[I originally wrote this for an internal presentation at Zwift. After reading Daniel Noah Halpern’s book about Tom Caldwell, the head of design at Riot Games, and about game design more generally, I decided to update it and post it publicly.]
Chris Crawford originally published The Art of Computer Game Design in 1984. In 2011, he did a refresh of the book and added some notes about predictions he made in the original edition and how they’d fared over time, as well as examining the overall trend in the computer games industry over the past 30 years. Crawford is a legendary designer, having founded the Games Research Group at Atari in 1982 under Alan Kay, the Chief Scientist there from 1981 – 1984. Crawford’s specialty was tactical or strategy games – his game Eastern Front (1941) was the first microcomputer game to compete with pen-and-paper strategy games in terms of depth and complexity. But he is most well known for his belief that computer games represent a still underdeveloped and underappreciated art form.
Crawford defines art as, “something designed to evoke emotion through fantasy. The artist presents his audience with a set of sensory experiences that stimulates commonly shared fantasies, and so generates emotion.” Given this definition of art, it’s easy to see how computer games can be classified as art. Game are not real, but great games are realistic, and that realism is what makes them so powerful. As a game company, maintaining that thread of realism and plausibility while also not being wholly bound by the constraints of the “real” word is an ongoing challenge. The confines of your specific world must be – above all else – consistent. It is inconsistency that is the death knell of a game, because you cannot master an inconsistent system.
Zwift’s Concept Z1 (aka Tron bike) has glowing wheels, but it’s still obviously a bike; New York’s glass sky roads defy imagination, but not the laws of physics. Every Apex Legend is more than human (or, in some cases, not human at all); can heal repeatedly and infinitely; has unique and remarkable abilities. But they are not omnipotent. They are not gods. They are bound by rules and by the reality of the Outlands. They can die (at least temporarily…). I’m still learning a great deal about the rules of Apex, and that’s why I play it so much. You need to understand your game as a player does – to have empathy – in order to preserve the integrity of your world.
Neither Zwift nor Apex is a solitary game. And because they are multiplayer games, they are defined by their community. And in doing so, both Zwift and Apex further embrace Crawford’s idea of art as games that, “create not the experience itself but the conditions and rules under which the audience will create its own individualized experience.” And it is that topic in particular – the creation of “rules and conditions” – that The Art of Computer Game Design focuses on.
At times, Crawford’s book is extremely technical. There’s a lot of focus on quantifying seemingly qualitative concepts such as “conflict” and “safety.” But Crawford is endeavoring to drive home that there is a process to all of this. Art is a creative endeavor, but that creativity does not exist without structure. In discussing interactivity – a core principle of games is that they must be “interactive” and not merely “participatory,” he states, “What is important about the modes of interaction is not their mechanical quality but their emotional significance.” Zwift is a single-input game – the only input (outside of the MTB trail; steering has come to the rest of Zwift, but it’s still a work in progress, as is the hardware…) is your power. But human power carries enormous emotional significance. There is literal blood, sweat, and tears behind it. And that’s why, in spite of being so simple, Zwift is so richly interactive. Climbing the Alpe is a profound expression of self.
Apex is less physically taxing, but its richness comes from competition and cooperation. Apex is a team-based game (and, truly, it is a game of trios); it is virtually impossible to win without your team. And a good Apex team is absolutely more than the sum of its parts. Whether playing with friends or with total strangers, you must work together to conquer 19 other teams. You must earn a victory in Apex. This emotional significance does not come cheaply. As customer support teams in the video game industry are well aware (see Cyberpunk 2077 for the latest, most poignant example…), the pain customers feel if the game mismanages their investment is massive.
Crawford also goes into depth on the difference between a simulation and a game – “Accuracy is the sine qua non of simulations; clarity the sine qua non of games.” And he gives a valuable lesson in his discussion of “transplanted” games, which he describes as, “one of the most disgusting denizens of computer gamedom.” Transplanted games are games which exist as fully formed ideas in an existing realm and which are transplanted into digital form. He cites virtual craps as an example of such a poorly realized replica. Part of the fun of playing craps is the throwing of the actual dice. The virtual roll of the dice is a poor substitute. As he says, “In one way or another, every transplanted game loses something in the translation.” He does qualify, “It may also gain something, but it always loses something.” And he argues that this is a universal truth. The book is always better than the movie based on it. Shakespeare inevitably loses some of its magic in anything other than the original iambic pentameter of Old English.
As much as the very technical sections – such as the one dedicated to understanding the, “mathematics of overdamped, underdamped, and critically damped systems” – are interesting, they’re probably not of much value to the person not directly involved in technical design. We most often experience damping as part of our car’s suspension; it’s what keeps your car from bouncing indefinitely once you hit a pothole. An example of an “overdamped” system is the Blackhole ultimate of Apex Legends’ Horizon, which pulls a static player inexorably towards the center; a critically damped system is the “draft lock” system where Zwift does work in the background to make it easier to draft and not slingshot past the rider in front of you when she speeds up or slows down.
When I first wrote this, it was with a focus on how Zwift – which in its current form largely is a “transplanted” game – might evolve beyond simply being an adequate representation and recreation of the outdoors, and how it can grow into something with merit for its own sake. Zwift can – and, I believe, should – be a unique and fully realized discipline of cycling. MTBing is not just what you do when the pavement stops; cyclocross is not just a way to ride a road bike in the mud; and BMX is not simply something for kids. They are all distinct and beautiful manifestations of what it means to ride a bike. And Zwift can be yet another such realization, even if the two wheels only ever exist in a virtual space. I think the first real steps in this direction were taken in 2020, with the first ever Virtual Tour de France and UCI Esports World Championships.
As I start work on a new game, I’ve found myself thinking about Crawford’s ideas more broadly. I believe that Apex Legends is most definitely art. Apex is expanding both as a story and as a competitive game. It’s growing both as an esport and as a fictional world full of characters with histories and backstories and feelings and relationships. And I believe these two paths are intrinsically linked. As I wrote in my post – Great Characters Matter – about why I wanted to join Respawn (and the Apex team more specifically), characters can be a critical part of creating immersion and the emotional significance Crawford speaks of. While that’s not the part of Apex that I will work on, understanding the importance of what makes a great game – and what makes a game art – must inform all aspects of design. The frame is an important part of how we view a painting. The location is an important part of how we appreciate a sculpture. And the art of a game must extend beyond the core game loop.
[Postscript: If you’re intrigued by the idea of game design, but not quite ready to jump in the deep end, the best book for people interested in a nice overview of Game Design is probably not this book, but Raph Koster’s Theory of Fun for Game Design, which does a great job of examining what actually makes something “fun.”]