Chris Crawford

Chris Crawford is a man who needs no introduction, but here's one anyway. He created one of the first RTS games back in the early 80s, he created some of the first management games, including a nuclear power plant sim and the legendary Balance of Power. He created the first socially-oriented game with Siboot, which was like the information economy cousin to Dani Bunten's commodity trading M.U.L.E. He worked with Alan Kay at Atari, and wrote the first book on game design. He founded the GDC in his backyard. He pioneered "serious" game design with Balance of Power and then Balance of the Planet.

He then charged out of a GDC lecture hall in 1992 while wielding a sword. In 1997 a 12-year-old read his column in Next Generation magazine and had his views on game design forever altered. In 2004 a stoned, nineteen year-old wanna-be game designer e-mailed him about interactive storytelling, wanting to get involved with Storytron, then known as Erasmatron 4. Crawford replied "I don't understand anything you're saying, but feel free to ask any questions about the book". In 2005 that wanna-be visited the backyard where that first GDC happened, and discussed interactive storytelling with the Crawdaddy and the guys behind Facade, who demoed their soon to be released Beta. In 2006 that designer's first GDC experience involved seeing Crawford emerge from the woods to tell the entire industry to basically fuck off.

That designer is me, by the way, and this man was a tremendous help to my early career and an occasional mentor. But this interview will not be biased by one-sided praise. I witnessed a rough tally of 17 delays in the Storytron schedule, was party to or witnessed an equal amount of design changes, and spent several months doing a contracted R&D prototype while the thing was in brutal Beta. QA x New Engine ^ Heisenbugs = Frustration. And yet, that road is finally winding down to a conclusion.

Storytron

We all know what you're trying to do with Storytron, you want to provide a design/creation platform for people to create toy-language driven, social play ridden, dramatically bidden storyworlds. Can you describe the underlying mechanics, the core algorithm, and what kind of play in particular arises from it?

The Storytron Engine is basically an event-driven system. Every time an Actor does something, that's an Event. Events trigger a variety of consequences. They can change the physical state of the storyworld. They can change the moods or relationships of Actors. Most of all, they can cause other Actors to react with their own plans, which become Events when they are executed. This makes the system very reactive (and very sensitive) to the player's actions.

How much room is there for the Diekto language interface to stretch and accommodate graphical or visual interfaces. I know that's not where you'd go volitionally, but if someone put a gun to your head and said "make this interface appealing to the Flash generation" how much could be done?

We've always had the goal of providing full 3D facial images with good emotional expressions; it's just that, until recently, we've never found a package that met our basic specifications of 1) accessible to independent authors; 2) capable of displaying many different character faces; 3) capable of putting any expression onto any Actor's face; and 4) integrable into our system. We continue to search, and we may have found one just this last week; we're still evaluating it.

I suggest, however, that a more accurate phrasing would refer to "the Flash Gender" rather than "the Flash Generation." It has been repeatedly demonstrated that males are more visually oriented and females are more linguistically oriented. One need only look at erotica to see the difference. Yet most young males refuse to recognize this truth, preferring instead to project their own tastes onto everybody. I have given up trying to convince them of the truth of this matter.

What are the design limitations in using Storytron?

There are two classes of design limitations: intrinsic and those that are specific to Storytron. Intrinsic limitations are those first noticed by people unfamiliar with the medium, who see interactive storytelling as a variation on games (in much the same way that artists a century ago saw cinema as a variation on theater). They rightly observe that interactive storytelling cannot do many of the things that can be done in games (just as artists a century ago rightly observed that cinema could not do many of the things that can be done in theater). For example, you can't have fast action, real-time play, or animation in interactive storytelling. That's intrinsic to the medium, and many people get hung up on those limitations. Such worries are a waste of time. Cinema will never have the intimacy of the theater -- but that doesn't mean that cinema is inferior to theater. It means that cinema and theater are two different forms of expression.

On the other hand, there are many limitations that are specific to the Storytron technology. The most serious, I think, is the steep learning curve for authors. We have been developing this technology for 16 years now, and a great deal of the technology arises from our years of experience grappling with the problems of interactive storytelling. Some of the abstractions we use are quite alien to many people, and they understandably recoil from the technology. We realize that people will be wary of the technology until they see what can be done with it. That's why building the demos is so crucial to our efforts. Unfortunately, a company with zero budget, a bunch of inspired volunteers, and only two full-time workers can't make rapid progress. I myself am deeply involved in everything that happens at Storytron, so I get at best 20 hours a week to work on the first demo, Balance of Power 21st Century. Some weeks I get not time at all to work on it. All very frustrating.

We (myself and the rest of the early group) encountered a fair bit of recombinatory explosions when combining large numbers of roles and variables; what strategies can storybuilders use to impose negative feedback loops and constraint the possibility space?

