Will Wright’s City in a Box

10 Jun
Will Wright, 1990

Will Wright, 1990

In “The Seventh Sally,” a story by the great Polish science-fiction writer Stanislaw Lem, a god-like “constructor” named Trurl comes upon a former tyrant named Excelsius, now exiled to a lonely asteroid by the peoples of the planets he used to terrorize. Upon learning of Trurl’s powers, Excelsius demands that he restore him to his throne. Trurl, however, is wise enough to consider what suffering Excelsius’s reinstatement would bring to his subjects. So, he instead fashions an intricate simulacrum of a kingdom for Excelsius to rule over.

And all of this, connected, mounted, and ground to precision, fit into a box, and not a very large box, but just the size that could be carried about with ease. This Trurl presented to Excelsius, to rule and have dominion over forever; but first he showed him where the input and output of his brand-new kingdom were, and how to program wars, quell rebellions, exact tribute, collect taxes, and also instructed him in the critical points and transition states of that microminiaturized society — in other words the maxima and minima of palace coups and revolutions — and explained everything so well that the king, an old hand in the running of tyrannies, instantly grasped the directions and, without hesitation, while the constructor watched, issued a few trial proclamations, correctly manipulating the control knobs, which were carved with imperial eagles and regal lions. These proclamations declared a state of emergency, martial law, a curfew, and a special levy. After a year had passed in the kingdom, which amounted to hardly a minute for Trurl and the king, by an act of the greatest magnanimity — that is, by a flick of the finger at the controls — the king abolished one death penalty, lightened the levy, and deigned to annul the state of emergency, whereupon a tumultuous cry of gratitude, like the squeaking of tiny mice lifted by their tails, rose up from the box, and through its curved glass cover one could see, on the dusty highways and along the banks of lazy rivers that reflected the fluffy clouds, the people rejoicing and praising the great and unsurpassed benevolence of their sovereign lord.

And so, though at first he had felt insulted by Trurl’s gift, in that the kingdom was too small and very like a child’s toy, the monarch saw that the thick glass lid made everything inside seem large; perhaps too he duly understood that size was not what mattered here, for government is not measured in meters and kilograms, and emotions are somehow the same, whether experienced by giants or dwarfs — and so he thanked the constructor, if somewhat stiffly. Who knows, he might even have liked to order him thrown in chains and tortured to death, just to be safe — that would have been a sure way of nipping in the bud any gossip about how some common vagabond tinkerer presented a mighty monarch with a kingdom. Excelsius was sensible enough, however, to see that this was out of the question, owing to a very fundamental disproportion, for fleas could sooner take their host into captivity than the king’s army seize Trurl. So with another cold nod, he stuck his orb and scepter under his arm, lifted the box kingdom with a grunt, and took it to his humble hut of exile. And as blazing day alternated with murky night outside, according to the rhythm of the asteroid’s rotation, the king, who was acknowledged by his subjects as the greatest in the world, diligently reigned, bidding this, forbidding that, beheading, rewarding — in all these ways incessantly spurring his little ones on to perfect fealty and worship of the throne.

When first published in 1965, Lem’s tale was the most purely speculative of speculative fictions, set as it was thousands if not millions of years in the future. Yet it would take just another quarter of a century before real-world Excelsiuses got the chance to play with little boxed kingdoms of their own, nurturing their subjects and tormenting them as the mood struck. The new strain of living, dynamic worlds filled with apparently living, dynamic beings was soon given the name of “god game” to distinguish it from the more static games of war and grand strategy that had preceded it.

The first of the great god-game constructors, the one whose name would always be most associated with the genre, was a hyperactive chain-smoking, chain-talking Southerner named Will Wright. This is the story of him and his first living world — or, actually, living city — in a box.


Will Wright has always been a constructor. As a boy in the 1960s and 1970s, he built hundreds of models of ships, cars, and planes. At age 10, he made a replica of the bridge of the Enterprise out of balsa wood and lugged it to a Star Trek convention; it won a prize there, the first of many Wright would get to enjoy during his life. When developments in electronics miniaturization made it possible, he started making his creations move, constructing primitive robots out of Lego bricks, model kits, and the contents of his local Radio Shack’s wall of hobbyist doodads. In 1980, the 20-year-old Wright and his partner Rick Doherty won the U.S. Express, an illegal coast-to-coast automobile race created by the organizer of the earlier Cannonball Run. A fighter jet’s worth of electronics allowed them to drive from New York City to Santa Monica in 33 hours and 39 minutes in a Mazda RX-7, cruising for long stretches of time at 120 miles per hour.

Wright was able to indulge these passions and others thanks to his late father, a materials engineer who invented a lucrative new process for manufacturing plastic packaging before dying of leukemia when his son was just 9 years old. His widow was very patient with her eccentric tinkerer of a son, similar in some ways to his practical-minded father but in others very different. Wright spent five years at various universities in and out of his home state of Louisiana, excelling in the subjects that caught his fancy — like architecture, economics, mechanical engineering, and military history — while ignoring entirely all the others. Through it all, his mother never put any undue pressure on him to settle on something, buckle down, and get an actual degree. When he told her in no uncertain terms that he wouldn’t be taking over the family business his father had left in trust for him, she accepted that as well. Yet even she must have struggled to accept the notion of her 22-year-old son running off to California with Joell Jones, a painter 11 years his senior; the two had bonded when Jones severed a nerve in her wrist and Wright built a gadget out of metal and rubber bands to allow her to continue to paint. The two would marry in 1984.

Given his love for electronic gadgetry, it will likely come as no surprise that Wright was snared quickly by the nascent PC revolution. Already by 1980 he had added an Apple II to his collection of toys, and with it computer programming and computer gaming to his long list of hobbies; his first computerized love was Bruce Artwick’s primitive original Flight Simulator. But it was only after moving to Oakland with Jones that he started thinking seriously about writing a game of his own. This first and arguably last entirely practical, commercial project of his life was apparently prompted by his now living permanently away from home, an adult at last. At some point even a dreamer has to do something with his life, and making computer games seemed as good a choice as any.

His first game was in some ways the antithesis of everything he would do later: a conventional experience in a proven genre, a game designed to suit the existing market rather than a game designed to create its own new market, and the only Will Wright game that can actually be won in the conventional sense. Like many games of its era, its design was inspired by a technical trick. Wright, who had moved on from his Apple II to a Commodore 64 by this time, had figured out a way to scroll smoothly over what appeared to be a single huge background image. “I knew the Apple couldn’t begin to move that much in the way of graphics around the screen that quickly,” he says. “So I designed the game around that feature.”

