RSS

Tag Archives: electronic arts

The Life and Times of Maxis, Part 3: The Sims


This article tells part of the story of Maxis Software.

I don’t think people realize how much tactical and strategic forethought goes into their daily lives. There’s a whole subconscious time-efficiency layer in our lives. In essence, real-time strategy is our lives.

— Will Wright

On the morning of October 19, 1991, two weeks after writing his last bit of code for SimAnt, Will Wright climbed out of bed and smelled smoke. Looking out the window of his home in the Oakland Hills of California he saw a dense black cloud blowing in from the east. He called emergency services, who informed him that the culprit was a grass fire that was under control.

But after his morning shower, the smoke was worse than ever; great acrid plumes of the stuff were engulfing the house and seeping past the window frames, making it hard to breathe. The emergency-services number now yielded only a busy signal.

Wright gathered up his wife Joell and their five-year-old daughter Cassidy and bundled them into the family car along with a neighbor couple. They shot down the road under a rain of ashes and cinder, choking on smoke so thick that it was all Wright could do to keep the car’s tires on the pavement. Finally, they burst out of the maelstrom into blessedly clear air. Wright continued holding the accelerator down until they were well away. He guesses that, had they been five minutes slower getting away, their names would all have been added to the list of 25 victims of the Oakland Firestorm of 1991.

A week passed before the family could return to the charred remnants of their house, still smouldering there under a glowering autumn sky. There was virtually nothing left — except for the ants, that is. (Wright was rather primed to notice them…) “The interesting part was to find that I wasn’t really attached to much,” Wright says. “I started assessing my material needs: a toothbrush, underwear, a car, a house… I was surprised how I didn’t miss stuff. The fact [that] we got out and none of our family was hurt seemed much more important.”

Wright found himself asking,

What is life made of? Rarely do you do that in your real life. When something like this happens you get a big picture. Where do I want to live? What sort of things do I need to buy? You see your life almost as a project in process. When you’re embedded in your day-to-day life, you don’t get that perspective. I started to wonder about all the things we have and how we purchased them for a reason. Why do we need X or Y or Z? Why do we think something will make [us] happier? It almost comes down to Maslow’s Pyramid of Needs.

A period of reassessment like this might end for some people in a decision to climb a mountain, walk a camino, quit a job, write a novel, join a charity, or move into a monastery. But Will Wright wasn’t that sort of man; he was someone who thought in systems and algorithms. He wondered if there was a game — no, a simulation, or maybe just a software toy — lurking in the prospect he was now facing in the real world, that of assembling the material components of day-to-day life from scratch. Back at work at Maxis, he sat down in front of his computer to find out. His first three games since co-founding the company — SimCitySimEarth, and SimAnt — had been simulations of complex systems with thousands of incidental living actors. Could you simulate the spaces and clutter of individual lives instead?

It would take him eight and a half years to answer that question affirmatively. The belated result of his close encounter with the Oakland Firestorm would be The Sims, the best-selling line of boxed computer games in history. Some would celebrate the series for offering a whole new vision of what games could do and the subject matter they could tackle, drawing in tens of millions of new players who would never have given SimCity — much less Diablo or Starcraft or any of the other hardcore standard bearers — a second glance. Some would condemn it for being just another empty monument to conspicuous consumption and shallow egotism, of a piece with the “reality” shows that were suddenly inescapable on American television screens. In the final reckoning, of course, it was a little of both.


Stories of how The Sims came to be have a tendency to fall into some well-worn patterns. We are told how Will Wright, the brave and pioneering creator, had to fight for his dream against the unimaginative suits who surrounded him for the better part of a decade, until he finally got to enjoy sweet validation as he watched it turn into the biggest game ever. There is something to this telling of the tale, but it’s missing a lot of nuance. First and foremost, it commits the fallacy of assuming that the version of The Sims which Wright conceived in late 1991 was essentially the same as the game which was released in early 2000. In truth, it went through a protracted process of evolution, from which it emerged with a radically different focus. Along the way, several other people became almost as instrumental as Wright himself in shaping it into a game with such extraordinary mass appeal. This is not, in other words, another case of a single genius auteur imagining and implementing something unprecedented, as Wright did when he made the original SimCity. The creation of The Sims was a more organic, at times almost accidental process. For Wright specialized in systems, not in people. To make an appealing simulation of the latter, he had to draw on folks whose strengths were very different from his own.

But all of that was still in the future when Wright was contemplating the altered nature of his existence in the wake of the Oakland Firestorm. At that time, it was accurate enough to say that no one else at Maxis knew quite what to make of his new hobby horse. That he could so thoroughly flummox even this bunch, the makers and publishers of some of the most esoteric games in the world, really says something. Indeed, it may very well say that he himself wasn’t entirely sure what he was doing. Some game designers are like military strategists, speccing out every element of a campaign beforehand. Some are more like mad scientists, mixing a bit of this with a bit of that to see what turns into a useful synergy and what just blows up in their face. Wright is the poster child for the mad-scientist school.

For a long time, he called his idea Dollhouse. As the name would imply, it was all about architecture and interior design. He had become enamored with a book called A Pattern Language, written in 1977 by a Berkeley architecture professor named Christopher Alexander and several colleagues. It can be read as an attempt to define a Western version of the ancient Chinese concept of feng shui, explaining how our living spaces can be arranged so as to bring a sense of peace and harmony to our lives. It defines everything from the amount of green space a city should have to how many floors a building should contain to how high a windowsill should be, all based on universal and eternal “archetypes” that are “deeply rooted in the nature of things.” “You can use this book to design a house for yourself with your family,” wrote Alexander. “You can use it to work with your neighbors to improve your town and neighborhood; you can use it to design an office, or a workshop, or a public building.”

Wright wanted to use it to design a computer game, one that would bring Alexander’s theories to the masses, in the same way that his earlier games had promoted those of Jay Wright Forrester, James Lovelock, and E.O. Wilson. As in those games, the dolls in this dollhouse would be purely supporting players to the real star of the show, the physical environment itself. Their reaction to the spaces you put them in would provide a way of “scoring” how well your layout conformed to the principles of Christopher Alexander. This was their one important function.

Jeff Braun couldn’t have been thrilled at seeing his partner going down this rabbit hole, an abstruse one even by Maxis’s usual standards; SimCity this was not. But as always, he did his best to support Wright, whom he unabashedly regarded as a “genius.” He arranged for Wright to explain his concept to a focus group, just to gauge their reaction. Setting something called Dollhouse in front of a bunch of self-identified gamers circa 1992 was always destined to be a fraught endeavor, but Wright and Braun were nonetheless taken aback by the vehemence of the rejection. “This was probably the most negative focus-group experience I have ever seen,” says Wright. “It was actually quite remarkable. They universally hated it.”

Soon after, Braun wheedled his partner into agreeing to lead the development of SimCity 2000, pulling him away from his living-spaces simulation for a year. After that, the concept got muddled up with SimWorld, this attempt to make all of Maxis’s games interoperable so that they were all essentially one big game in the end. So, you would be able to pick a building in a SimCity city and zoom in to tinker with the fine details of its exterior and interior using Dollhouse. Part of the reason that Wright was so enamored with Yoot Saito’s SimTower was that it was already blurring some of the lines between SimCity and Dollhouse. (The Sony PlayStation port of SimTower even let you explore your skyscraper from a crudely implemented first-person perspective.)

