GDC 2011: Opinion: Understanding Crawford - Why Chris Crawford Is Misunderstood
In 2006 Chris Crawford stood up at the GDC rant-session and candidly informed a room full of people dedicating their professional lives to game creation that they were wasting their time, servicing a dead patient, and by implication a bunch of idiots.
He didn’t say the last part so explicitly, but gauging from the crowd’s reaction, he might as well have. This event made him infamous with a generation of game developers now in their thirties or older, who came of age in the industry when the Playstation series of consoles lead a boom in what is now referred to as traditional AAA game development.
Here’s a guy some of them had heard of with good tidings, a guy who worked at Atari back in the day, wrote the first book on game design ever published, made a number of firsts with his games about King Arthur’s struggles, managing nuclear plants, cold war diplomacy, environmental policy and alien courtship rituals, and he’s suddenly throwing their entire assorted careers in the garbage.
The result was a general bad taste for the man, cementing his image as a crank, a mad scientist bitter from too many years in the woods, outside the funded annuals of the game industry, where marketing and occasionally profits happen.
But it wasn’t always this way, and perhaps Crawford, like the Aurthurian legend he’s so fond of, will be known as the once and future king of game design.
The Way Games Ought To Be
For a younger generation, and the generation of his contemporaries, Chris meant something different.
If it weren’t for Chris Crawford, I would have never though of an imp in Doom 2 as having any similarity to a Nazi. Had I frequented any forums discussing the game, Godwin’s law would have taken over eventually, but this was a special kind of context.
The year was 1997, I was 12 years old, and I was reading Crawford’s column - “The Way Games Ought To Be” in Next Generation Magazine, perhaps the best print publication on the art and science of games ever distributed. He was describing the level “Barrels of Fun” from late in the game, where you spawn amidst a room full of explosive barrels and have to out run the cascading blasts after an Imp sees you and belligerently fires away.
By the literacy of Doom players, as I was at the time, I thought it was a clever level design, but Crawford’s critique made me think differently. He dissected the level as needlessly requiring replay, having no context for the player the first time through, and questioned why games put this onus of painful reincarnation on their players.
I relished this kind of contrary thinking and vacuumed up each column as they came out, it was my favorite part of reading the magazine, and I was disappointed when they changed the columnist - although then managing editor Neil West did a good job as well, just without the same gusto.
The next time I came across Crawford’s work was the summer before college, I read some mention of him on this very site, and proceeded to his personal collection of essays. It was the closest thing to a formal design education that I had encountered at the time.
He had just released a new book, On Game Design and I read the whole thing on 3 hour car-trip to my place of study. The foundation of these ideas, when formal game design training was scarce, lead me to the driving themes of my career: games about people and procedural content. Others had similar experiences; Darius Kazemi, Lead Data Analyst at Blue Fang, comments:
“I have never played one of Chris' games. It's specifically his writing that had a huge impact on me. Chris was the first person I was aware of who was writing about game design in a serious way. Back in 2003 was when I first started learning about game design and development, and there was very little out there in terms of literature.
Aside from Crawford's work, you had Rouse, Salen/Zimmerman, and that's about it. Actually for me, Chris' most influential book is The Art of Interactive Design, which goes beyond video games and is more about general user interface design for interactivity. That book was life-changing for me: it changed me from the kind of person who uses software and devices into the kind of person who critically engages with software and devices.”
Erasmatron and Interactive Storytelling
After reading On Interactive Storytelling the following year, I decided to contact him. Both books’ covers featured a portrait of the man’s face staring eerily at the reader, masking a compendium of opinionated theories written in a Buckminster Fuller-esque inflammatory writing style - if that’s what you were into, it delivered.
Little did I know what I was getting into. I wrote a sort of rambling, pseudo-intellectual email about various concepts discussed in the book, and he politely informed me that he didn’t know what the hell I was talking about, but if I could phrase my questions in an intelligible form he’d be happy to respond. Somehow I didn’t see anything wrong with that.
Eventually I decided the best way to pursue my goals would involve using his then-titled Erasmatron technology, he was pleasantly flabbergasted to hear about it and invited me to join the pioneer group advising him on the development of his great masterwork.