Raid on Bungeling Bay on the Commodore 64

Raid on Bungeling Bay on the Commodore 64

Raid on Bungeling Bay owed a lot to Choplifter and a little to Beach-Head, sending you off in a futuristic helicopter to strike at the heart of the evil Bungeling Empire, returning when necessary to your home base for repairs and more ammunition. The most impressive aspect of the game, even more so than its graphical tricks, was the sophisticated modeling of the enemy forces. The Bungeling factories would turn out more advanced hardware as time went on, while your ability and need to disrupt supply lines and to monitor and attack the enemy on multiple fronts created a craving for at least a modicum of strategy as well as reflexes.

Wright sold Raid on Bungeling Bay to Brøderbund Software, who published it in 1984, whereupon it sold a reasonable if hardly overwhelming 30,000 copies on the Commodore 64. But, in contrast to so many of its peers, that wasn’t the end of the story. Hudson Soft in Japan took note of the game, paying Brøderbund and Wright for the right to make it into a cartridge for the Nintendo Entertainment System. Wright claims it sold an astonishing 750,000 copies on the NES in Japan and later the United States, giving him a steady income while he played around with the ideas that would become his next project, the one that would really make his name.

As it happened, the first project merged into the second almost seamlessly. Wright had written a tool for his own use in creating the Bungeling Empire’s cities, a little world editor that would let him scroll around a virtual space, laying down tiles to represent land and sea, factories and gun turrets. He realized at some point — perhaps after his game had shipped and yet he was still tinkering with his world inside the editor — that he found this task of creation much more compelling than the act of destruction that was actually playing the game. Might there be others who felt like him? Based on the success of Electronic Arts’s Pinball Construction Set, a program he hugely admired, he thought there just might be.

One fateful day Wright shared his world editor and his still half-baked ideas about what to do with it with his neighbor Bruce Joffe. An established architect and urban planner, Joffe had studied under Jay Wright Forrester at MIT, generally regarded as the founder of the entire field of system dynamics — i.e., using a computer to simulate a complex, dynamic reality. When he saw Wright’s little Bungeling Empire cities, Joffe was immediately reminded of Forrester’s work. He wasted no time in telling his friend that he really needed to check this guy out.

Even though the two have never to my knowledge met, Jay Wright Forrester and Will Wright were a match made in heaven; they shared much beyond the name of “Wright.” Both, to name one example, got their start in the field of simulation with a flight simulator, Jay Wright Forrester trying to build one and Will Wright trying to figure out how Bruce Artwick’s Flight Simulator really worked.

Driven by his desire to make a flight simulator, Forrester had been instrumental in the creation of Whirlwind, the first real computer, in the sense that we understand the term today, to be built in the United States. [1]The more canonical example in American textbooks, the ENIAC, could only be “programmed” by physically rewiring its internals. It’s probably better understood as an elaborate calculating machine than a true computer; its original purpose was to calculate static artillery firing tables. As in so many things, politics plays a role in ENIAC’s anointment. The first computer programmable entirely in software, pre-dating even Whirlwind, was EDSAC-1, built at Cambridge University in Britain. That such a feat was first managed abroad seems to be just a bit more than some Americans in Silicon Valley and elsewhere can bring themselves to accept. The flight simulator never quite came together, but an undaunted Forrester moved on to Project SAGE, an air-defense early-warning system that became easily the most elaborate computing project of the 1950s. From there, he pioneered economic and industrial modeling on computers, and finally, in the late 1960s, arrived at what he called “urban dynamics.” Forrester’s urban modeling created a firestorm of controversy among city planners and social activists; as he put it in his dry way, it “was the first of my modeling work that produced strong, emotional reactions.” He was accused of everything from incompetence to racism when his models insisted that low-cost urban public housing, heretofore widely regarded as a potent tool for fighting poverty, was in reality “a powerful tool for creating poverty, not alleviating it.”

Of more immediate interest to us, however, is the reaction one Will Wright had to Forrester’s work many years after all the controversy had died away. The jacket copy of Forrester’s book Urban Dynamics reads like a synopsis of the simulation Wright was now about to create on a microcomputer: “a computer model describing the major internal forces controlling the balance of population, housing, and industry within an urban area,” which “simulates the life cycle of a city and predicts the impact of proposed remedies on the system.” When Wright’s neighbor Joffe had studied under Forrester in the 1970s, the latter had been constructing physical scale models of his urban subjects, updating them as time went on with the latest data extracted from his computer programs. If he could build a similar program to live behind his graphical Bungeling Empire cities, Wright would have found a much easier way to study the lives of cities. At about the same time that he had that initial conversation with Joffe, Wright happened to read the Stanislaw Lem story that opened this article. If he needed further inspiration to create his own city in a box, he found plenty of it there.

Never one to shy away from difficult or esoteric academic literature, Wright plunged into the arcane theoretical world of system dynamics. He wound up drawing almost as much from John Horton Conway’s 1970 Game of Life, another major landmark in the field, as he did from Forrester. Wright:

System dynamics is a way to look at a system and divide it into, basically, stocks and flows. Stocks are quantities, like population, and flows are rates, like the death rate, the birth rate, immigration. You can model almost anything using those two features. That was how he [Forrester] started system dynamics and that was the approach he took to his modeling. I uncovered his stuff when I started working on SimCity and started teaching myself modeling techniques. I also came across the more recent stuff with cellular automata [i.e., Conway’s Game of Life], and SimCity is really a hybrid of those two approaches. Because his [Forrester’s] approach was not spatial at all, whereas the cellular automata gives you a lot of really interesting spatial tools for propagation, network flow, proximity, and so forth. So the fact that pollution starts here, spreads over here, and slowly gets less and less, and you can actually simulate propagation waves through these spatial structures. So SimCity in some sense is like a big three-dimensional cellular automata, with each layer being some feature of the landscape like crime or pollution or land value. But the layers can interact on the third dimension. So the layers of crime and pollution can impact the land-value layer.