That said, the people around Wright still didn’t entirely get what he was trying to do with Dollhouse. “It just felt kind of peculiar, you know?” says the Maxis researcher, writer, and producer Chris Trottier, whose first assignment as a new hire was to travel to real tropical islands and take pictures for SimIsle. (Nice work if you can get it…) “Will’s goal for it was that it was about efficiency and architectural layout and that the people were just there as a byproduct. The people were what measured your success in setting things up efficiently.”

But in trying to model the behavior of the people well enough to measure the quality of the player’s architectural and interior-design efforts, Wright would wind up inadvertently discovering a whole different, more immediately appealing game.

It began in 1995, when Jeff Braun agreed to bring on a bright young Stanford graduate named Jamie Doornbos to help program the dolls in Wright’s dollhouse. Doing that well enough to properly grade the player’s layouts proved to be far more difficult, complex, and time-consuming than Wright and Doornbos had thought it would be. Luckily, it was also more fascinating than they had ever imagined. Together they realized that the fun of the game may not be where Wright had thought it was. Wright:

We were trying to figure out how we could simulate an open-ended system where the behaviors [of the people] were expandable and they had the level of intelligence that we would require for the game, so that they could basically live out their whole home life and we could simulate it reasonably. So Jamie and I spent about a year and a half just working on the behavior model, as a little research project. At some point it just started really working out, and really looking pretty good.

It turned out that controlling the little people was actually more interesting. It was still fun designing houses for them, but controlling their lives actually turned out to be far more compelling, so the whole project took a turn toward the people. I kept the architecture tools in there, but then I just really started focusing more on the people and objects and their behaviors and relationships — all that sort of stuff.

Thus the stars of Dollhouse became the dolls, with their houses as supporting players, instead of the other way around.

As they did so, the simulation came more and more to resemble an oddity from the previous decade called Little Computer People. Created by Rich Gold and David Crane, Little Computer People was published by Activision in 1985 in the midst of that company’s artsy-fartsy period, when the warehouse full of cash it had collected on the pre-crash Atari VCS market gave it the luxury of venturing down a number of such avant-garde, high-concept alleyways. The “game,” if that’s what you chose to call it, was a bit like the Maxis-published El-Fish aquarium simulator, but with a human in the fishbowl. A little man lives in a little house inside your computer, and you can watch as he and his dog go about their day on a realistic albeit compressed schedule, with 24 hours inside the computer passing for every six outside of it. Depending on the time of day and his personality and mood, the fellow might prefer to relax with a book in his favorite armchair, make music on his piano, exercise, chat on the phone, wrestle with his dog, watch television, listen to his record collection, or play a game of cards with you — when he isn’t sleeping, brushing his teeth, or showering, that is.  You can try to get him to do your bidding via a parser interface, but if you aren’t polite enough about it or if he’s just feeling cranky the answer is likely to be a big fat “No!”

Little Computer People.

The most unique aspect of Little Computer People was that all of the little people were, well, unique. Each floppy disk that shipped had a different personality magnetically encoded on its sectors. You might wind up with a sluggard or an athlete, a Pigpen or a neat freak, according to the luck of the draw. Rich Gold envisioned Little Computer People becoming merely the beginning of a whole product ecosystem which would let you watch and nudge your little people in all sorts of different environments: at school, at work, during a night on the town, etc. Maybe you could even put your little person together with someone else’s to see how they got on! All of those plans were scrapped when sales were not up to snuff — but they were an uncanny reflection of the direction that Will Wright’s game would later go.

This was probably not completely coincidental. Wright has stated that he himself played Little Computer People in the mid-1980s; thus his early habit of referring to the characters in his own game as “little people.” During an online chat hosted by CNN in 2001, he said that “I’ve since gotten to know several who were involved with that project, and many of them gave valuable feedback, especially Rich Gold.” The most obvious differences between the two projects had to do with resources. Whereas Little Computer People had to run on an 8-bit Commodore 64, Dollhouse was implemented for a 32-bit Mac or Windows machine with a thousand times the storage and processing power. And whereas Little Computer People was made by two men in a handful of months, Wright had the privilege of tinkering for years, and would have a team of several dozen at his disposal before all was said and done.

For right now and quite some time to come, though, it was just Wright and Doornbos. They rechristened Dollhouse “Project X,” reflecting the reality that its goals were becoming more open-ended rather than less so, even more of a “little research project” than a coherent game. Which isn’t to say that the tech wasn’t cool and innovative. At bottom, Wright and Doornbos strove to create a simulation of the psychologist Abraham Maslow’s Hierarchy of Needs. This was a huge leap into the wild blue yonder for a project that had begun as a simulation of architecture and interior design. “I spent a lot of time thinking about how to make the behavior of these people very robust, interesting, and plausible, no matter what kind of environment you put them in,” says Wright. Chaim Gingold described the approach in his (highly recommended) 2024 book Building SimCity: How to Put the World in a Machine.

Wright crafted a simulation of a person’s needs. A simple prototype showed [how] a person’s hunger, bladder, and other needs evolve over time. If the person peed in their pants or fell asleep, a dialog box popped up. Doornbos incorporated this “motive engine,” a concise set of equations and coefficients, into a new Project X prototype focused on imbuing the people and things of this dollhouse world with behaviors.

“SimAntics,” as the language came to be known, went through many iterations, but from the start it had a graphical editor inspired by the diagrams Wright drew to explain how people would do things like go to the bathroom, brush their teeth, and make breakfast. These behavior scripts, which were compiled into a compact bytecode, told people and objects how to do things like select activities, walk around, animate, and, in the case of insensate objects like appliances, break. As in SimAnt, intelligence emerged from the coupling between agent and environment. Objects broadcast advertisements that claimed to service a person’s changing motives. A toilet, for instance, would advertise its ability to relieve the bladder motive. Crucially, the custom language scaffolded the design conversation between Doornbos and Wright, who could discuss its limitations. As a result, SimAntics was highly extensible; new objects could easily plug into Project X’s dollhouse world and describe both why and how a person might use them.

In a sense, then, the system worked by inverting the expected order of things, by making the people dumb and the objects smart. “This all happens behind the scenes,” Will Wright said, “but basically the fridge yells at [a person], ‘I can satisfy Hunger Five!’ and the chair says, ‘But I’m all about Comfort Four!’ and the bath says, ‘Hygiene Seven right here!'” This approach was rather brilliant in that it meant that you could add new objects without touching the core “needs” simulation, and have the inhabitants of the world respond believably. As we will see, this modularity, and the nearly infinite extendibility it brought with it, would be key to what Project X would become; stated in terms of filthy lucre, it was capable of turning one product into a potentially limitless line of same. In the meanwhile, notes Chaim, “Doornbos wasn’t sure if Project X would be fun, but he loved working for Wright.”

For a while, the two men were allowed to tinker in peace; again, Jeff Braun always prided himself on letting his resident genius do his thing without outside interference. Yet Wright and Doornbos would only be able to ignore the clouds that were gathering outside their blue-sky bubble for so long. As Maxis’s financial situation began to look worrisome, the now publicly-traded company’s board could hardly fail to note that its star designer hadn’t finished an original game since SimAnt almost five years ago, hadn’t contributed anything tangible to the bottom line since leading the SimCity 2000 effort. Nevertheless, it was Braun rather than Wright who was dismissed in August of 1996. Wright was then more or less forced by new CEO Sam Poole to take a hiatus from Project X and help out on SimCopter, the long-delayed proof of concept for SimWorld. Pressed to focus on something more realizable and practical than his people simulator even after SimCopter was finished, he spent time drawing up plans for a simulation of, of all things, the infamous German airship Hindenburg. He gave that up, he admits, only because “the Hindenburg had a swastika on the tail. And even if we took the swastika off, a lot of people have this association in their mind of the Hindenburg as a Nazi symbol.”