Before going into the history of Erasmatron-cum-Storytron, it’s important to note Chris Crawford’s indelible penchant for obscure intellectual texts from Greek antiquity and the renaissance period. Any self-respecting game designer should fancy him or herself a scholar of all things interesting, whatever that may mean to the individual.
In Crawford’s case, it’s quoting Cicero, studying the cutting-edge translations of the works of Erasmus, and applying this body of literature to names like “Phrontisterion”, the annual conference he held in his backyard during the mid ‘00s. Phrontisterion is Greek for “think-tank” and was a reincarnation in location and format of the very first Computer Game Developers Conference, which would later become the GDC, an event trafficked by tens of thousands.
Instead of asking why he would cloth his brain-children in such arcane terms, realize that this is part of who Chris Crawford is; it doesn’t matter how accessible the idea is, as long as it is true, a platonic form that exists regardless of popular opinion.
I met the man for the first time at Phrontisterion VI, which would prove to be the last such conference. Andrew Stern and Micheal Matias had just released Facade and I had watched Erasmatron’s slow, primordial soup stir and boil for several months. In early July of 2005, we all converged in Chris’ backyard to eat baby carrots and compare notes.
The atmosphere exhibited a low barometric pressure and a high conceptual intensity. We sat in spacious camping chairs in a big circle and went over a series of topics in a manner fitting Socrates and company at his symposium. In youth I had imagined Crawford as a steely engineering type, with a cool voice that commandeered respect, but the man I encountered was closer to Yoda, who might admonish your search for a great game designer by saying “there is no such thing.”
Oddly though, his nasally voice induces a sort of trance-like state, which may be why he’s considered a great public speaker. The use of wild, illustrative gestures and manic pacing also enhances his delivery. It was with that poise that the group tried to dissect the differences and meaningful similarities between Crawford’s and Facade’s distinct approaches to dramatic play.
Storytron As A Training Tool
One of the ideas posted in Chris’ essays was that interactivity should have a certain symmetry between the expectations staged for the player and the things the player could actually do. Too much of the prior and you get Spore, too much of the latter and you’ve get Dwarf Fortress. Facade, which can still be downloaded, suffered from too many expectations relative to a finite pool of story beats the AI characters could execute and narrow interpretations of seemingly vast natural language inputs.
Chris’ engine, which he’d later renamed Storytron in a stroke of marketing sanity, had the opposite problem. Its input structure was identical to the data-structure that generated both the world and the resulting series of events, a logo-graphic “toy” language called Deikto (another obscure name deriving from the history of artificial language).
To borrow another of Crawford’s theories framing interaction as a conversation, Facade would pretend to listen while waiting to speak; Storytron would listen emphatically while speaking mere shadows of the machinations lurking beneath the surface.
Its design was bold, making a leap to a monolithic engine that could allow non-programmers to parametrize the tendencies of any scope of possible storytelling scenarios, storyworlds, that in practice were too interactive. He actually achieved that, but considering the massive scope of the project and the relative lack of resources on hand, I wouldn’t be able to see it for some years yet.
Paul Eres, an independent designer and programmer, notes:
“When I met Costik [Greg Costikyan] in NYC (he invited me and Wynand to dinner because of the Manifesto Games, Immortal Defense thing, which never materialized) he said something like ‘Crawford is great, but I think he's wasting his life.’ Costik saw the Storytron project as like something that'll never be finished.”
It was finished, however, after a few more years of intermittent delays. In 2007 I got involved with a serious game company that was later successfully acquired, and the CTO wanted to explore Storytron’s potential as a training tool. I may have the honor of being the only person paid to work with the tool, and I was one of a handful that had access to the early Alpha.
Beyond the Alpha bugs, the editor experience was less than the artistic dream Crawford had envisioned. But, to be fair, I have weak programming skills, and was able to laboriously tailor the weighted reactions of about 30 verbs.
The client ended up really liking the resulting demo, but there were difficulties in actually deploying the technology in a way that could be utilized. Crawford had the engine architected to run in a proprietary manner, meaning serious game applications, using it to power a Flash game with action-oriented front-end designs, or tying into third party platforms like Facebook were express “no-no’s”, as I somehow imagine Chris saying.
