of fun. The people within this industry are friendly and accessible. This week is our opportunity to do nothing but talk about our passion, and as such, we tend to be a cheery bunch while we're there. Newcomers have the opportunity to meet venerated designers such as Will Wright or Chris Taylor in a casual setting (and beg them for jobs); grizzled veterans have the opportunity to sit around drinking beer with other grizzled veterans and mumble about how things were back "'afore games had all these colors and sounds in 'em."

Microsoft dominated the event, as Microsoft loves to do, but its openhanded presence was not unwelcome. Everything from the lobby bar to the lobby steps was festooned with Xbox ads, new technology initiatives, plus a few tantalizing banners about what we can expect from the upcoming Longhorn operating system.

The end of every day means very painful feet, a very sore back, a legal pad full of notes, a brain full of exciting new ideas, and a complete misapprehension about the time of day. Especially if you're from the east coast or, worse, one of the poor fools brave enough to visit from Europe and Asia. It's exhausting, but in a good way.

Summits and Superpowers

The first two days of the GDC encompass high-level tutorials and major academic issue meetings, such as the IGDA's Business Summit and the new Serious Games Summit. These multiday events feature several speakers, roundtable discussions, and real-life problem-solving activities run by various industry luminaries.

Two bad things are happening in gaming today.

One: young people, excited by gaming, come into the industry—only to be burnt out by the ridiculously low pay, ridiculously long hours, and often thankless relationship with publishers. As a result, the industry's best new talent has a habit of leaving within five years and moving into far more lucrative (if less enjoyable) technical jobs in the business and government sectors. This dilemma is generally referred to within the industry as the "Quality of Life" problem, and it's become a topic of major discussion in recent years. The IGDA Business Summit focused attention on the Quality of Life problem and various other nuts-and-bolts issues surrounding the industry as a whole. The IGDA's annual Quality of Life report, a conglomeration of all relevant research into industry recompense and satisfaction, is also unveiled at the GDC.

Two: games were then, are now, and threaten to always be viewed by outsiders as a bland and useless entertainment medium. Those educators who know anything about modern gaming realize that it represents a form of learning so drastically superior to every conventional educational method in existence that if leveraged properly, interactive "games-based" learning could foment a revolution in how human beings gather and interpret knowledge. People in the know, from business to education to intelligence to military, are very interested in working with the game industry to create electronic simulations capable of more than just entertainment. While this will by no means reduce the entertainment value of games designed for that purpose (any more than documentaries eliminated narrative movies), leveraging games for other roles is a potentially worthwhile subject bearing further discussion. That was the topic of the Serious Games Summit, a first at the GDC this year, and by all accounts a whizbang success.

The callsign for this year's GDC was "evolve," and it's more appropriate now than ever before. The game industry is changing, hopefully for the better, in several significant ways. Games are becoming more intense narrative experiences. They're being applied outside of pure entertainment. Their technology still represents the forefront of computer programming and development. If anything, the next few years will be less evolution than revolution, but "revolve" doesn't make as good a tagline.

But the real conference starts on Wednesday, when those holding Classic passes descend on San Jose in their thousands.

Three Trends

There are some 400 lectures and discussions available to passholders at the GDC, each usually lasting an hour. As such, chances are good that you'll miss at least something that you wanted to see because it overlaps something you wanted to see more; remember that most presentations and lecture scripts are posted at the GamaSutra Network sometime after the conference ends, so you can at least pick up the gist of a lecture you weren't able to attend.

Three philosophies of game design are coming to the fore in the next year or so: emergence, behavioral development, and scripted linearity. There were tons of lectures on each of these subjects, and even those on other topics generally made some mention of them. The good news is that these three philosophies, or at least parts of them, can coexist in the same game, so we're likely to see some interesting variation in upcoming titles.