In June of 1997, the struggling Maxis was purchased by Electronic Arts. One might have guessed at the time that this event would represent the final kiss of death for Project X. Luc Barthelet, the manager whom the new corporate overlords sent in, was not at all a blue-sky type. On the contrary, he sounded almost like a caricature of a soulless gaming executive when he insisted that the time of Maxis as a maker of quirky little simulations of unusual things was well and truly over: from now on, he said, they would make only “top-ten” games.

And yet Barthelet surprised everyone by recognizing that Project X could be just that — recognizing it far better, in fact, than anyone had during Maxis’s years of independence, including Will Wright himself. He saw that all of the computers that were being sold in the wake of Windows 95 and the dot.com craze were being bought by new demographics with wildly divergent interests from conventional gamers. If you could give those people something fun and engaging to do on the computer when they weren’t surfing the young Web over creaky dial-up connections, you could open up and own a whole new market. So, rather than cancel Project X, he put roughly half of Maxis’s resources into it, while the other half worked on the safer and more predictable bet known as SimCity 3000. With the injection of employees and resources came a new focus on turning the research project into a product that could be sold; Will Wright says that it was “probably Luc more than anybody else [who] was responsible” for turning Project X into a proper game. It was rebranded simply The Sims, a name that had been used for the anonymous, interchangeable people in the company’s simulations since the days of the original SimCity. They would be anonymous and interchangeable no more; they would be the whole point of this latest game.

Whatever else they may have been, Will Wright and Jamie Doornbos weren’t touchy-feely guys. They had been and remained all about their algorithms, which had the potential to become the plumbing of a game, but were nothing like the sum total of one. The colleagues who joined them had the assignment of turning a research project into a playable, saleable game.  Claire Curtin and Roxana Wolosenko, two women who joined the project in co-designer roles, strove to humanize the algorithms. Ditto the aforementioned Chris Trottier, another woman, who came aboard as an associate producer. The ranks of the team came tantalizingly close to gender parity, in an industry that was at least 90 percent male outside of Maxis’s offices. This surely had much to do with how appealing The Sims would become to a wide and divergent swath of folks, gamers and non-gamers alike.

As Maxis’s later promotional materials would advertise, the principal facets of the game were building, buying, and living. You made a place for your Sim to dwell, filled it with stuff, then helped her to deal with the fallout of your choices. It quickly became clear to the team that they needed to dial back some of Wright and Doornbos’s behavioral algorithms in order to give the player something to do during the living phases. In fact, in the finished game, Sims don’t do much of anything on their own volition other than eating. You have to tell them when to go to bed and when to get up, when to wash the dishes and water the plants — and yes, even when to go to the bathroom. If you fail to send them to the toilet in a timely manner, a messy and embarrassing accident will follow. Some future players would find this elevation of the mundanities of life to the heart of the gameplay loop weirdly riveting; others would find it mind-numbingly boring. But no one could deny that it was daringly different.

The Sims would be a lonely experience if it was, like Little Computer People, all about one person alone in a house. It ought to be possible to have friends and families living together, ought to be possible to have outsiders come over for a visit, for barbecues and pool parties — both kinds of pool — and even romantic trysts. At bottom, the game would remain a nesting simulator — there was no possibility for leaving the house and its environs — but at least there would be relationships inside the nest. Indeed, these would become the real heart of the game in the eyes of most of its players. This was no small achievement. Games tend to be good with structured fictional genres like mysteries, the kind that have goals which are never in doubt and are achievable by taking logical, concrete steps. Relationship drama is the opposite of that. Maxis pulled it off well enough to mimic a frothy sitcom or soap opera if not deathless literature — which was more than anyone else had managed.

The first and trickiest problem to solve in this context was the question of how the Sims should communicate with one another. Maxis wanted to create the impression of conversations without delivering literal meaning; as Will Wright noted, even if they included five CDs worth of voice-acted stock phrases, the conversations would soon start to sound repetitive and moronic, pulling the player right out of the fantasy. Maxis’s first idea was to use a language that few of their players would be able to identify, much less understand: they toyed with such tongues as Ukrainian, Estonian, and Navajo. But with them too, familiarity bred contempt, when it became clear that the Sims were repeating the same stock phrases over and over, even if the player didn’t know what they meant. Maxis finally settled on a gibberish that has become known as Simlish. Manipulated to suit the mood of an encounter, a fairly small amount of formless mumbling, provided by just two talented improv actors, could be stretched a surprisingly long way.

The speech bubbles convey the broad contours of what is being discussed. Here the topic is swimming.

Maxis created a look and personality for The Sims that were familiar from other forms of media but new to games. In place of the tacky-looking digital Arnold Schwarzeneggers and Pamela Andersons who dominated hardcore gaming, they offered up relatively normal-looking people — idealized to some degree, yes, but possessed of normal proportions and normal sartorial instincts. The people and environments were given an airbrushed sheen that was, if not quite realistic, sanguinely adjacent to real life in the turn-of-the-millennium United States. The aesthetic and the vibe of the game were of a piece with those of a “hangout” television show like Friends or Dawson’s CreekThe Sims too was real life, only better, with no truly intractable tribulations to suffer through, with stakes that were big enough to engage but not big enough to make anyone uncomfortable. We might go so far as to call it the first hangout game.


The music contributed as much as the visuals to making The Sims stand out from the gaming crowd. In place of the bombastic John Williams-inspired orchestral soundtracks and jackhammer metal tunes of other games, it featured perky, pleasant, melodic tracks with a range of inspirations spanning from Antonio Vivaldi to Keith Jarrett. Created by a whole stable of professional composers and musicians, herded by Maxis’s in-house sound man Jerry Martin, at its best the music of The Sims transcends the artifact to which it has been attached to become something pure and lovely in its own right. A surprising number of players would go so far as to buy EA’s official soundtrack CD, so that they could listen to the music without playing the game; many considered it a fine accompaniment to household chores, which seemed thoroughly appropriate. As Alex Robert Ross wrote in an article for Vice, “it wouldn’t be a stretch to say that more people have heard Jerry Martin’s minimal jazz than Thom Yorke’s measured howls on ‘Everything in Its Right Place.’” (This is a reference to Radiohead’s Kid A, arguably the most critically acclaimed album to be released in the same year as The Sims.)


In the spirit of being a welcoming space — real life with the sharp edges sanded off — The Sims broke ground by not just permitting same-sex romantic relationships but thoroughly normalizing them, treating them as just another thread in the fabric of humanity. This makes quite a contrast to Maxis’s swift backpedaling the moment the kissing lads of SimCopter were discovered. Gay relationships wound up in this game more by accident than any premeditated effort to stand up for equal rights. Then, once they were there, no one had the heart to take them out. It strikes me that a lot of positive change in the world happens in exactly this way.