It was like bringing the Holy Grail back from the desert and then telling everyone they could only look at it through a window inside a confessional. The discussions on business model that took place on Storytron forums considered subscriptions and pay-per-play, but in retrospect it seems like the assumption of the engine working only for a specific Java client on their own site limited vast potential for use as middleware.
Eventually Storytron was officially released in late 2008, with a sequel of Chris’ 1985 hit, Balance of Power. In the wake of this, Storytron gained very few players and precisely zero content-creators outside the company proper. Thus, Chris’ 18 years in the wilderness, from his Dragon Speech where he faux-galloped out of the conference room holding a sword, to the initial release of that dragon’s purportedly severed head, seemed to be a failure.
In early 2010, Chris announced that he would be shutting down the main site and trying to find a new strategy for Storytron. Meanwhile, the audience expansion that Chris has been alluding to, one of breadth over depth, happened in a massive way on Facebook, but lo, not in the way that he had imagined.
Rather than try to simply paint a complex but endearing portrait of a man, I want to try and answer a question with this article: Does this man’s contribution to the art and science of game design justify his adversarial attitude towards what many game developers hold sacred?
To judge that, we have to look at two separate things -- the degree that his writings on game design has informed the following generations of designers, and the weight of his playable work.
Paul Eres, who I met years ago via the Storytron forums, adds:
"I think he has an unintentionally negative effect on a lot of game designers (probably because he's the only one with a coherent/systematic theory of game design), mainly that after reading him, they tend to undervalue aesthetics, story, stuff like that, and only care about interaction. But those things are pretty important too. Imagine Super Metroid with just the mechanics, no ambient atmosphere with creepy music, it wouldn't be as fun even though nothing changed but the graphics and music.
He’s probably the smartest game designer who ever lived (perhaps only rivaled by Will Wright). His dedication to Storytron/Erasmatron is pretty great, I like how persistent he is, even though it'll probably go nowhere. I actually like his non-game design writing even better, like his 'how to think' series, or his essay on how he forged a sword, but his game design writings themselves have come to feel too analytical to me upon re-reading, like when he subdivided games into puzzles and games and various other subcategories, it felt a bit taxonomic.
I think his writings are best when they are descriptive, rather than prescriptive; e.g. when he talks about how things are rather than what people should do or should make. People should make what they desire to make.
He gave me the sense that games are important and can change people if used well, which probably influenced my games like Immortal Defense and Saturated Dreamers."
Greg Costikyan, co-founder of Manifesto Games, another quixotic attempt at shattering the status quo of the games industry that didn’t go entirely to plan, has a longer term view on Crawford:
“I first met Chris when I interviewed at Atari in 1983 or thereabouts -- his group, dealing with experimental games, was the one I wanted to work for (though not the one where I was offered a job). After the interview, he gave me a copy of The Art of Computer Game Design, which I read cover-to-cover in my hotel room that very night -- and remember thinking that this was one of the greatest intellectual moments of my life.
A few years later, he was looking for a publisher for Balance of Power and having a hard time finding one; I was then head of the publishing group at a tabletop game company, and even though we published no digital games, I offered to publish it, figuring we could learn the ins and outs of another distribution channel. Sensibly, he found another publisher (Mindscape) that had some idea what it was doing.
We have kept in intermittent touch in subsequent years, often argued vociferously (but with due respect on both sides), and I consider Chris a friend.
I don't think people today realize the impact Balance of Power had; in the last few years, we've gotten happily used to the fact that conventional media pay attention to games, but in the 80s, -none- did. Chris Crawford was the first designer to be profiled by the New York Times Magazine (I believe the second was Will Wright, 20 years later). It was an astonishingly ambitious, and astonishingly successful (by the standards of the then-industry) game. Nothing remotely like it is published today -- alas.
I could go on about his other work, games like Guns & Butter, Patton vs. Rommel, and Trust & Betrayal, but I don't think it's necessary; it was clear, during Crawford's period as a game developer, that he was an -auteur- operating at the top of his form, not always creating -commercially- successful products, but always pressing at the edges of the possible, always advancing the state of the art. He was an inspiration.