Emergence is the new buzzword in the industry, and unfortunately it is too complicated a process to discuss in detail here. In a nutshell, emergence refers to a chain of causal events where the result is not immediately discernable based on the initial step or link in the chain. An example might be something like this: in Black & White, your creature can relieve himself anywhere on the map. Some effort in toilet training can ingrain in him the rule that he should always go on the villagers' wheat fields. This clod of protein-rich fertilizer causes the crops to grow faster, thereby increasing the speed at which villagers accumulate food. The more food you have, the faster the village grows; the faster the village grows, the higher your population; the higher your population, the more worshippers you can assign to your temple; the more worshippers, the more miracle power you can wield to rain fire and brimstone on your enemies. That's a linear chain.

But if you look at step one ("Your creature can poo anywhere he wants to poo"), it has no direct or obvious bearing on the final step: "You get more miracle power." By training your creature to go in the wheat fields, you have produced an emergent result. This trend drastically increases player improvisation and creative problem-solving, which in turn makes game worlds seem more immersive and real. Peter Molyneaux has opened the Pandora's Box of emergence with his upcoming Xbox title Fable; we'll see whether today's technology and design strategies are sufficiently advanced to accommodate his ambition.

Behavioral design employs the psychological discipline of the same name to invisibly direct player actions through a series of task/reward scenarios. Most people, playing Everquest at four in the morning, are waiting until they get "one more level" before going to bed. When they finish all of the tasks necessary to get that level (say, killing nineteen orcs), they generally stop playing. The trick in behavioral design is to create a network of overlapping goals, presumably requiring very different tasks, so that each reward the player receives isn't succeeded by the sort of lull in play that generally leads to the player leaving the game. We see this implemented in nearly every game to some degree or another—it's the classic "let me finish this map," or "let me beat this boss," or "let me get that item." As designers apply more psychological theory to their task/reward networks, it will become possible to engage the player far more effectively and, with luck, keep them playing longer.

"Linearity" has been a bad word in this business for far too long. Even I have railed against it in the past, but I see now that I may have been in error. People see linear gameplay as heavyhanded, closed to improvisation, and damaging to replayability. In truth, it's none of those things, as long as it is well implemented. Think about it: how many times have most people played through Half Life, a game so linear that you as a player quite literally have no choice in where to go or what to do once you get there? Yet no one complained that Half Life was linear because it employed narrative and scripted sequences so elegantly, and incorporated them into the gameplay so gracefully, that players felt they were having (and controlling) a deeper experience than they actually were.

This is beneficial to game designers because it allows them to retain narrative control, to tell the story they want to tell—which can be very difficult in nonlinear games. Morrowind is nonlinear and players missed huge swaths of story because their travels didn't lead them in those directions, yet most people are not replaying Morrowind because, though there's a lot they missed the first time around, the undertaking is too big and the first twelve or fifteen hours of the game are always identical. Warren Spector of ION Storm, a highly respected designer responsible for Ultima Underworld and Deus Ex, has moved toward this philosophy, eschewing aggressive nonlinearity because of the design headaches it causes.

All this is part and parcel of the argument that games are coming into their own as a narrative force. This creates a new challenge for developers, however: the narrative capability of current games is the equivalent of a child's coloring book. Very, very few story-driven games go beyond the most trite and formulaic narrative arcs. Time and again at lectures, I heard respected designers bemoan the fact that narrative games were still being written for teenage boys, which isn't acceptable when one considers that our audience consists of a great deal more than that. Great effort will be needed over the next few years to bring gaming narrative out of the Me-Tarzan-You-Jane doldrums they're currently in.

It Means XTREME-Ribonucleic Acid!

Microsoft scared me on Wednesday afternoon, not by flinging itself out of a dark corner and shrieking "boo!" but by encapsulating in a few sentences an alarming truth that I'd never really consciously put together before.