This description subtly reveals something about the eventual SimCity that is too often misunderstood. The model of urban planning that underpins Wright’s simulation is grossly simplified and, often, grossly biased to match its author’s own preexisting political views. SimCity is far more defensible as an abstract exploration of system dynamics than as a concrete contribution to urban planning. All this talk about “stocks” and “flows” illustrates where Wright’s passion truly lay. For him the what that was being simulated was less interesting than the way it was being simulated. Wright:

I think the primary goal of this [SimCity] is to show people how intertwined such things can get. I’m not so concerned with predicting the future accurately as I am with showing which things have influence over which other things, sort of a chaos introduction, where the system is so complex that it can get very hard to predict the future ramifications of a decision or policy.

After working on the idea for about six months, Wright brought a very primitive SimCity to Brøderbund, who were intrigued enough to sign him to a contract. But over the next year or so of work a disturbing trend manifested. Each time Wright would bring the latest version to Brøderbund, they’d nod approvingly as he showed all the latest features, only to ask, gently but persistently, a question Wright learned to loathe: when would he be making an actual game out of the simulation? You know, something with a winning state, perhaps with a computer opponent to play against?

Even as it was, SimCity was hardly without challenge. You had to plan and manage your city reasonably well or it would go bankrupt or drown in a sea of crime or other urban blights and you, the mayor, would get run out of town on a rail. Yet it was also true that there wasn’t a conventional winning screen to go along with all those potential losing ones. Wright tried to explain that the simulation was the game, that the fun would come from trying things out in this huge, wide-open possibility space and seeing what happened. He thought he had ample evidence from his friends that he wasn’t the only one who liked to play this way. They would dutifully build their cities to a point and then, just like Excelsius in the story, would have just as much fun tearing them down, just to see what happened. Indeed, they found the virtual destruction so enjoyable that Wright added disasters to the program — fires, earthquakes, tornadoes, even a rampaging Godzilla monster — that they could unleash at will. As with everything else in SimCity, the motivation for a player consciously choosing to destroy all her labor was just to see what would happen. After all, you could always save the game first. Wright:

When I first started showing the Commodore version, the only thing that was in there was a bulldozer, basically to erase mistakes. So if you accidentally built a road or a building in the wrong place you could erase it with the bulldozer. What I found was that, invariably, in the first five minutes people would discover the bulldozer, and they would blow up a building with it by accident. And then they would laugh. And then they would go and attack the city with the bulldozer. And they’d blow up all the buildings, and they’d be laughing their heads off. And it really intrigued me because it was like someone coming across an ant pile and poking it with a stick to see what happens. And they would get that out of their system in about ten minutes, and then they would realize that the hard part wasn’t destroying, but building it back up. And so people would have a great time destroying the city with a bulldozer, and then they would discover, “Wow, the power’s out. Wow, there’s a fire starting.” And that’s when they would start the rebuilding process, and that’s what would really hook them. Because they would realize that the destruction was so easy in this game, it was the creation that was the hard part. And this is back when all games were about destruction. After seeing that happen with so many people, I finally decided, “Well I might as well let them get it out of their systems. I’ll add disasters to the game.” And that’s what gave me the idea for the disasters menu.

Wright asked Brøderbund to look at his “game” not as a conventional zero-sum ludic experience, but as a doll house or a train set, an open-ended, interactive creative experience — or, to use the term the market would later choose, as a “sandbox” for the player. Wright:

I think it [sandbox gaming] attracts a different kind of player. In fact, some people play it very goal-directed. What it really does is force you to determine the goals. So when you start SimCity, one of the most interesting things that happens is that you have to decide, “What do I want to make? Do I want to make the biggest possible city, or the city with the happiest residents, or the most parks, or the lowest crime?” Every time you have to idealize in your head, “What does the ideal city mean to me?” It requires a bit more motivated player. What that buys you in a sense is more replayability because we aren’t enforcing any strict goal on you. We could have said, “Get your city to 10,000 people in ten years or you lose.” And you would always have to play that way. And there would be strategies to get there, and people would figure out the strategies, and that would be that. By leaving it more open-ended, people can play the game in a lot of different ways. And that’s where it becomes more like a toy.

But Brøderbund just couldn’t seem to understand what he was on about. At last, Wright and his publisher parted ways in a haze of mutual incomprehension. By the time they did so, the Commodore 64 SimCity was essentially complete; it would finally be released virtually unchanged more than two years later.

SimCity on the Commodore 64

SimCity on the Commodore 64

For the moment, though, nobody seemed interested at all. After halfheartedly shopping SimCity around to some other publishers (among them Cinemaware) without a bite, Wright largely gave up on the idea of ever getting it released. But then in early 1987, with SimCity apparently dead in the water, he was invited to a pizza party for game developers hosted by a young businessman named Jeff Braun. Braun, who envisioned himself as the next great software entrepreneur, had an ulterior motive: he was looking for the next great game idea. “Will is a very shy guy, and he was sitting by himself, and I felt sorry for him,” Braun says. In marked contrast to Brøderbund, Braun saw the appeal of SimCity before he ever even saw the program in action, as soon as a very reluctant, thoroughly dispirited Wright started to tell him about it. His interest was piqued despite Wright being far from a compelling pitchman: “Will kept saying that this won’t work, that no one likes it.”

Braun nevertheless suggested that he and Wright found their own little company to port the program from the Commodore 64 to the Apple Macintosh and Commodore Amiga, more expensive machines whose older and presumably more sophisticated buyers might be more receptive to the idea of an urban-planning simulation. Thus was Maxis Software born.

Wright ported the heart of the simulation from Commodore 64 assembler to platform-independent C while a few other programmers Braun had found developed user interfaces and graphics for the Macintosh and Amiga. The simulation grew somewhat more complex on the bigger machines, but not as much as you might think. “It got more elaborate, more layers were added, and there was higher resolution on the map,” says Wright, “but it had the same basic structure for the simulation and the same basic sets of tools.”