Chris is a man who prides himself on not reading fiction, thinking of it as a waste of time; and yet he has devoted the last two decades to attempting to solve the problem of true interactive literature. For my part, I don't think this is a problem susceptible of solution; interaction (inherently divergent) and narrative (inherently linear) are orthogonal --- not necessarily in direct opposition, and products that try to harness the values of both have, in some cases, been very interesting, but I genuinely think the quest is a red herring.
And while, with Crawford, I dream of the days when computer games will be a viable medium of artistic experience -- and, like Crawford, despair of the tastelessness, cowardice, and mendacity of our publishers -- yet I believe that the "viable medium of artistic experience" will ultimately be found in entities we can plausibly call "games" and not in some imaginary realm of "interactive literature." Nor do I think we have totally failed to produce games worthy of being called art; I'm perfectly happy to argue that, say, Civilization is a sublime product of the human soul.
So yes, as Paul Eres suggests I said, I do actually think Chris has been -- maybe not wasting his life, but spending a great deal of lifespan attempting to slay a dragon that probably cannot be slain, and might not be worth the slaying. It is, of course, entirely his right to do so; and if I view his quest as Quixotic, I can do no less than honor it.”
To be fair, Chris plays Civilization sometimes and seems to enjoy it.
Taking another perspective on the hard problem that Crawford has thrown himself at, freelance game designer Chris DeLeon discusses Crawford’s work in response to a question:
"Many of the popular games emerging from companies and individuals right now have complex problems that need to be addressed. How relevant do you believe Crawford’s perspectives are to the next generation of game designers who are trying to create solutions for game design issues?
[...] Getting a computer to process algorithms and crunch math for graphics and physics isn’t a hard problem (that’s what computers do), nor is getting humans to interact with one another a particularly hard problem (that’s what humans do).
To modernize the case further: nor is getting compulsive people addicted to meaningless things a particularly hard problem (that’s, unfortunately, what compulsive people do). Show me something instead that helps compulsive people not develop harmful addiction to shallow, meaningless things, and then I’ll be impressed.
By comparison, understanding how complex meaning is conveyed experientially from computer output to human user is a nut we still haven’t cracked, and it’s a challenge that encompasses the intersection of so many different fields... that most specialists feel that it isn’t their problem to worry about or take on, and that even if it was somehow thrust upon them, they wouldn’t have the right preparation to tackle it.”
DeLeon added via email:
“Crawford's work inspired me to consider the possibility that what's worth doing isn't the same as what's currently fundable, that people lost in today often can't distinguish between forward thinking and wild ideas - that there will always be an audience out there for work with meaning that we cannot allow puerile, short-term interests to drown out.”
Agustin Perez Fernandez, an independent designer/programmer from Argentina whose trance-game, Mantra, will be featured in this year’s Experimental Gameplay Workshop, added:
“For me Crawford is a like a “guerrillero” in the videogame world. But he is a good “guerrillero”, if he had lived in another country or time he probably would be involved with the French revolution or with Che Guevara. He has really strong ideals and maybe they can be wrong, (I personally don’t think so, but many “guerrilleros” are wrong) but he fights for them at any cost. That’s something very brave and respectable. That was how he inspired me; he made me think that in videogames we can have ideals and utopias too and that it is very important to fight for them.”
Gregory Weir, creator of I Fell In Love With The Majesty of Colors, commented:
“I disagree with most of Crawford's theory, but without him I might not be making games at all. Crawford was the first person to look at video games as an art form, and we owe him a tremendous debt for that.”
Finally, Jason Rohrer, who appeared with Crawford in a 2009 episode of the German interview show, Into The Night, has this to say:
“Chris is one of my game design heroes, though I completely missed playing his games when I was growing up. As I came into my own as a designer, I read two of his design books, and after that I went back to play some of his games.
Though his games were over fifteen years old at the time that I played them, I found them to be groundbreaking even by modern standards. For example, in Trust and Betrayal, he was modeling characters on an emotional level. Chris was clearly decades ahead of his time.”