It was the Microsoft keynote speech, given by Microsoft Games honchos Robbie Bach and J Allard. All I had in me was coffee and a banana muffin from Starbucks, and I'd already been on my feet for like seven hours. That probably contributed to the yawning void that opened in my gut when Bach said something to the effect of: "Gamers are expecting top shelf titles with multi-million dollar budgets, professional voice acting, Hollywood scripts, shaded graphics, realistic physics, lifelike AI, and maybe sixty hours or more of gameplay. Compare that to what gamers expected in 1994 and then consider this: they expect all that, and they still expect to pay only fifty dollars for it."

It's no secret that most game studios operate perpetually on the razor-thin margin separating red from black. As gamers come to expect more, games cost more to make—but the product still has the same price tag. These days, a game that sells 400,000 copies worldwide is considered a success. 400,000 × $50 = $20,000,000. Twenty million bucks seems pretty good until you realize that the publisher gets most of that; studio royalties can be as low as 10% in some cases, and games rarely cost only two million bucks to make. Thus the studio has to get its publisher to pony up for development of its next game, which means it gets roped into another contract that grants it a lot less than it grants the publisher, and if its next game flops, well ...

Microsoft, actually being one of the fairer and more generous publishers in the biz, shouldn't necessarily be expected to try and help beleaguered studios out of this mire. And their solution doesn't necessarily help, but it doesn't hurt either; all told it seems more like a marketing ploy than a software tool, but hey.

Their new XNA initiative is a software "ecosystem" (marketing people ... geez) that gloms all existing Microsoft game development tools into one bite-sized package. Thus PC developers now have access to some of the custom development tools previously only available to Xbox developers, while Xbox developers can integrate more tightly with PC-only tools in their games. It's not like DirectX, not at all; that was a brand-new software development environment. XNA is really more of a brand name.

And no, it doesn't actually refer to "Xtreme-riboneucleic Acid," though I think it would be funny if it did; the "X" symbolizes the new cross-platform nature of the tools, and the "N" and "A" are short for "Next-generation" and "Architecture," respectively. You can check here for more details of the initiative, including some keen streaming demos.

Alas, Robbie and J were coy when it came to details about the Xbox 2, and even more coy when it came to the future of Windows as a gaming platform. Of course, they both loudly insisted that Microsoft was committed to supporting Windows as a gaming environment "then, now, and in the Longhorn!" (gah ... marketing people).

The cynic in me sees the increasing trend of clumsy console ports and the low price of the console systems and thinks that the PC may be in trouble. This warning has been issued before, and far be it from me to cry wolf, but if you look at 2003's crop of games and count the number of sloppy console ports that play like an ungainly postscript on the PC, it's kind of upsetting. Here's to the hope that Microsoft and developers alike bear in mind the relative power of a contemporary gaming PC versus a console.

The Oscar Goes To ...

Contrary to what Spike TV would have you believe, it is the Developers Choice Awards that represent the Emmys of our industry. The awards are extra special because, as the name implies, all voting is done by industry vets, much like the SAG awards. There's something deeply satisfying about being recognized for an achievement by your peers and colleagues.

Once again, the Developers Choice Awards were handed out with great fanfare and acclaim at the GDC. The show followed the Independent Games Festival, a wonderful recent edition to the week that spotlights and recognizes excellence in nonstudio games around the world. I'd say more about it, but I was late and arrived only in time for the last three minutes of the IGF.

I'll spare you a list of winners and tearful thank yous except to say that Star Wars: Knights of the Old Republic deservedly won Game of the Year and generally dominated the other categories as well. You can look here for a complete rundown of the awards show.

Why I'm Afraid of John Carmack

The enigmatic wunderkind behind DOOM and Quake has long been pursued by GDC sponsors to give a keynote speech, but until this year he has steadfastly declined. (By the way, if you're interested in learning more about the history of id Software and the personalities it harbored, you can't do better than David Kushner's wonderful Masters of DOOM for all the gory details). Apparently, pressure on his wife was the key to success, as this year Carmack tossed off a fifty-minute lecture on the benefits and drawbacks of bleeding-edge engine design.