The story in this case involves a programmer named Patrick J. Barrett III, who was himself gay. Unlike his predecessor Jacques Servin, who had deliberately inserted kissing men into SimCopter as an act of provocation and been fired for his trouble, Barrett really did do so accidentally, according to his own telling. He joined Maxis in October of 1998, after the decision had already been quietly made not to allow homosexual relationships; this was still a dismayingly commonsense one in the late-1990s United States. But nobody ever told Barrett about it. So, when he was asked to implement some algorithms found in a planning document for handling romantic relationships, he gave them equal weight for everyone. And once they were in the game, it seemed to all of his colleagues too sad and cowardly to even contemplate going back and taking them out. Nobody made a big deal out of it. Everyone “just pretty much ignored” that the rules had been changed, says Barrett. “The topic didn’t come up again.”

Under the new rules, every Sim started out in the dead center of a spectrum. If you prodded her to direct toward the opposite gender actions that were tagged as “romantic,” she would drift toward the heterosexual end; vice versa if you pointed her toward her own gender. If you played your cards just right, you could even keep her poised right in the middle of the spectrum — effectively, bisexual, or asexual if you preferred.

The low-key way in which the game tackled the subject would wind up being more powerfully validating to some of its players than a box festooned with rainbow flags and Pride regalia would have been. Patrick Barrett:

At the time, it wasn’t considered “normal” to be gay or lesbian. Some even saw it as dangerous. But in The Sims it was normal and safe to be a gay person. It was the first time we could play a game and be free to be ourselves within. It was a magical moment when my first same-sex Sims couple kissed. I still sometimes wonder how in the world I got away with it.

In keeping with the law of the land at the time, gay Sims could do everything except marry one another and have children. The picture of happy domestic cohabitation the game presented was enough to make you wonder why not — and that, of course, was a potent statement in itself.

The Sims had its coming-out party at the E3 trade show in May of 1999. It caused a buzz there that belied its modest booth in a back corner of the show floor. That E3 took place in the heart of Hollywood, and seemed to many attendees to mark the opening of a new chapter in gaming culture’s relationship with the rest of pop culture. Rock star David Bowie was there, exuding relaxed cool as he signed autographs, gave interviews, and even put on a concert, all to mark the inclusion of his songs, voice, and visage in the upcoming Omikron: The Nomad Soul. Despite bearing no surface resemblance whatsoever to Omikron, The Sims seemed somehow to be part and parcel of the same trends that had landed Bowie at this gathering. Indeed, taken purely as games, The Sims actually stood out far more than Omikron. It had no dragons, no spaceships, not even any guns. It struck many as an argument for the vitality and relevance of the medium that was even more convincing than cameos from an aging rock star.

Chris Trottier’s memories of the show are of near-pandemonium: “There were just rows of people, and we were just trying to project our voices. We were completely hoarse. That was the turning point.” Two female Sims kissed there on the screen in front of the assembled crowd on the very first day of the show. This provoked more excitement than controversy: “You might say that they stole the show,” says Patrick Barrett. Given the demographics of the attendees, it was probably for the best that the kissers were two women rather than two men: “I guess straight guys that make sports games loved the idea of controlling two lesbians.”

After that E3 show The Sims, which had heretofore received little outside attention, became a fixture in the gaming press. Some of the hardcore set reflexively kicked against it for watering down their hobby in some amorphous way, but most looked forward to seeing how it would turn out.

Just as had happened the previous year to SimCity 3000The Sims wasn’t quite ready to go in time for that Christmas buying season. For the second year in a row, Luc Barthelet had the unenviable task of convincing his bosses at EA to choose quality over timing. And for the second year in a row, he succeeded.

The Sims shipped in February of 2000, exactly one year after SimCity 3000. It immediately became clear that, while SimCity 3000 had been a massive success by all of the usual metrics, The Sims was going to blow all of those same metrics right out of the water. It was exactly the right game at exactly the right time, feeding a latent demand from all of those new purchasers of personal computers for something absorbing and engaging, but also lighter, more approachable, less zero-sum, and less “gamer dark” than the industry’s standard fare. In many ways, it was the fulfillment of a dream that the games industry had been flirting with since Trip Hawkins founded EA back in 1982, with only localized and intermittent signs of success. This was the moment when digital games went fully and irrefutably mainstream, the moment when mothers and sisters, shopkeepers and hairdressers joined their nerdy brothers, sons, and husbands in the hobby. Most of these folks would never self-identify as gamers, but they played. Boy, did they play.


You start out by creating your Sim, giving her an appearance, a gender, and a personality profile defined along five continuums. In keeping with a game that is proudly goalless, the emphasis is not on making the best character, as in many CRPGs; it’s on making the character that you want to play.

Next you need to build a house for your Sim to live in. You only have $20,000 to start with, so you have to economize. Note the bar graphs of her current needs. These lie at the heart of the gameplay loop. Your goal is to never let them drop any lower than they have to. Yes, this means that going to the toilet is as important as love and friendship. Perhaps this is more true to life than most of us would like to believe…

Now you need a job. You can advance along ten different career tracks in the original game, more if you’ve installed expansions. You don’t actually follow your Sim to work, but her performance on the job is dictated by how well her needs are being met at home. Send her to work tired or hungry, and she’ll under-perform and eventually get fired; send her out in tip-top condition, and a promotion could soon be coming, with the requisite higher salary attached. (This is where criticisms of the game as a consumption simulator can begin to take root…) In one of the game’s more baffling omissions, there are no weekends. Your Sim needs to go to work every single day. How’s that for a capitalist treadmill?

The Sims may have seemed like a kinder and gentler game by the standards of its time, but it’s much more unforgiving than its sequels. Your Sims can die in a housefire or even starve to death. Public assistance doesn’t exist in its dog-eat-dog economy, but one type of social worker does — the type who can carry off your Sims’ crying children if they aren’t being properly cared for. Predictably enough, some players found their fun after they got bored with being nurturing in making life hell for their Sims, from bulldozing their houses to killing their pets.

Naked Sims are discreetly blurred out, and there is no explicit sex.

Sex is certainly implied, however, through home accessories like a “vibromatic heart bed” well-suited to “play.” This is contrasted with the more conventional “mission bed”; the folks at Maxis were winking and nudging so hard they must have developed permanent twitches. Sims can have extramarital affairs, or even engage in openly polygamous relationships. In retrospect, it’s a wonder more people didn’t notice and complain about this stuff.



The commercial trajectory of The Sims was unprecedented in the history of computer games. It shifted a million units within its first few months. It duplicated the achievement of SimCity 3000 by becoming the best-selling computer game of its release year; then it duplicated that feat again in 2001, all in an intensely novelty-driven industry where most titles went in and out of print in well under a year. In 2002, the two-year-old The Sims officially snatched the title of best-selling computer game of all time away from the nine-year-old Myst. It just kept on keeping on from there. Its total sales reached 16 million copies by 2005, numbers comparable to those an upper-echelon pop star could expect to put up with a hit album, a product with a far cheaper price tag and a far lower technological barrier to entry. Only four albums released in 2000 managed to sell as many or more copies than The Sims (and no, the critics’ darling Kid A was not among them.)

But that first boxed game was only the beginning of the marketplace story. Will Wright’s decade-old idea for an interoperable universe of games finally bore proper fruit now, as Maxis turned their hands to creating expansion after expansion, to add more environments, more stuff, more relationship options… just more, more, more. Take your Sims out on dates; take them on vacation; give them pets; even teach them to cast spells. The Superstar pack let you turn them into celebrity-tabloid fodder, which felt like closing a circle; as I noted earlier, The Sims had always had some of the same appeal and even many of the same visual cues as a fishbowl reality show like The Real World. The expansions packs included cameos by hot stars of the day like Avril Lavigne, Christina Aguilera, and Drew Carey, further blurring the lines between gaming culture and generalized pop culture. There were seven of them in all released between 2000 and 2003. Will Wright has confessed to some discomfort with the sheer quantity of them, but Maxis kept the quality just high enough, and they too sold in the tens of millions. Needless to say, lots of people — most prominently among them Will Wright and Luc Barthelet — got very, very wealthy in the process.