Jason then described Crawford’s influence on his own multiplayer storytelling game, Sleep is Death:
“[I made the game] not necessarily in response to "errors", but instead as an immediate way to achieve some of the desired goals that were taking so long to achieve using a single-player approach.”
The Downfall of Storytron
The problem with Storytron was not that it was ill-conceived, rather that it was too powerful. Making good use of it required not only mastering a fairly cumbersome editor that may, even to an artistic type, make an actual scripting language seem superior. Once you get through the leg work of putting all these scripts in place, you face the mammoth task of getting the storyworld to work as a game.
The term “game” was taboo among the early Storytron participants, so eager were they to break free of everything that came before and start from ground zero, as if the engine were the monolith in 2001: A Space Odyssey -- but I mean “game” in accordance with Crawford’s definition of good interactivity. You’ve got to be able to listen to feedback in order to reasonably formulate and adjust plans to pursue a goal, whether self-made or imposed by design.
In Storytron you could brag to someone Medium-Big and set in motion inscrutable ripples in floating point matrices under the hood of a dozen characters, closing off verb options in later situations and opening trap doors. If only you chose to brag Little-Small instead, then they would have thought you more intelligent without lowering their trust as much, and wouldn’t have chosen to tell other characters you’re a jerk behind the scenes! Except you have no way of knowing this!
So not only does a designer have to tailor scripts for a recombinant explosion of verb intersections, you had to tailor vectors of information -- sometimes props but usually characters bearing complex re-telling verbs with your name on it. Otherwise the experience would be largely unintelligible, more Waiting for Godot than Hamlet. Crawford wanted to create both a new plane of interactivity and the means by which it could be shared with the masses, and in trying to do both at once he sucked himself into a warp (or perhaps Sapir-Warp) best described as a mix of M.C. Escher’s Drawing Hands, and Edvard Munch’s The Scream.
Super-Saiyan Chris Crawford
But he’s not finished. Pulling himself through an ontological warp of his own making, Chris Crawford has emerged as a sort of game design Super-Saiyan, transformed, back from the otherworld, as unlimited as a 19-year-old in a dorm room but with all the right intentions.
So now, he’s going to sit down and make another storyworld, one for the epic saga that he’d been in love with for many years, Le Morte D’Arthur. Researching this article and revisiting his materials, I realize that the biggest benefit I’ve gotten from the man isn’t just ideas or career direction, but an enhanced appreciation of life, because Crawford is driven by a realization of his own mortality and a desperate wish to live a worthwhile life. Why waste your time doing anything less?
Appropriately, Arthur’s story is one that combines all manner of hardship - warfare, adultery, a conception brought through deception, betrayal, patricide - and contrasts it with the ideas of civic duty, of equitable social engineering, of economic productivity, dutiful love and obsession with perfection via the Grail.
The story has many game-like elements weaved in via the roles of the different characters, and though there may be some cognitive dissonance in creating dynamics within the cast of a very old, fixed story, perhaps he can make some mechanics out of the re-tellings and variations that have sprung up over the years. He’s throwing away Storytron and approaching the project from whole cloth, like Arthur tossing Excalibur back in the lake.
As far as myself and many others are concerned, it doesn’t really matter if Chris nails down interactive storytelling and drinks its blood out of the divine cup. His legacy is highly likely to have inspired the further development of several vectors that will transform the art of game design and provide humanity other benefits than mere entertainment: social games that actually generate positive social functions, procedural content engines, serious game applications that educate and proselytize.
But, if Chris can complete Le Morte D’Arthur, taking the modest scope of just doing a revolutionary game - rather than a revolutionary not-quite-game engine, complete with revolutionary not-quite-game-design-tool - he will have made his masterpiece and be worthy of a Lifetime Achievement Award.
I could see him walking up on the stage, fully assured of the meaninglessness of all the pomp, perhaps highly tempted to say “I told you so!” and rebuke the fickle approvals of the crowd. But like Don Quixote returning to La Mancha, to the warm greetings of everyday people slowly appreciating that someone had the courage to take on those damn windmills, he’d be smiling on the inside.