And I do mean he tossed it off, which is why I'm afraid of him. He spoke without pause for fifty minutes, never once referring to notes or a teleprompter. He rattled off numbers and statistics from memory, with scarcely an "er" to punctuate his sentences. It was quite the spectacle.

The focus of the speech was on what's "done" versus what needs more time in the oven. Audio, he said, is pretty much complete in games, as is control structure. There's not improvement on the horizon for sound or the keyboard/mouse combo. Graphics, physics, and AI, however, still have a long way to go—especially AI, since developers of games tend to employ AI techniques that were not designed for gaming. More aggressive research into artificial intelligence for games, he said, would greatly improve the science.

I like John, though he and I disagree on some fairly pivotal issues, including the role of games as art and the future of the industry. His speech, however, contained nothing that I could really find fault with. It effortlessly summed up the benefits and drawbacks of staying on the very cutting edge and laid out Carmack's vision of the next three or four years of hardware and software. Considering that hardware and software designers produce their products at least in part based on what Carmack says, it's not unreasonable to assume that his forecast is going to be pretty accurate. He also assured us that DOOM 3 (which looks astounding—more on that later) will be shipping "very soon"—Activision is still saying mid-April, but I think the smart money should be on a May release.

In response to a question about the growing scope and cost of development projects, he said that it wouldn't surprise him at all if we see video games sporting hundred million–dollar budgets within the next ten years. That's kind of scary, especially when you go back to our earlier math lesson and consider that a game would have to sell two million units at fifty bucks to break even with such a budget. Plenty of games even today are played by two million people, but once you take into account piracy, rentals, and swapping, it comes down to just a couple hundred thousand sold. It will be interesting to see if Carmack's suggestion bears out in the next few years.

The reason for this is the increasing scale of game development. Carmack said that when they were developing DOOM, designers could churn out a reasonably good level in thirty minutes. It would need several hours of polish and balance before it made it into the game, of course, but a playable version could be done in half an hour. Levels in DOOM 3 and similar modern games take literally man-months to get into the most basic playable format—which is why we're seeing longer development cycles and higher budgets.

It was a very good speech, and an eye-opening one, especially in the wake of the Microsoft keynote. When two industry heavyweights that generally butt heads agree on such a pivotal issue, it's worth taking notice.

Better Living Through Chemistry

The Expo section of the Conference is open throughout the week and is composed of organizations that cater in part to the development industry: nVidia, Metrowerks, Havok, Microsoft, and, inexplicably, the State of Maryland were among those hundred or so with large, flashy booths. Since the Expo is sparsely attended during the day (most attendees are in lectures or at some other activity), the GDC includes a three-hour "booth crawl" period where there's nothing else to do but visit the Expo floor.

It's almost criminal how much alcohol is poured into Conference attendees by the various exhibitors during this period. Knowing that game developers are a pliable bunch when tempted with Sam Adams, thousands of gallons of premium beers were handed out by dozens of exhibitors on Thursday night, the only hook being that you had to see what it was they were pimping—and since it was usually something like the physics model in Serious Sam 2 or upcoming GPU architectures, it wasn't much of a sacrifice.

ATI and others were showing off the next generation of game engines, and they are a quantum leap beyond what we're seeing today. Finally, after nearly two years of finger-tapping, developers have been able to unleash the true power of the pixel shader engines that have until now lain largely dormant in the hearts of our video cards. Expect dynamic, movable lights, photorealistic environments, integrated physics interacting realistically with the game environment, genuinely curved surfaces, accurate body rendering and ragdolls—basically everything that we've been eager to see since the square blocks collided in Pong thirty years ago. Assuming you have the technology necessary to run the next gen games (by "next gen" I'm referring to the generation of which Half Life 2, DOOM 3, and Serious Sam 2 are members) with all the goodies turned on, you're likely to find your breath taken away by what you'll see.

Booth crawl over, the beer-fueled attendees staggered across the street in the rain to the big auditorium where a new GDC event called GameHotel was about to make its debut. No one knew quite what to expect from GameHotel; just that it had premiered to great acclaim in Paris. The ad in our Conference schedule said "You'll walk away from GameHotel with your brain bursting with fresh ideas and a deeper understanding of games as a driving force of today's pop culture." Not so much.