One can go so far as to say that Maxis, alongside the makers of early MMORPGs such as EverQuest, invented the modern approach of treating games as services, to be kept alive almost indefinitely by a steady stream of fresh content and a serious focus on community. By no means has this been an unmitigated positive for gaming — DLC and loot packs have deservedly garnered a terrible reputation in some corners, and that’s without even discussing such grift-adjacent boondoggles as Star Citizen and Shroud of the Avatar — but in this early, comparatively innocent age it worked a treat. As a supplement to the expansion packs, Maxis put up a lot of fresh content on The Sims website, which also gave fans the opportunity to share their own mods, objects, screenshots, and stories. Importantly in an era when CD burners had become commonplace, you needed a valid serial number to access all of this and become a full-fledged member of a community that was, estimated Will Wright, half of the reason people played the game. As he said, “If you register the game, you get more than a pirate would get.” This too doubtless drove sales.

The knock-on effects were enormous. The Sims essentially created the casual-game aesthetic that persisted throughout the aughts and beyond on portals like Big Fish Games and in series like Diner Dash: bright, colorful, pleasant environments and people, coupled to fictional premises with personal rather than epic stakes. Like The Sims itself, these games were the polar opposite of the ones preferred by the hardcore set, with their demanding, ultra-violent quests to save one more dark-fantasy world or other. You could dismiss the casual scene as Hallmark Channel pablum if you liked, but the fact remained that many people — especially women, who came to make up well over half of the casual player base — felt they had more than enough stress in their day-to-day lives already, and didn’t feel a need to seek out more of it in their interactive entertainments. Meanwhile, for publishers like EA, whose acquisition of Maxis had paid off beyond its wildest expectations, The Sims and what it spawned became the ultimate validation of the premise with which I opened this trio of articles: that the real money in games does not come from targeting gamers. By many reckonings, casual games would be outselling and out-earning hardcore games by a considerable margin by 2010.

A game this huge was bound to attract journalists, columnists, and academics. And did it ever. The Sims was SimCity times ten in that respect. It was widely lauded for finally cracking the puzzle of how to make a game with genuinely broad appeal, long after other misbegotten attempts like the “Siliwood” movement had been cast into the dustbin of history.

Other aspects of the game were perchance less laudable. It was hard for some to ignore how it restricted the simulated inhabitants of its world to conformist, bourgeois lives, in which happiness was derived from having the right interior accessories and a mate with a suitable income. In a contemporary journal article, Russell W. Belk pointed out how conventional and even retrograde — same-sex relationships excepted — the assumptions and expectations built into the game really were.

Many elements in this game represent a morality that is centered in the suburban America of the 1950s. Characters must get a job, get up early to catch their carpool (no telecommuting allowed), advance through a career path, earn more money, buy more things, clean house, and practice good personal hygiene. The Sims manual recommends having one spouse in a family stay at home and keep house. If possessions are bought on credit and the credit bills go unpaid, they are repossessed. The dating model of Hot Date is also retro in this age of hanging out and hooking up. Among the ways to impress a date (with greater intimacy being the payoff) are to bring flowers, go shopping together, give gifts, and take the date to more expensive bars and restaurants. The one who asks the other out does all the paying, and those with more money have a distinct advantage. Children must get enough sleep and food so that they don’t get picked up by the welfare worker or sent to military school. Maids and gardeners are also throwbacks to an earlier era when such pseudo-slavery was more democratized.

“A look at the homes and people uploaded by players onto The Sims’s website,” notes Belk, “will show much that mimics what is and has been and little or nothing that anticipates what might be.”

The Sims trivializes or ignores most of what really matters in human existence. Friends and romantic partners are interchangeable, a matter of changing numbers on a graph which you can see right there on the screen. Philosophy and religion and spirituality are nowhere to be found. Your Sim will never have her life changed by a book, sit at the bedside of a dying loved one, have an epiphany while out on a long walk, serve a country or a cause, create a work of art (even though she can learn to paint), commit a flagrant act of kindness, attend a funeral, or become the victim of a crime and have to choose whether to forgive or hate the perpetrator. The life the game presents is a plastic pantomime of the real thing.

There are no such things as social problems in The Sims. If your Sim has a problem, it’s her — or rather your — own fault. The only way to get ahead in its picket-fence world is to play the game of life according to the rules of the day. Some folks called it thinly veiled propaganda for the American consumer economy, of a piece with the hollow words of President George W. Bush when he said that the best way to be a patriot in the wake of the September 11 attacks was to go spend money at Disney World. (Never fear, Maxis had you covered with the Vacation expansion…)

Personally, I don’t get too up in arms about this commoditization of the human experience. I’m not even inclined to get too upset about the way the game places the needs of your Sim’s bladder on a level footing with her social needs; after all, it really is pretty darn hard to focus on anything else when you desperately need a toilet. I presume that few of the people who bought or played The Sims did so expecting life-changing epiphanies. Yet I can’t quite bring myself to ignore the perhaps uniquely American emptiness that yawns at the center of it either. In the end, it’s a social-status simulator in which everything and everyone are transactional. I’m sure that this is more an artifact of its algorithmic nature than any conscious or even unconscious attempt to inculcate a set of values, but it still leaves me with an uncomfortable feeling.

Setting questions about values aside, many old-school gamers responded to The Sims by asking why they should want to take out virtual trash and water virtual plants when they do plenty of that sort of thing in real life. In this case, too, I kind of find myself agreeing with the critics. Is it worth taking time out from your real life to lead a mundane virtual one? Will Wright said to Computer Gaming World shortly after The Sims’s release that, based on Maxis’s early playtesting, six out of every ten people loved it, two were lukewarm on it, and two hated it. I lie somewhere between the second and third groups, for reasons that were pinpointed by reviewer Chris Lombardi in the same issue.

The fundamental challenge of The Sims, as in life, is to find the time to meet all of your Sim’s basic needs and still have time to boogie. Each day, you face questions like: Do I spend my two hours of free time tonight playing pool with my friends or studying for a promotion? Computer games until the wee hours or a good night’s sleep? A bath tonight or in the morning? It’s the juggling act that each of us performs every single day, each in our own way and to varying degrees of success. It’s life in all of its drab, humdrum detail, as well as in the little victories and achievements that make it all worthwhile.

What can I say? These do not strike me as Interesting Decisions in a ludic sense, any more than “life in all of its drab, humdrum detail” strikes me as an overly compelling ludic premise. For me, The Sims is a fiddly, relentless minutiae simulator. The problem is doubtless compounded by my genetic difficulty with games that don’t give me concrete goals to work toward. (This same difficulty is the reason that I struggle with the entirety of Maxis’s catalog, SimIsle excepted.) Sheepish as I am to admit this, if The Sims gave me a to-do list, maybe I would suddenly feel differently about it.

But having said all that, I’m well past the stage of life where I judge people for their entertainment preferences or try to convince them to like the things that I do. So, by all means, play The Sims if you enjoy it, subject only to the one caveat that I attach to all games: don’t do so too much, to the point of neglecting your real life that is filled with all of those other, more ineffable experiences that The Sims does not and cannot include.