SimCity on the Macintosh

SimCity on the Macintosh

While Wright and the other programmers were finishing up the new versions of SimCity, Braun scared up a very surprising partner for their tiny company. He visited Brøderbund again with the latest versions, and found them much more receptive to Wright’s project this time around, a switch that Wright attributes to the generally “more impressive” new versions and the fact that by this point “the market was getting into much more interesting games.” Still somewhat concerned about how gamers would perceive Wright’s non-game, Brøderbund did convince Maxis to add a set of optional “scenarios” to the sandbox simulation, time-limited challenges the player could either meet or fail to meet, thus definitively winning or losing. The eight scenarios, some historical (the San Francisco earthquake of 1906, the fire-bombing of Hamburg in 1944), some hypothetical (a nuclear meltdown in Boston in 2010, the flooding of Rio de Janeiro in 2047 thanks to global warming), and some unabashedly fanciful (a monster attack on Tokyo in 1957), were all ultimately less compelling than they initially sounded, being all too clearly shoehorned into an engine that had never been designed for this mode of play. Still, Brøderbund’s perceived need to be able to honestly call SimCity a game was met, and that was the most important thing. Brøderbund happily agreed to become little Maxis’s distributor, a desperately needed big brother to look after them in a cutthroat industry.


SimCity shipped for the Macintosh in February of 1989, for the Commodore 64 in April, and for the Amiga in May. Some people immediately sat up to take notice of this clearly new thing; sales were, all things considered, quite strong right out of the gate. In an online conference hosted on June 19, 1989, Wright said that they had already sold 11,000 copies of the Macintosh version and 8000 of the Amiga, big numbers in a short span of time for those relatively small American gaming markets. Presaging the real explosion of interest still to come, he noted that Maxis had had “many inquiries from universities and planning departments.” And indeed, already in August of 1989 the first academic paper on SimCity would be presented at an urban-planning conference. Realizing all too well himself how non-rigorous an exercise in urban planning SimCity really was, Wright sounded almost sheepish in contemplating “a more serious version” for the future.

SimCity for MS-DOS

SimCity for MS-DOS

SimCity would begin to sell in really big numbers that September, when the all-important MS-DOS version appeared. Ports to virtually every commercially viable or semi-viable computer in the world appeared over the next couple of years, culminating in a version for the Super Nintendo Entertainment System in August of 1991.

SimCity for Super Nintendo

SimCity for Super Nintendo

It’s at this point that our history of SimCity the private passion project must inevitably become the history of SimCity the public sensation. For, make no mistake, a public sensation SimCity most definitely became. It sold and sold and sold, and then sold some more, for years on end. In 1991, the year it celebrated its second anniversary on the market, it still managed to top the charts as the annum’s best-selling single computer game. Even five years after its release, with Wright’s belated “more serious” — or at least more complicated — version about to ship as SimCity 2000, the original was still selling so well that Maxis decided to rename it SimCity Classic and to continue to offer it alongside its more advanced variant. In that form it continued to sell for yet several more years. Shelf lives like this were all but unheard of in the fickle world of entertainment software.

In all, the original SimCity sold at least 500,000 copies on personal computers, while the Super Nintendo version alone sold another 500,000 to console gamers. Spin-offs, sequels, and derivatives added millions and millions more to those numbers in the years that followed the original’s long heyday; at no point between 1989 and today has there not been at least one SimCity title available for purchase. And, believe me, people have continued to purchase. SimCity 2000 (1994) and SimCity 3000 (1999) both became the best-selling single computer games of their respective release years, while post-millennial iterations have sold in the millions as a matter of routine.

But almost more important than the quantities in which the original SimCity sold and the veritable cottage industry it spawned are the people to whom it was selling. By the time they signed Maxis to a distribution contract, Brøderbund had long since demonstrated their knack for getting past the nerdy hardcore of computer users, for bypassing Dungeons & Dragons and military simulations and all the rest to reach the great unwashed masses of Middle America. Brøderbund’s The Print Shop and their Carmen Sandiego series in particular remain icons of ordinary American life during the 1980s. SimCity must be added to that list for the 1990s. Beginning with a June 15, 1989, piece in no less august a journal than The New York Times, seemingly every newspaper and news magazine in the country wrote about SimCity. For a mainstream media that has never known quite what to make of computer games, this was the rare game that, like Carmen Sandiego, was clearly good for you and your kids.

SimCity even penetrated into the political sphere. With a mayoral election pending in 1990, The Providence Journal set up a contest for the five candidates for the post, letting each have his way with a simulated version of Providence, Rhode Island. The winner of that contest also wound up winning the election. More amusing was the experiment conducted by Detroit News columnist Chuck Moss. He sent Godzilla rampaging through a simulated Detroit, then compared the result with the carnage wrought by Coleman Young during his two-decade real-world reign as mayor. His conclusion? Godzilla had nothing on Mayor Young.

If the interest SimCity prompted in the mainstream media wasn’t unusual enough, academia’s eagerness to jump on the bandwagon in these years long before “game studies” became an accepted area of interest is even more astonishing. Articles and anecdotes about Will Wright’s creation were almost as prevalent in the pages of psychology and urban-planning journals as they were in newspapers. Plenty of the papers in the latter journals, written though they were by professionals in their field who really should have known better, credited Wright’s experiment with an authority out of all proportion to the fairly simplistic reality of the simulation, in spite of candid admissions of its limitations from the people who knew the program best. “I wouldn’t want to predict a real city with it,” Wright said. Bruce Joffe, the urban planner who had set Wright down the road to SimCity, responded with one word when asked if he would use the program to simulate any aspect of a city he was designing in the real world: “No.” And yet SimCity came to offer perhaps the most compelling demonstration of the Eliza Effect since Joseph Weizenbaum’s simple chatbot that had given the phenomenon its name. The world, SimCity proved once again, is full of Fox Mulders. We all want to believe.

In that spirit, SimCity also found a home in a reported 10,000 elementary-, middle-, and high-school classrooms across the country, prompting Maxis to offer a new pedagogical version of the manual, focused on techniques for using the simulation as a teaching tool. And SimCity started showing up on university syllabi as well; the construction of your own simulated city became a requirement in many sociology and economics classes.

Back in May of 1989, Computer Gaming World had concluded their superlative review of SimCity — one of the first to appear anywhere in print — by asking their readers to “buy this game. We want them to make lots of money so they’ll develop SimCounty, SimState, SimNation, SimPlanet, SimUniverse… billions and billions of games!” The hyperbole proved prescient; Maxis spent the 1990s flooding the market with new Sim titles.

SimEarth on MS-DOS

SimEarth on MS-DOS

Jay Wright Forrester’s follow-up to his book Urban Dynamics had been Global Dynamics, an inquiry into the possibility of simulating the entire world as a dynamic system. Wright’s own next game, then, was 1990’s SimEarth, which attempted to do just that, putting you in charge of a planet through 10 billion years of geological and biological evolution. SimEarth became a huge success in its day, one almost comparable to SimCity. The same year-end chart that shows SimCity as the best-selling single title of 1991 has SimEarth at number two — quite a coup for Maxis. Yet, like virtually all of the later Sim efforts, SimEarth is far less fondly remembered today than is its predecessor. The ambitious planet simulator just wasn’t all that much fun to play, as even Wright himself admits today.