I hate to disappoint, but GameHotel was, frankly, moronic. I think it goes without saying that what works in France is unlikely to work anywhere else; GameHotel is living proof of that. Talking about games and game trends in a talk show setting is fine—if it's a talk show. Doing it in front of a drunken live audience that would much rather be looking at the various corollary exhibits, and adding thrill to the process by announcing loudly that So-and-So is "Checking in to the GAMEHOTEL!" every time a new guest comes along, then banishing them with a "So-and-So, your TAXI HAS ARRIVED!" is really just annoying. The various guests, who included Masaya Matsuura of Dance Dance Revolution fame, Phil Harrison of SCE Europe, and the entire Furi Furi Action Figure Design Company, at least had the decency to look mortified that they were part of such a contrived spectacle of nonsense, which really just made the whole experience that much more painful for the audience.

What could have been done in one hour took almost two and half—though most of the audience had bolted long before then—and this was after a long, long day of sitting in lectures and taking notes. All I wanted was a trip to the bar and some room service, but I'd made the ill-advised decision to attend GameHotel with my editor at the IGDA, who, though a wonderful guy, as one of the conference sponsors had a vested interest in the event's success. Abandoning the freak show in the middle, no matter how painful it was to endure, would have gone poorly for me in the long run.

Max Greed and Max Payne

The GDC isn't about showing off new stuff, though a few of the lecturers did manage to slide in some demonstrations of upcoming games, and some exhibitors were offering interactive demos of one or two high-profile soon to be released titles. Some of what we did see:

Peter Molyneaux used live demos of Fable and Black & White 2 during his lecture on gameplay innovation; both look terrific. Black & White 2, especially, may be a pleasant surprise to those who found the original lacking in direction. The game focuses much more on strategy and much less on endless management of spiritual growth. Visually, of course, it's a stunner—the new miracle powers are incredibly satisfying to watch. In fact, it looks to me that Black & White 2 is what Molyneaux wished Populous could have been, had the technology been around back in the day.

Fable has been getting a lot of press because it is such an ambitious experiment in emergent gameplay. However, I'd wait until a few reviews have been penned (and the early awe has died down) before buying this Xbox title. Molyneaux's games have a history of wild ambition and generally wind up biting off more than they can chew in the long run. It usually takes a sequel to iron out the flaws and produce some really compelling gameplay—sometimes a sequel that Molyneaux himself has nothing to do with. Consider the clumsiness that was Dungeon Keeper versus the grace and style of Dungeon Keeper 2. Anyway, Fable promises to allow the player to guide a little boy through his entire life, having adventures, meeting people, slaying monsters, and so forth, all in a living world with more or less complete freedom of choice. If it does work, then this game may be the reason some of us holdouts buy an Xbox.

We only got a glimpse of DOOM 3 at Carmack's lecture. Those who have seen—and been correspondingly flabbergasted by—the screen captures of the game really need to see it in action before they can appreciate how visually stunning the title truly is. I refuse to hand this game an A+ without playing it, but assuming that the graphics we saw in that short demo are representative of what a casual gamer can expect, then DOOM 3 will at least look good.

Thief 3, of which project lead Randy Smith showed screen captures only during his excellent lecture on emergent gameplay design techniques, is absolutely beautiful and looks like it will offer some very exciting new gameplay, but considering the ugly, ham-fisted abortion that was the PC version of Deus Ex: Invisible War, I'm going to withhold judgment. Watch for an upcoming Thief retrospective at Four Fat Chicks to whet our appetite for the future of the franchise.

Serious Sam 2 was used by some of the physics middleware companies, and by ATI, as a tech demo for upcoming shooter engines. The new Serious Engine looks good, really good, though not quite as good as DOOM 3 and Half Life 2 (which, along with Duke Nukem Forever, was nowhere to be found at the GDC). Still, most people play Serious Sam for the level of violence and irreverent humor rather than for the visual experience. It's one I plan to pick up regardless.