Whether you happen to be a fan or not, there is no denying that The Sims changed the face of gaming forever, possibly even more so than Will Wright’s previous masterstroke SimCity. In so doing, it opened up what had become a dismayingly homogeneous industry to fresh aesthetics, themes, and approaches, paving the way for other games that I do enjoy more than this one. For this reason alone, Wright must be considered one of the most important — not to mention commercially successful — people ever to work in his field. It may be that the only person whose games have reached more and more diverse people and created more change in their image is his old friend Shigeru Miyamoto. That’s not bad for a man who never really wanted to be a game designer at all, who preferred the label of software toy-maker. Under either heading, Will Wright transformed the way that people play.



Did you enjoy this article? If so, please think about pitching in to help me make many more like it. You can pledge any amount you like.


Sources: The books Games That Sell! by Mark H. Walker; Building SimCity: How to Put the World in a Machine by Chaim Gingold; Game Design Theory & Practice (2nd ed.) by Richard Rouse III; Time, Space, and the Market: Retroscapes Rising, edited by Stephen Brown and John F. Sherry Jr.; A Pattern Language: Towns, Buildings, Construction by Christopher Alexander, Sara Ishikawa, and Murray Silverstein with Max Jacobson, Ingrid Fiksdahl-King, and Shlomo Angel. Computer Gaming World of May 2000; Retro Gamer 115.

Online sources include Tristan Donovan’s interview with Will Wright for the old Gamasutra site, “SIMply Divine: The Story of Maxis Software” by Geoff Keighley for GameSpot, “Will Wright Inspired to Make The Sims After Losing a Home” by Tracey Taylor for Berkeleyside, “The Kiss That Changed Video Games” by Simon Parkin for the New Yorker website, “The Untold Story of The Sims, Your First Favorite Jazz Record” by Alex Robert Ross for Vice, and an online chat with Will Wright conducted by CNN on January 20, 2000.

Where to Get Them: The Sims and all of its expansions can be purchased as a “Legacy Collection” on Steam.

 
 

Tags: , , ,

The Life and Times of Maxis, Part 2: SimWorld

Would you trust these two men with your investment? Jeff Braun and Will Wright shortly before Maxis’s IPO.


This article tells part of the story of Maxis Software.

Maxis came into 1994 riding high on the wings of SimCity 2000, the hit of the previous holiday season. They followed up that success with… SimHealth: The National Health Care Simulation. This may sound like satire, but it was all too real.

To be clear, SimHealth was not, say, a playful look at the life of a hospital, with eccentric doctors and bizarre made-up diseases. It wasn’t even an earnest simulation of performing surgery. No, this was exactly what it said on the tin, a simulation of healthcare in the most macro sense. It was, to put it another way, literally a simulation of government bureaucracy. If you were getting bored with saving fantasy worlds from evil wizards and fighting the Second World War over and over, said Maxis, you could buy this game and engage with some truly exciting dilemmas. “Who gets health insurance? What does it cover? Who pays for it? How do you control costs? How will your plan affect business, physicians, specialists, insurance companies, lawyers, medical research, and the average citizen?” Hours and hours of fun for the whole family, am I right? Um… am I right?

Ironically, this most defiantly esoteric Maxis game ever arose from the same set of circumstances that gave birth to the multi-million-selling SimCity 2000. The venture capitalists at Warburg Pincus Ventures, who had traded $10 million in cash for 30 percent of Maxis’s equity and a seat on its board in 1992, thought that the company could find a profitable sideline — or possibly even a mainline — in making simulations for Corporate America to use for internal training and planning. The sticking point was that neither Will Wright nor any of the other idealists around him had much interest in such things. Jeff Braun’s solution was to buy a little software house called Delta Logic, built around a couple of refugees from the wreckage of Digital Research, the firm that had made the most popular business-oriented personal-computer operating system in the world until Bill Gates and Microsoft came along. Since then, John Hiles, Greg Rossi, and their colleagues at Delta Logic had been tinkering with social simulations of various stripes. They were now to be rebranded Maxis Business Simulations, given a lot more money than they had ever had before, and set loose to realize the venture capitalists’ dreams. “There’s the potential for this new business to be bigger than the games business,” enthused Braun dutifully.

That didn’t happen, although a surprising number of projects came and went during the less than two years that Maxis Business Simulations was a going concern. There were a simulation of an oil refinery for Chevron, a simulation of the power grid for the Pacific Gas & Electric Company, a simulation of a telephone network for a communications consultancy, a simulation of pollution management for the Environmental Protection Agency, and a simulation of the closure of a military base (?) for the new Maxis subsidiary’s hometown of Monterey, California. With the exception of a prototype version of SimRefinery, none of the actual software that resulted from these initiatives has been preserved, to whatever extent it came to exist at all. Constant churn was a fact of life at Maxis Business Simulations. It seems safe to say that the idea of doing corporate or civic planning via interactive simulations was one of those that looked better on paper than it tended to turn out on the computer screen.

SimHealths origin story isn’t that different from most of the other Maxis Business Simulations projects. It began when the Markle Foundation, a bedrock of old-school American philanthropy, requested a simulation to illuminate one of the hottest political debates of the 1990s, as it seems to be of every decade: what, if anything, to do about the uniquely strange and byzantine American healthcare system. President Bill Clinton, who in January of 1993 became the first Democrat to occupy the Oval Office in twelve years, made reforming healthcare on a more efficient, equitable basis the flagship issue of his first term. So far, so good. But then he decided to place the project under the stewardship of his unelected wife Hillary Clinton, who was already at this early date a lightning rod for Republicans and no small percentage of independents. It turned into an epic fiasco, burning up most of the president’s political capital without benefiting anyone and setting the stage for a brutal midterm shellacking at the hands of the Republicans. Indeed, many political scientists argue that the current, hyper-partisan era of American politics dates from those 1994 midterms, when Republicans under their future Speaker of the House Newt Gingrich first learned to characterize their Democratic opponents as not just misguided about means but evil in intent.

Is this the weirdest game ever to be reviewed in Computer Gaming World?

If this is the case, then SimHealth stands today as an oddly touching artifact of the Before Times, when there was still more than lip service paid to the notion of politics as a good-faith quest for the policies that yield the most good for the most people. In that spirit, the Markle Foundation decided to make SimHealth available to everyone as a boxed… well, I suppose the word “game” is the only one to use, considering that it was sold in the games section of those stores who were willing to sell it at all, and considering that it was given a two-page review by Computer Gaming World magazine, alongside the likes of Wing of GloryAlone in the Dark 2Carriers at War IIRavenloft, and Aces of the Deep. (One of these things was not like the others…)

SimHealth even played like a game, albeit one with unusually sober — not to say boring — subject matter; in some ways, it was actually more of a game than Maxis’s earlier SimEarth or SimLife. You start out with your political party having just won an election, and have to try to maintain your mandate and with it your chosen approach to healthcare through the elections that follow. Everything about it is incongruous. Once it was decided to release SimHealth publicly, the main Maxis office pitched in with art drawn in the house style; inevitably, it brings to mind nothing so much as good old SimCity. All told, SimHealth stands today as one of the more bizarre games ever to be put in a box and sold. One doesn’t know whether to hail it for its chutzpah or laugh at it. Whatever else you can say about it, it certainly didn’t embrace the then-prevalent gaming mono-culture of sci-fi, fantasy, and war.