But then, one could make the same complaint about many of Maxis’s later efforts, which simulated everything from ant colonies to office towers, healthcare systems (!) to rain forests. New Sim games began to feel not just like failed experiments but downright uninspired, iterating and reiterating endlessly over the same concept of the open-ended “software toy” even as other designers found ways to build SimCity‘s innovations into warmer and more compelling game designs. Relying heavily as always on his readings of the latest scientific literature, Wright could perhaps have stood to put away the academic journals from time to time and crack open a good novel; he struggled to find the human dimension in his simulations. The result was a slow but steady decline in commercial returns as the decade wore on, a trend from which only the evergreen SimCity and its sequels were excepted. Not until 2000 would Maxis finally enjoy a new breakthrough title, one that would dwarf even the success of SimCity… but that is most definitely a story for another time.

Given its storied history and the passion it once inspired in so many players, playing the original SimCity as well for the first time today is all but guaranteed to be a somewhat underwhelming experience. Even allowing for what now feels like a crude, slow user interface and absurdly low-resolution graphics, everything just feels so needlessly obscure, leaving you with the supreme frustration of losing again and again without being able to figure out why you’re losing. Not for nothing was this game among the first to spawn a book-length strategy guide — in fact, two of them. You need inside information just to understand what’s going on much of the time. There are games that are of their time and games that are for all time. In my perhaps controversial opinion, the original SimCity largely falls into the former category.

But, far from negating SimCity‘s claim to our attention, this judgment only means that we, as dutiful students of history, need to try even harder to understand what it was that so many people first saw in what may strike us today as a perversely frustrating simulation. Those who played the original SimCity for the first time, like those who played the original AdventureDefender of the Crown, and a bare handful of other landmark games in the history of the hobby, felt the full shock of a genuinely new experience that was destined to change the very nature of gaming. It’s a shock we can try to appreciate today but can never fully replicate.

You can see traces of SimCity in many if not most of the games we play today, from casual social games to hardcore CRPG and strategy titles. Sid Meier, when asked in 2008 to name the three most important innovations in the history of electronic gaming, listed the invention of the IBM PC, the Nintendo Seal of Quality… and, yes, SimCity. “SimCity was a revelation to most of us game designers,” says Meier. “The idea that players enjoyed a game that was open-ended, non-combative, and emphasized construction over destruction opened up many new avenues and possibilities for game concepts.” Many years before Meier’s statement, Russell Sipe, the respected founder of Computer Gaming World, said simply that “SimCity has changed the face of computer-entertainment software.” He was and is absolutely correct. Its influence really has been that immense.

(Sources: Magazines include Amazing Computing of October 1989; Game Developer from April 2006; MacWorld from April 1990; Computer Gaming World from May 1989; Compute! from January 1992; The New Yorker from November 6 2006. Newspapers include The San Francisco Chronicle from November 3 2003; The New York Times from June 15 1989; The Los Angeles Times from October 2 1992. Books include The Cyberiad by Stanislaw Lem; The SimCity Planning Commission Handbook by Johnny L. Wilson; Game Design Theory and Practice by Richard Rouse III; The City of Tomorrow and Its Planning by Le Corbusier; The Second Self by Sherry Turkle. Current and archived online sources include John Cutter’s blog; Game Research; articles about Will Wright and Sid Meier on Wired; The Next American City; Reform; GameSpot; a 1989 talk given by Jay Wright Forrester, which is hosted at MIT; First Monday; Taylor Francis Online. And finally, there’s the collection of Brøderbund archives I went through during my visit to the Strong Museum of Play.

Beginning with SimCity 2000, the more playable later iterations of the franchise are all available for purchase in various places online. For those of an historical bent who’d like to experience the original, I offer a zip that includes the first three versions — for the Macintosh, Commodore 64, and Amiga.)


1 The more canonical example in American textbooks, the ENIAC, could only be “programmed” by physically rewiring its internals. It’s probably better understood as an elaborate calculating machine than a true computer; its original purpose was to calculate static artillery firing tables. As in so many things, politics plays a role in ENIAC’s anointment. The first computer programmable entirely in software, pre-dating even Whirlwind, was EDSAC-1, built at Cambridge University in Britain. That such a feat was first managed abroad seems to be just a bit more than some Americans in Silicon Valley and elsewhere can bring themselves to accept.

Tags: , , , , , ,

48 Responses to Will Wright’s City in a Box

  1. Ken Rutsky

    June 10, 2016 at 6:10 pm

    Nice, as always.

    My first brush with SimCity was in 1993-4, my senior year of high school. The art department had a Computer Art course, which was really just a fun playground for the art students to mess around in. Our projects encompassed laying out brochures and fliers, using scanners and photo manipulation software, 3-d modeling, MIDI music and the like.

    Mr. Greene, our teacher, also ran a yearlong contest with SimCity 2000. Whoever built the largest and most successful city (to be determined by class votes) would win a little prize. I was obsessed with the game, coming in during lunches and study periods and staying after school in order to come to grips with this fascinating piece of software. A couple of my classmates were similarly addicted, and we shared new tips and discoveries among ourselves (all of our cities ended up having huge hills covered with waterfalls…hydro power was the ticket!).

    My city eventually became on of the largest and most successful, but fell victim (as so many projects in that class) to a computer crash that corrupted my save file, so I was disqualified from the competition. Damn Macs.

    (I did make the best piece of MIDI music, but I was the only one in class who knew both how to play an instrument and use MIDI…)

    • Jimmy Maher

      June 11, 2016 at 7:18 am

      In the original SimCity, it was either coal or nuclear power. A choice of pest or cholera if ever there was one, especially as the nukes were hard-coded to always melt down at some point.

      • Gnoman

        June 12, 2016 at 8:19 pm

        I don’t think they were, at least not in the original. I’ve had several multi-century cities in both the original MS-DOS version and the SNES one and I’ve never seen a meltdown outside of a scenario. As far as I know, nukes just have a very small chance of meltdown. While it could be said that this makes it hardcoded in the sense that “if you leave the simulation running long enough, even the most unlikely event will occur”, I don’t think that’s what you meant.