In general, combinatorial explosions are good! You want the story to be able to branch off in a hundred different directions. Yes, it can be difficult keeping track of all those possibilities. I don't consider that a fundamental obstacle. We are all trained to think of stories instantially rather than algorithmically. That is, we tend to think in terms of one particular story with one particular protagonist facing one particular set of challenges. To create interactive storytelling, we need to make climb up to the next level of abstraction, thinking in terms of cause and effect rather than character and situation. People have no problems thinking about such issues as money, politics, and law in such abstract ways. I believe that there are already people who can master this concept of dramatic abstraction, and as they show everybody else how to do it, more and more people will come to understand and participate. But right now it is pretty esoteric.

How would you describe the underlying dynamic of a Storytron world? Or do you think you've given users enough strong mathematical tools to sculpt dynamics as different as Balance of Power is from Balance of the Planet?

We've certainly given the authors plenty of algorithmic tools for handling all manner of different situations. If we have erred here, it is in being too lavish on the algorithmic side. Some of our algorithmic operators -- 'blend' is an obvious example -- are really pretty unconventional. However, we've been using this system for a long time and we have a good idea of what authors need and what they don't really need. For example, we provide no operators for exponentiation and we're pretty sure that this will not impede anybody.

How many subscribers would you need to see a positive cashflow on your operation?

I can't really discuss our the details of our business plan at this stage.

How many Storybuilders would you need to get enough content to attract that subscriber base? That is, assuming 90% of the people who try don't produce quality content (Sturgeon's law).

I can't really discuss our the details of our business plan at this stage.

What happens if you see a bigger profit margin than EA (2.5% in 2006, I believe, not too tough). What happens if you lose money but at least one strong work is produced?

I can't really discuss our the details of our business plan at this stage.

In 1992 you addressed this hard problem (or perhaps, a wicked problem) as "the dragon". You said: "I'm going to die someday, and before I can do that I've got to face you, eyeball to eyeball. I've got to look you right in the eye and see what's inside." Are you now content to die?

Not yet. Storytron isn't complete yet. I have not yet produced a truly artistic work with it. Balance of Power 21st Century is a warmup project for me. Once that is done, I will return to my old dream of an Arthurian storyworld. I have a liter of water that came from the well at the foot of Glastonbury Tor. I've had it for maybe 20 years now, and I have promised myself that I will crack it open on the day that I complete my Arthurian storyworld. No sooner. The Arthurian project will be about the sadness of leadership, of having a clear vision of a better world if the people you lead could just rise above their petty disputes and work together -- and knowing that they won't.

Balancing Power

Balance of Power is looked on as the first "serious" game. What was your creative process in deciding to make a geopolitics game, and what was your design process in finding the interface and algorithms you used?

Actually, I did a couple of other serious games before Balance of Power. There was Energy Czar for the Atari, published in 1981. It was a simple energy-environment simulator. There was also Scram, a nuclear power plant simulator, also published in 1982. But obviously Balance of Power was the big one. It was pretty simple at first: I wanted to design an "un-war" game: something that emphasized diplomacy over military strategy. I immersed myself in the material for some time, then I set to work. Fortunately, I figured out about then what I now call "Crawford's First Law of Software Design": Always ask, 'what are the verbs? What does the user DO?' So I wrote down a bunch of possible actions that should be available to the President of the United States: intervene in internal conflicts, provide weapons support or financial aid. Those provided the core of the game. Later on I added the ability to offer defense pacts to threatened nations. This got the mechanics of the game going pretty well. There were lots of design issues that came in later, and the game had serious problems right up until the end, but that's the way it is with any game: it's never fun until just before the end. Expecting a game to be fun 80% of the way through is like tasting a cake before it's fully baked. You just have to wait until the end; the thing is never fun until the end.

The algorithms were fairly straightforward; I wrote an entire book explaining them. It was mostly just a matter of coming up with the right buffering factors to keep things from going hyperbolic. The interface was also pretty obvious: a map that you could directly interact with. The map took a lot of manual labor to build -- we didn't have much in the way of software tools in those days, so I had to hand-draw a huge map and digitize it -- literally -- by hand. I remember crouching over this huge hand-drawn map on graph paper on the floor with a tape recorder, calling out "North, 2 west, 3 north, 1 west, north, east, 2 north, 2 east..." I then sat at the keyboard, listening to the tape and keying in text strings: 1N2W3N1W1N1E2N2E...

What's your current appraisal of "serious games".

I don't know what other people are doing in the field. That's fair; none of them seem to know anything about my work 20 years ago.

(Ed: I know several people working in that sector and they´re generally quite fond of Crawford´s work, BoP at least, if not the older stuff.)

What's your appraisal of Paolo Pedercini's work?

Interesting. The architecture is similar to the kind of thing I was doing in the early 80s. Of course, the cosmetics are much better than we could do back then.

What's your appraisal of Jason Rohrer's work?

Rather like somebody trying to write a tragedy with crayons of many colors.

What's your underlying design thesis for the new Balance of Power?

That multilateral diplomacy is usually more effective than military means -- but military means sometimes work.

How does the gameplay of the new BoP compare to the original?

Much more emphasis on the relationships with other countries. The original BoP was definitely a bipolar world: only the USA and the USSR have the ability to affect the world; everybody else is a victim. The new BoP2K is strongly multilateral. The USA has a few more options than other countries, but the USA cannot act unilaterally without serious repercussions.