The discussion that sprang up around SimHealth was, to the extent it sprang up at all, dogged by the same fallacies that swirled around SimCity. For all that SimHealth presented itself as an early example of a “serious game” with something to say about the real world, there was no reason to believe that this purported “simulation,” created by a few harried programmers in well under a year, could possibly be modeling anything as complicated as a national healthcare system for one of the more populous and diffuse countries in the world with any sort of accuracy. Approaching it as a gamer would — because how else was he supposed to approach it? — Computer Gaming World’s Keith Schlesinger quickly found a winning strategy.

All you have to do is adopt an extreme libertarian ideology, eliminate all federal healthcare (including Medicare!), and cut other government services by $100-300 billion per year. The polls never sink below 57 percent favorable, and the behavior point total is almost a perfect 400. You can set the game to “Run to the End” and breeze to victory while hardly touching the mouse or keyboard. Institutions wax and wane, but the “invisible hand” of the free market keeps everything pretty much in line.

There is one catch: the entire private insurance industry goes bankrupt in the first three months of the game! That means no one has any protection even from catastrophic disease and injury. One wonders what people are doing about this. Why aren’t [they] rioting and burning down Main Street?

Once they’re through celebrating this endorsement of their views, even the extreme libertarians among us may, if they’re being honest with themselves, feel a twinge of concern at the pernicious idea that simulations like this one are somehow modeling abstract truths instead of being a morass of oversimplifications, biases, and highly questionable judgment calls on the part of their designers. I do suspect that the anecdote above arose by accident rather than out of any deliberate attempt to instill an ideological slant into the software. In fact, to their credit I suppose, the designers of SimHealth included an “Assumptions” menu which lets players tinker with the most fundamental rules of the simulation itself. But it does rather raise the question of what the point of it all is. What am I meant to learn from SimHealth if even the designers aren’t sure what the rules should be? The whole project was noble and well-meaning — and also kind of useless.

Unlike SimCity, though, SimHealth came and went quickly enough that the debate around it hardly got started. By the time it was released to retail in early 1994, Maxis was already in the process of divesting themselves of their brief-lived Business Simulations division. “The gap between building simulations for entertainment and simulations for businesses was too great,” said an older and wiser Jeff Braun. “One was a product, the other was a service.” The former Maxis Business Simulations would soldier on independently for a while as Thinking Tools before closing up shop for good.

In his (excellent) article on this side quest, Phil Salvador of The Obscuritory states that Maxis sold about 50,000 copies of SimHealth. I find that number extremely surprising, assuming it’s correct — not because it’s so low but because it’s so high. My guess would have been more in the range of 5000 copies — and that’s if I was feeling generous.


Regardless of the exact numbers that SimHealth did, we can say with confidence that Maxis’s second and last Sim title of 1994 sold far more. SimTower: The Vertical Empire still stands today as the best-remembered Sim game of the 1990s that isn’t a SimCity. That said, this particular apple didn’t fall that far from the SimCity tree. Your task in SimTower as well is that of building a functioning community — but here you’re to do it all inside a single skyscraper. It feels only appropriate that the game was conceived and coded in Japan, where people have long been in the habit of living small and vertically.

Yoot Saito, the mastermind of SimTower, was once memorably described by Jeff Braun as “a rock-star groupie who became a rock star himself.” He fell in love with SimCity not long after its release: “I didn’t leave the Macintosh for almost 24 hours. I so much admired the creator of this genius and crazy software.” He chased Will Wright into a hotel lobby during a MacWorld exhibition like a screaming girl from A Hard Day’s Night. He then proceeded to tell his cornered hero about his idea for a SimCity inside a single building. One thing led to another, and in late 1994 Maxis brought his SimTower to the West.

Your goal in SimTower — and yes, you do have a goal — is to eventually erect a five-star building, meaning one that has at least 100 floors and at least 10,000 more or less happy residents, with a cathedral as the cherry on top of the imposing sundae. How you do this is entirely up to you. Your biggest challenge, the direct equivalent to SimCity’s need for efficient roadways and subways, is your elevator system; get this wrong and your gleaming utopian tower will soon collapse into impatience and discord. As Yoot Saito would be the first to tell you, the rest of the game was really born out of his odd fascination with elevator systems.

Late one night, I was waiting in my building’s lobby for an elevator. There are two cars in my building’s shaft. I pushed the button, but instead of the closest car coming to my floor, the farthest car arrived first. In my usual questioning way, I asked myself, “What happened here? How are these cars logically synchronized?” That was the beginning of SimTower.

Yoot Saito was truly a man after Will Wright’s heart.

SimTower has many of the cute, juicy touches that lent personality to the Japanese ports of SimCity. Add a restaurant to your tower, and a food critic will stop in to eat and pass judgment. Play the game on Christmas, and Santa Claus will fly overhead. SimTower isn’t as deep or complex as SimCity 2000, but this was probably more of a plus than a minus for the nascent casual-gaming market of the mid-1990s. It became a surprise hit for Maxis, giving them their second holiday season in a row with a new game near the top of the bestseller charts. Suddenly the bottom line was looking pretty darn good.

As a result, Jeff Braun made a fateful decision, one that he would later deem “a big mistake.” He took Maxis public on May 25, 1995. It “was just sort of the thing to do, and it just felt right,” he says. This year would prove the first of the dot.com boom, highlighted by Netscape’s $3 billion public bow in August. Maxis’s own $35 million IPO may have been a vastly more modest affair on the face of it, but the cachet of SimCity and SimTower were enough to get Jeff Braun and Will Wright mentioned in some of the same conversations as Netscape’s dynamic duo of Marc Andreessen and Jim Clark. In December of 1995, Braun showed up alongside Andreeseen in Newsweek’s list of “The Net 50,” despite the fact that Maxis had yet to release a game that connected to the Internet in any sort of way. Newsweek was blissfully unfazed: “Look for Braun to start moving his world of fantastic simulations to the Internet sometime next year.” Maxis took out a lease on the glass-enclosed sixth floor of a glitzy high-rise in Walnut Creek, California.

In addition to the generalized furor over all things tech, the Maxis IPO was fueled by the excitement surrounding educational and edutational software. These categories were widely considered to have a very lucrative future in the offing, as school administrators bought the new generation of multimedia computers and looked for things to do with them in classrooms at the same time that parents were doing the same at home. If Maxis perhaps wasn’t quite of the category, their games’ non-violent nature and engagement with real-world topics placed them solidly adjacent to it. Their new Sim games of 1995 reflected a desire to align themselves still more closely.

SimTown was a kid-friendly, simplified version of SimCity, wrapped in a bright, cartoony aesthetic. “This town is run by kids!” screamed the back of the box. Unfortunately for Maxis, children tend to require less simplification than many adults think they do, even as they can smell adult condescension from a mile away. SimTown wasn’t a big seller.

SimIsle: Missions in the Rainforest was a worthier effort with a lot more heart and soul put into it, so much so that I’m prepared to call it the hidden gem in the 1990s Maxis catalog. Developed in collaboration with a British house called Intelligent Games, it plunks you down in a tropical archipelago of 30 islands. But instead of settling for being just another “SimCity except somewhere else,” SimIsle brings some fresh ideas to the table. Each of the islands constitutes a set-piece scenario, complete with a set victory condition and a defined endpoint. As Computer Gaming World wrote in its review, SimIsle has “a lot of good old SimCity charm with a bit more game than is customary in the series.” It replaces the usual Sim abstractions with a distinctive and compelling sense of place, fed by digitized photographs and an evocative soundtrack. At times, it feels less like a Maxis Sim game that it does an early version of Tropico.