        In later games, where building life cycles were implemented and periodic replacements were necessary it is quite correct to say that they will melt down – that is the consequence of not refreshing your infrastructure, and most other power plants either had a similar drawback or just polluted heavily.

        • Jimmy Maher

          June 13, 2016 at 9:56 am

          Interesting. Johnny Wilson’s SimCity Planning Commission Handbook, my source for that comment, says that meltdowns depend strongly on the age of the plants. He says they’re 100 percent safe for the first 75 years, but thereafter are guaranteed to melt down at some point between 75 and 150 years of age.

          The source code (for Metropolis) doesn’t bear this out, however. There the chance of meltdown is strictly random, with age of the plant playing no role at all. Every phase, a nuclear plant has a 1 in 30,000 chance of meltdown at the lowest difficulty level, 1 in 20,000 at medium, and 1 in 10,000 at hard. (There are 48 phases in a year, so each one corresponds to just a little over a week.) This does also seem to square more with your anecdotal experience.

          Assuming the Micropolis source is indeed exactly the same as the original when it comes to the mechanics of the simulation model, as its maintainers say it is, this provides an another interesting illustration of the Eliza Effect that was so prevalent with SimCity — i.e., people crediting the simulation with much more sophistication than it really possesses. Thanks!

  2. David Boddie

    June 10, 2016 at 11:37 pm

    When I started reading this article, I wondered if you’d already covered Utopia ( in another strand of gaming history.

    • Jimmy Maher

      June 11, 2016 at 7:24 am

      Interesting. I feel certain I’ve heard of Utopia, but, no, I’ve never written about it and never thought about it at all when writing this piece. It doesn’t appear that Will Wright was aware of it either.

      • AguyinaRPG

        June 12, 2016 at 4:23 am

        Don’t worry, it’s not really relevant in the sense of ludic narrative. It’s merely got the same management theme, as does Fortune Builder on the Colecovision (an outwardly capitalist game!).

  3. Keith Palmer

    June 10, 2016 at 11:58 pm

    Learning (not that long ago) SimCity was released for the Commodore 64 evokes elegiac “sunset’s last glow” impressions for me (as does knowing “Prince of Persia” was released for the Apple II), but we are getting to a point where I was actually playing some of these computer games around when they came out (if not on my family’s home computer; we stuck with the Tandy Color Computer 3 to the end of 1992, which I believe was because of my father’s conviction a new computer shouldn’t cost as much as even generic MS-DOS boxes did…) I don’t know how deep I ever got into SimCity, which was installed on a PC at my high school, but I do remember tidying up with the bulldozer the run-down houses that kept popping up in residential zones. The next instalment, in any case, sounds that much more interesting.

    • Jimmy Maher

      June 11, 2016 at 7:27 am

      I think you were supposed to address the underlying causes of poverty rather than just plow the houses under with a bulldozer. Your solution sounds a little like those of certain politicians. I do hope you also built a ring of police stations around the area in order to be “tough on crime.” :)

  4. Allan

    June 11, 2016 at 4:34 am

    typo: “Carmen Sandigo”

    SimCity is also available as “Micropolis” on some Linux systems—with source code:

    • Jimmy Maher

      June 11, 2016 at 7:29 am


      I do plan to dig into the source code just a little in my next article.

  5. Alex Smith

    June 11, 2016 at 5:36 am

    Small correction: Raid on Bungeling Bay was licensed by Hudson Soft for release on the Famicom, not Nintendo. Hudson also had a major hit in Japan with Broederbund’s Lode Runner. Also, I believe the Famicom version came first. The arcade version appeared on Nintendo’s VS. System hardware, which was essentially a Famicom repackaged in an arcade cabinet. Nintendo released many of its own Famicom games on the system and also licensed hits from its third party console publishers, such as Hudson’s console port of Bungeling Bay.

    • Jimmy Maher

      June 11, 2016 at 7:32 am

      Thanks! Correction made.

  6. Felix

    June 11, 2016 at 1:51 pm

    Oh dear. I must have read The Cyberiad around the same time as I was struggling with SimCity on a Speccy, but I never quite made the connection, I think. And I was SO surprised to learn from a computer magazine years later that Raid on Bungeling Bay, a game I had noticed on the NES, was made by the same Will Wright…

    It was many years until I could “graduate” to SimCity 2000 on a 486, which remains one of my favorite games ever. And even more years before I was able to try out the original again in the form of Micropolis on Linux. And while the latter is so limited in comparison, I still spent quite a few hours playing it again.

    As for the game’s educational value, I think it’s very real, however unrealistic the simulation. Like with Civilization and its theme park view of history, the point is learning to make decisions in the face of complex situations, and having to deal with any consequences, expected or not, that arise. (Another game that’s highly educational in the same way is Settlers, with its simulated economic chain.) And while some of the rules do need to be explained, others really want to be learned by trial and error. A much better way to train future decision makers than making them learn dry theory by heart, in my opinion.

  7. Anonymous

    June 11, 2016 at 2:59 pm

    For anyone interested in playing these games, it just so happens that SimCity 2000 and SimCity 4 are on sale this week at, at $1.50 and $4.00, respectively.

  8. Gerry Quinn

    June 12, 2016 at 5:39 pm

    I used to love Sim Life on the Amiga, though I don;t think it was a huge hit in general. It was very hard to get a stable ecosystem going – in a relatively small space everything tended to go chaotic. Realistic, in retrospect. But I never tired of trying.

  9. Abraham Limpo Martinez

    June 12, 2016 at 11:07 pm

    I would like to point that although SimTower was distributed by Maxis, it was the fruit of a small team of japanese developers and was only distributed by Maxis after it was a success in Japan. A bit like what happened with Raid on Bungeling Bay but in reverse.

  10. CdrJameson

    June 13, 2016 at 7:15 am

    Shouldn’t it be just:

    “constructing primitive robots out of Lego”

    Without having to use that odd ‘s’? cf. wood, brick, stone, plastic, tarmac etc. etc.

    • Jimmy Maher

      June 13, 2016 at 10:03 am

      Sounds odd to me to put it that way, as “Lego” isn’t a material in the same sense of those others. Maybe I should change it to “Lego blocks?” Yeah, that’s the ticket!

      • Lisa H.