How likely is it that China will drive a massive devaluation of the dollar in 2009?

Very unlikely. While it would be to their advantage to do so, they simply do not have the means to pull it off. Remember, in absolute terms their economy is still only the fourth largest in the world. Their GDP is still barely one-fourth that of the USA. They simply don't possess the economic heft to drive the dollar in either direction.

Yet.

(Ed: China has recently surpassed Japan as the largest holder of dollar FOREX reserves.)

How much does monetary policy play into your model relative to foreign policy?

Not at all, sadly. The model is already too complex for most people to appreciate. I toyed with several ideas for adding trade considerations to the storyworld. But they got ripped out later.

History

When you were at Atari, what kind of forecast did you project for the game industry and the evolution of the medium?

It's all in The Art of Computer Game Design. I got some of it right, and some wrong. I never saw the Internet coming.

What motivated you to start the GDC?

A desire to build a community of game designers. I knew that there were lots of game designers out there, but game designers had no way of sharing their knowledge. I knew that a professional association of some sort was necessary, but several wise people told me that a conference would accomplish everything a professional association could do, without all the petty politics. I founded the conference and ran it as a nonpolitical entity. I simply refused to allow politics to intrude into the collegiality of the operation. That, ultimately, was a major factor in my undoing.

How instrumental was Ernest Adams in making the GDC what it is today?

The answer to that depends on what you mean by what the GDC is today. Ernest joined GDC well after it had established itself. He brought strong political skills to the board of directors.

Did your designer friends' opinions of you change after the Dragon Speech?

No, I think they saw the Dragon Speech as a more intense expression of beliefs that they had known about for years.

I'm going to paint an alternative history: it's 1996, you decide to barrel forward on Le Morte 'd Artur, using a lot of comparatively ad-hoc algorithms that aren't suitable for generalized interactive storytelling, but work ok for that specific context. You use smoke and mirrors to fill in the gaps and smooth the edges. You release it on PC, it sells 30-50 thousand copies, a publisher approaches you about doing a Playstation version but they want to put in more graphics, you don't bite. It receives mixed reviews, some lauding it as a bold step forward, others claiming it's Final Fantasy with some dialog gimmicks and crappy art assets. The best reviews come in the form of academic papers written in 2000 and forward. If this came to pass, how would your current work be different?

I'd have developed more games along the lines of Le Morte D'Arthur, with stronger storytelling content, steadily improving the internal dynamics. The end result would be a technology that is much more powerful for storytelling than what we have with Storytron, but unusable by anybody but me.

Futures

Which is more likely: inflation or deflation?

Inflation. Deflation requires very special economic conditions. When governments face economic crises, they almost always respond by inflating the currency. Deflation is good for creditors, inflation is good for debtors. The United States is very much a debtor nation now, so inflation is its best short-term response to economic difficulties. Of course, steep inflation will lead to the dollar losing its status as the primary currency for international transactions, which would cost us about 1% GDP growth per year.

Which is more likely: peak oil and environment driven collapse, or globalized Clean Green Fascism(tm)?

Definitely the former, although I don't think we're looking at a collapse.

There are currently a wave of content-creation platforms designed to create different kinds of content, Storytron is one of them, but perhaps the only one oriented to a completely unprecedented form of interaction. There are evidently a lot of unemployed people, and definitely more young people in game design programs than there are game design jobs. Where do you see this going?

My expectation is that some of the eager young creatives who are being shut out by the games industry will jump onto the Storytron platform as one in which their talents can be utilized. Of course, many of these people will be more interested in games than in storytelling, and they won't enjoy working with the Storytron technology. But there will be some serious storytellers who, I believe, will pounce onto this technology and do wonderful things with it.

Will humanity overcome stupidity and find true freedom? Will the human spirit prevail? If so, how instrumental will play be in this perseverance?

No and no. Never forget that we are, at heart, Pleistocene hunter-gatherers who are adaptable enough and intelligent enough that we do a pretty good job of faking it as civilized people. However, all this civilization stuff really is just a show. Deep down in their hearts, people are always ready to kill, torture, or oppress those who are different. There really isn't much difference between a modern political rally and a bunch of hunter-gatherers preparing a war party to raid and kill another tribal camp. A debate on a political blog is pretty much the same as a battle between hunter-gatherer groups, except that it uses words instead of rocks and spears.

Can play help? Only to the extent that we develop the ability to express artistic concepts through the medium of play. At the moment, despite all the chest-thumping to the contrary, games have plenty of artistry but no art. Game designers don't set out to communicate something in their hearts, they set out to put something in their wallets. They are very, very good at accomplishing their goals, and I respect a job well done. But when they start talking out of both sides of their mouths, wrapping themselves in the robes of the artist while they subordinate everything to financial success -- well, they're welcome to lie to themselves, but that's not how I work.

Technology gives us power, but art gives us the wisdom to wield power beneficially. Game designers nowadays are technologists, concentrating on ever more power in their work. But someday, there will be interactive entertainment that makes people look inside their hearts, ponder their relationships with others, and consider their place in the universe. It's coming. It will happen. I may not be the first one to actually pull it off, but I know I'm moving in the right direction.