Rome Total War wasn't demoed at the GDC, but I read about it on the plane and it looks awesome. I studied Roman history in college and take accuracy seriously; it looks like the Total War folks have not only made an effort to develop a fun, accurate game, but have produced a real visual stunner as well. Alas, we'll probably not see this one until the end of the year. Maybe we'll get a Caesar 4 in the interim.

The Sims 2 is nearing completion—I met one of the project leads and he used the phrase "90% complete," and it should be a big hit among existing Sims fans. The new graphics are nice, gameplay appears to be more or less the same (but prettier), and though I doubt it will hook any who didn't find the original terribly thrilling, it will be a success all the same. Everything Will Wright touches turns to gold these days—a far cry from the bleak past of SimAnt.

Max Payne project lead Markus Maki gave a really, really interesting lecture on the history and tribulations of that franchise, complete with early tech demos and various motion capture examples. Had I seen it before reviewing Max Payne 2, I may have written the piece somewhat differently, but it was fascinating to see how the title evolved over the years. Though Markus didn't come out and say so, it's reasonable to assume that Max Payne 3 is now or will soon be in development at Remedy. He also acknowledged that Remedy takes the complaints about both games' brevity seriously and insisted that they intend to address that shortcoming with the third installment.

I wanted to see even more stuff (hence the "greed" subtitle above), because some of the new technology that's on the brink of shipping is so exciting that I'm eager to get an early look at it. But as mentioned before, the GDC is about making games rather than showing games—E3 is for showing games—so I and others had to content ourselves with what we did see.

See You Next Year

The overall trend of the GDC was one of increased interest in academia and game scholarship, more focus on behavioral and psychological techniques of engaging the player, and really amazing new technologies. It's really not that common that we see sudden, revolutionary leaps in technological achievement—the last time was when games went to 3D—and pixel shaders, plus the new system busses and 64-bit processors, promise to provide such a leap.

Many developers have for a long time scoffed at the idea of a scholastic foundation of game theorists. Whether they like it or not, though, a game academia is growing and becoming increasingly influential in the development process. Consultancies are appearing, and teams such as the Freeman Group are promising much richer narrative experiences by taking over the writing of games from people who aren't really qualified to do it.

The next step is to engage in more serious game criticism. There's a difference between criticism and review—one is scholarly analysis and the other is a laundry list of features—and until recently reviews have dominated the press landscape. While paper magazines such as PC Gamer will continue reviewing titles and stay out of the scholarly growth (as they should), the subculture of game academics is increasingly finding sites such as this one to reflect on the nature of games and what they mean to those who play them.

Longtime industry scholar and developer Ernest Adams gave one of the last lectures of the Conference, titled "The Philosophical Roots of Game Design." Ernest is a major advocate of gaming academia, and one of the few academics with a long career in game design. People tend to listen to him, even those who dismiss academics as a "those who can't, teach" crowd. His lecture was wonderful, deeply insightful, and carried with it some marvelous revelations regarding the history of game design, its literary foundations, and the juxtaposition of emotion and technology that gaming represents. Anyone who is interested in games as additions to our lives rather than just ten minutes of empty fun here and there would do well to read his excellent lecture, which adeptly sums up the academic theory of gaming and can be found here.

People seem to leave the GDC a bit overwhelmed. Those with long plane rides home are the fortunate ones; they have ample opportunity to review the literally dozens of pages of notes they took over the course of the week and arrange those jottings into coherent thoughts and strategies. Even those who do not have a direct hand in the creation of games will find it easy to fill their week with material of interest.

Ultimately, games, the future of games, is about handmade reality—the sort of custom-machined perception that now exists only in human dreams. Though we are far away indeed from the realization of that goal, the knowledge that the industry is actively moving toward it is a welcome sign of what's to come.