        June 13, 2016 at 7:19 pm

        I think the company themselves likes to have the things referred to as “LEGO bricks” (note capitalization), but colloquial “Legos” is pretty common.

        • Jimmy Maher

          June 14, 2016 at 9:16 am

          Bricks it is, then. Much as I want to support one of the biggest national success stories here in Denmark, however, all caps I just can’t bring myself to do. ;)

  11. Wade

    June 13, 2016 at 8:24 am

    The game completed for the Apple IIGS by Burger Bill, too, but never released. A video of that version plus a bit more info:

  12. Alexander Freeman

    June 15, 2016 at 3:57 am

    Oh, man, I remember getting Sim City Classic for Windows 3.1 when I was 12! I once actually got a city up running smoothly, but some random earthquake wrecked it, and I somehow accidentally saved my game AFTER that. I was never able to replicate my success although I suspect I could it again now.

    Anyhow, this may be slightly off topic, but what are your thoughts on simulation/emergent systems for adventure games? I know Adventure was going for something like that, but I’m not so sure it succeeded since I don’t see how you’re supposed to plan out a cave expedition if you don’t already know what’s inside the cave, and I don’t see how you can know that in Adventure except through trial and error (assuming you don’t consult something like a walkthrough of course).

    • Jimmy Maher

      June 15, 2016 at 8:19 am

      It’s a big topic, but:

      I’ve never seen an attempt at entirely procedurally generated storytelling that felt all that compelling to me. I have, however, seen some blended approaches that were quite successful.

      From a modern perspective at least, I’m not sure I’d put early games like Adventure and Zork in the category of complete successes, since, as you alluded, most of their emergent/logistical aspects — limited light sources, the thief and dwarfs, etc. — just feel like annoyances rather than real challenges. Beyond Zork comes closer, but in my view falls down a bit due to some philosophical clashes between its adventure and CRPG sides that it doesn’t quite know how to resolve. As the comments to my article on Beyond Zork will attest, however, there are quite a few who hold a more favorable view of that game than I do.

      I’ll be writing about the Quest For Glory series before too long. I judge that to be perhaps the earliest completely successful blending of an emergent CRPG with set-piece adventure-game puzzles and plot that I’ve seen. We can maybe talk about this more when I get to that topic. In the meantime, if you haven’t played the Quest For Glory games, they do come highly recommended, both to appreciate their gameplay innovations and just because they’re a hell of a lot of fun.

  13. Alexander Freeman

    June 15, 2016 at 5:43 pm

    Well, it’s interesting you should mention the dwarfs. Jason Dyer thinks they and the pirate make a compelling emergent story of cat and mouse:

    I have indeed played QFG. I really like the original game (not so much the remake, though). I also like the second although a lot was underimplemented, and it doesn’t have the “you’re on your own” feeling of the first. The less said about the third, the better. I like just about everything about the fourth, including the “you’re on your own” feeling that it brings back, although I didn’t like the point-and-click interface, which had the effect of making many puzzles not much of puzzles at all, and the bugginess of course. I’ll have to finish the fifth one some day.

    • Jimmy Maher

      June 16, 2016 at 8:29 am

      I just found them annoying. ;)

      I’ve actually been playing the second Quest for Glory on and off recently, and, yeah, it’s definitely a little more rough around the edges than the first. The linear plot means you spend lots of time just waiting around for things to happen. Still heaps of charm, though, which makes it hard for me to judge the game too harshly. It’s been so long that I can’t really speak to the later Quest for Glory games until I give them another play.

      • Peter Smith

        June 17, 2016 at 2:38 pm

        Trial by Fire is probably my favorite. The setting is just more fresh than the kind of Germanic standard of the first one. Which I also love, and which also has tons of wonderful details–like the frost giant “from the northlands” who speaks in the alliterative verse of Anglo-Saxon poetry… hmm, they’re all pretty great actually. I also love Wages of War (again, a setting totally unusual to games, plus some nice open-ended themes) and I’m not sure how it got its bad reputation.

        • Alexander Freeman

          June 21, 2016 at 3:46 am

          “I also love Wages of War (again, a setting totally unusual to games, plus some nice open-ended themes) and I’m not sure how it got its bad reputation.”

          Seriously? Well, OK, this is why it has such a bad reputation:

          First of all, it has a threadbare story. The main plot plays out in a far too straightforward a manner. You are told at the very beginning that demons are probably to blame the tensions breaking out in Tarna, and that sure enough turns out to be true. Discovering the demons’ secret plot might have been a surprising plot twist if it had only been subtly hinted at. As it is, though, the “story” consists of just going out there to confirm what you’ve been told, so there’s no sense of suspense or build up at all.

          There’s also hardly any character development. As a result, all the QFG3-only characters are completely forgettable.

          The game’s challenge is shallow, even for a point-and-click game. Combat is pathetically easy. Once you hit the enemy, the recoil lasts so long you can keep whaling on him ’til he drops dead. The puzzles are just a little more difficult. In some cases, the player is even TOLD how to solve these puzzles by some NPC.

          Many skills and spells are of little use and, as such, are a waste of skill points, only something that becomes evident far after you’ve spent them in the game. The thief in particular has little opportunity to use his skills. The wizard’s side quest involves making a magic staff, which anti-climactically has little utility.

          Although the setting is indeed unique for an RPG, that can’t make up for the fact it’s almost as threadbare as the story. The land of Tarna, at first glance, looks massive. Undulating savannahs, spacious grasslands, forbidding jungles, and various terrain features such as waterfalls, rock formations, and waterholes, combined with the city Tarna and the Simbani village, make for what seems to be a very large and interesting game world. Unfortunately, this illusion lasts only minutes once the player realizes the supposedly large city of Tarna consists solely of an inn, an apothecary, a royal hall, one house, a temple, and a bazaar. The royal hall and the inn are even only visited during cut scenes.

          Even the jungle is really small. Gone is the screen-by-screen navigation used in the first previous games of the series. Instead, the wilderness is a four-screen map decorated with the occasional landmark that’s worthy only of a single visit. As a result, the jungle is actually a vast wasteland.

          The Simbani village has only 6 places: 3 huts, 2 practice areas, and a place where you can play a board game.

          Finally, the ending is almost non-existent and serves only to advertise the next game. This, perhaps more than anything, underscores the fact that the third game is just some filler game intended as padding between the two “real” games.

      • Corey Cole

        July 7, 2016 at 2:00 am

        It’s interesting that I’ve seen the comment about “having to wait” in Quest for Glory II a few times in the last few years, but never when the game was in initial release. It may be that – even with DOSBox – that newer computers speed up game play enough that player actions don’t take up enough time.

        Remember that Sierra games topped out at (and, in the better-programmed ones, were throttled to) ten FPS. Quest for Glory was real-time, so you can probably do far more actions in a game day now that you could originally.

        We also assumed that players would talk to all of the NPC’s multiple times – their dialogue changes as the game progresses – and would spend a lot of time exploring the city and the desert.

        So while it’s possible that the pacing has always been too slow, that might also be an artifact of playing on much faster computers, and being more sophisticated players.

        The time scale in Quest for Glory II was 150 real seconds to one game hour, or one real hour to a game day. Since the Hero will typically sleep 8-10 hours, you have 40 minutes or less to complete a game day. With a typing interface, places to go, and people to talk to, we thought that was pretty tight timing. :-) By today’s standards, it might have been too generous.

        • Jimmy Maher

          July 7, 2016 at 8:31 am

          Partly it may be an artifact of my being too seasoned a an adventure-game player. When I see a time limit, my instinct is to heavily optimize my play. This goes doubly when it’s a Sierra game, since they weren’t particularly known for their fairness, mercy, or player-focused design sensibility. The Quest For Glory games are very much an exception to that rule, but still… that name on the box can scare a guy.

          Also, I imported my character from the first game, which I think probably makes a big difference in the experience of the game. He was strong enough that he could win most of the fights from the outset, meaning there wasn’t much point to grinding. I didn’t particularly mind this, as I don’t enjoy grinding, but it probably affected my experience somewhat. No idea, of course, how many players back in the day imported versus started from scratch…

  14. ZUrlocker

    June 18, 2016 at 10:32 pm

    Great post! I loved SimCity and played the original PC version and SimCity 2K for many months in the early 90s. Hard to explain the appeal but it was definitely innovative and addictive.

    Is there an easy-to-install Mac OS/X version of the open source Micropolis port?

  15. Pedro Timóteo

    June 20, 2016 at 10:59 am

    Found a typo while re-reading: “Rio de Janero” (should be “Janeiro”).

    • Jimmy Maher

      June 20, 2016 at 11:58 am


  16. Soh Kam Yung

    July 1, 2016 at 7:42 am

    Great article. Thanks for putting it up.

    BTW, there a “the the” typo: “Wright ported the the heart of the simulation from Commodore 64 assembler to platform-independent C …”

    • Jimmy Maher

      July 1, 2016 at 8:10 am


  17. Corey Cole

    July 7, 2016 at 1:26 am

    Interesting article. Something I have always wondered about SimCity, but didn’t come up – in the mid-1970’s, I took several “political science simulations” classes at UC Santa Barbara. Teams of players made decisions each turn, which were entered into the simulation. The TA’s posted printouts showing the results prior to the next class session.

    One of these simulations was a city simulation in which the players were city planners. Their decisions affected the city’s growth, pollution, and so on. When I first played SimCity, I assumed it was a PC port of the game I played in College. I wonder whether Will Wright ever “played” the mainframe simulation and was influenced by it.

    In another class, we played Guetzkow’s Inter-Nation Simulation. One of my first games was a simplified version of INS on PLATO. Each team played leaders of an imaginary country and tried to balance productivity with military strength and diplomacy.

    Incidentally, if you have questions about Hero’s Quest / Quest for Glory or Lori’s and my backgrounds, we’ll be happy to answer them by email.

    • Jimmy Maher

      July 7, 2016 at 8:24 am

      Sounds like it might have been something either written by Jay Wright Forrester or derivative of his work. Urban Dynamics was published in 1969, I believe, and was quite influential in academia (in addition, of course, to later influencing Will Wright so heavily).

      I’d been planning to reach out to you anyway, so will definitely take you up on your offer. Will be in touch soon. Thanks!

  18. Peter Piers

    July 27, 2016 at 6:21 pm

    “In that spirit”

    Stray italics there.

    • Jimmy Maher

      July 28, 2016 at 7:08 am


  19. DZ-Jay

    April 22, 2017 at 4:35 pm

    Just a pet peeve of mine:
    >> “… from Commodore 64 assembler…”

    You program in “Assembly Language,” or just “Assembly”; you use an “Assembler” to assemble the source into object code.

    Afterall, you don’t say “he wrote in ‘compiler’…” or “he ported his code from ‘gcc”…” when the source is written in ‘C’.


    • Steve Taylor

      January 25, 2022 at 11:56 pm

      Hey DZ-Jay – My first job was writing 8 bit games in Z-80 and 6502, from around ‘86 to ‘90. Neither saying “I wrote assembly” or “I wrote assembler” presses my pet peeve button – and it’s a button that’s usually pretty easy to press.

      The big question is “What am I doing replying to a comment that’s five years old?”

      • Steve Pitts

        November 22, 2022 at 12:36 pm

        Belated agreement with my namesake here. I wrote my first lines of assembler code (on an IBM mainframe) back in 1981 and have referred to it as assembler ever since, as does much of the industry. It may be a pet peeve for some but like so much inexactitude in English it persists because it has become common parlance.

  20. JJ Tynes

    September 16, 2017 at 9:17 am

    I look at what Will Wright has done in recent years and everything seems a dead end. Spore, Syntertainment, Thred. None of them show any real innovation. And when you hear Will talk, he seems consumed with himself and his sense of self-importance. Hey, and what’s up with the swapping out of the first wife, then the second. Three times is not the charm. Three times is the out. Seems he has lost his way on character issues and it has affected his work.

  21. Wolfeye M.

    September 26, 2019 at 7:00 am

    I remember playing SimCity 2000 on a school computer. Can’t quite remember if it was in elementary or middle school, but it was in the 90s. I was pleasantly surprised a game that the school would allow was actually fun. Fun, but also frustrating, because I never quite figured out how to get a city to last very long. Most of them ended up smashed by monsters and UFOs.

  22. Ben

    October 13, 2023 at 10:37 pm

    Raid on Bungeling Bay (in caption) -> (Bay in italics)

    where it’s becomes -> where it’s become

    • Jimmy Maher

      October 20, 2023 at 12:17 pm



Leave a Reply

Your email address will not be published.

This site uses Akismet to reduce spam. Learn how your comment data is processed.