Some of the islands are to be developed for tourism, some for resource extraction or industry, some as wildlife preserves. There’s an environmental message to SimIsle, but you’re never beaten over the head with it in that way that can make some other Maxis games feel so simultaneously noble and dull. “If you want to explore what happens when you tarmac an entire island, that’s your affair,” said lead designer Matthew Stibbe. “Alternatively, you can see if it is possible to marry economic growth, with all its benefits, to ecological safety.” The game succeeds in making you want to preserve and conserve as much as possible, by making its archipelago an interesting and even exciting place to explore. Finding a new species of flora or fauna or an unsuspected ancient ruin is a genuine thrill. Destroying such things can be devastating, but sometimes you feel like you have no choice if you want to satisfy a scenario’s goals. You often have to settle for preserving what you can. These agonizing tradeoffs convey a different sort of realism than most Maxis games, showing the world to be a messy place where exploitation and conservation are less binary choices than a precarious balancing act.

In keeping with this Sim game being not only about nature, science, and other such abstract forces, it features, for the first time ever — outside of the SNES SimCity, that is — individual people. You can accomplish things on your island only by directing “agents,” each with a picture, a capsule biography, a personality profile, and a unique set of skills and abilities, from “flora and forestry” to “construction,” “negotiation” to “local ecology.” You can even find agents with underworld connections, if that’s your bag; setting up a drug trade is a possibility for the unscrupulous player. Your agents can improve their skills with experience and training, further tightening the bond between you and them. The lack of just such a personal touch is one of the reasons that most Maxis Sim games leave this player feeling kind of cold. Everybody is different, of course, but I have to say that SimIsle is the direction that I wish Maxis had begun to take much earlier and had stuck to thereafter.

The agents system of SimIsle smacks oddly of Sir-tech’s cult classic Jagged Alliance, which came out the very same year.

Indeed, the places where SimIsle disappoints me most are the ones where it misses opportunities to take its freshly “game-like” approach still further. The archipelago itself positively begs for a larger campaign to structure your play. As it is, though, SimIsle maintains no memory of what you’ve done before as you go from island to island; nor can you take the same team of agents with you. But this failing isn’t enough to keep it from becoming my personal favorite Maxis game. I first played it shortly after its release, played it again more briefly for this article, and really wish I had the time to play it some more. I highly recommend it to anyone who is left a little cold by the rest of Maxis’s catalog for any or all of the same reasons that I am.

Sadly, SimIsle was not rewarded for its uniqueness. It was no more commercially successful than SimTown, being written off by most gamers as just more of the same, from a publisher that had become rather too much of a known quantity. From the perspective of the marketplace, the bloom had come off the Sim rose just as the post-IPO firm was supposed to be taking flight.

SimGolf wasn’t a bad effort as golf simulations went, but it never had a prayer against the Links juggernaut.

So, by 1996, the Sim line was flailing, casting about for the right combination of familiarity and freshness that could make it feel relevant again. There were no fewer than five new Sim-branded games that year, and they were well and truly all over the place. SimPark was a simulation of a wildlife park, but with the kid-friendly branding and simplifications that had made their dubious debut in SimTownSimTunes was a visual music-making system that could have been released under any of a dozen educational-software imprints. And SimGolf was nothing more nor less than a straightforward golf simulator, dumped onto a market that was already full to bursting with the things. One of the most distinctive brands in the industry just a couple of years earlier, the Sim line was now being diluted to the point of meaninglessness. Was it for adults or children? Was it games or educational software? Was it products in the mold of SimCity or just anything upon which you could conceivably slap the label of “simulation?” Maxis seemed to want the answer to all of these questions to be “Yes,” full stop. The result was just confusion.

Even the stuff that ought to have been in their wheelhouse didn’t quite pan out. SimCity 2000: Network Edition added online-multiplayer functionality to the venerable city-builder; rumors of this product were presumably the reason that Newsweek had buzzed about Jeff Braun “moving his world of fantastic simulations to the Internet.” But the sales of the finished product weren’t anything like Maxis had hoped. This may have been because they did nothing to update the graphics or interface, leaving the Network Edition looking like exactly the three-year-old game it otherwise was. But more likely, it was down to the questionable value proposition of a multiplayer SimCity. How was such a thing supposed to work, given that competition was so conspicuously absent from the franchise’s DNA? Maxis’s approach was to let you divide a city into different zones, then place one person in charge of each of them, to compete or cooperate as they chose. It felt more forced than exciting.



You may have noticed that the name of Will Wright has appeared in this article much less than it did in the one before it. There’s a reason for that: after completing (the non-networked) SimCity 2000, he had stepped back from day-to-day game development to work on various “research projects.” One of these tried to make a reality out of the idea he had presented at the 1990 Game Developers Conference, for a new ecosystem of interoperable games. He was no longer so bold — or naïve? — as to imagine the entire industry joining him in this effort right away, but he did think it could be a very good thing for Maxis games at least to learn to talk to other Maxis games, as summarized in Game Developer magazine:

Most Maxis games occupied some of the same intellectual real estate, which made them natural candidates to extend into each other. For example, one of the central strategies a player can use in SimCity is to develop a mass-transit system to help the city grow. In A-Train, a game Maxis imported from the Japanese company Artdink, the goal is to help a community expand using a commuter rail line. In the same vein, the monster towers of the future in SimCity 2000 are similar in nature to those players build in another Japanese game imported by Maxis, SimTower.

Thus the SimWorld project was born. The goal of this project was twofold. It would allow Maxis to create future games with the capability to extend and merge into other preexisting and future games. And it would leverage SimCity and the millions of hours of playtime that people have invested in SimCity saved games into new markets.

The most intriguing Sim title of 1996 was also the first concrete manifestation of all this. SimCopter was a helicopter simulator, marking the brand’s first foray into first-person vehicular simulators of any sort and Maxis’s first foray into 3D graphics. But this was a helicopter simulator with a difference. You could take a city you’d built in SimCity 2000, load it into SimCopter, and fly around looking at your handiwork close-up. No one could say it wasn’t a cool gimmick.

SimCopter did have other things to enjoy as well once you got past the gimmick. The flight simulator itself was pleasantly forgiving, making it stand out in a genre that usually wanted to make games only for the hardcore of the hardcore. Best of all, SimCopter took a lesson from SimIsle in structuring the time you spent with it more than did your typical Maxis game. If you play in “career” mode, you take the role of an urban helicopter pilot tasked with putting out fires — mostly figuratively, occasionally literally. In addition to extinguishing the literal conflagrations, your assignments can involve traffic jams, rooftop rescues, crowd control, police chases, medical evacuations, and VIP taxi flights. If you do well, you earn money and points and advance to better helicopters, bigger cities, and more difficult (and profitable) missions.

It’s surprisingly engaging, although it can leave you wanting more in the longer term. The reality that there are less than ten types of mission in all, procedurally generated over and over, sets in well before the campaign is over. “We could not miss Christmas, which meant we really couldn’t aim too high,” admits Will Wright. “Had we had another six to eight months to work on it, I think it would have turned out much, much better.” (How many times has a game developer said this?)

Immediately after its release, SimCopter and Maxis were embroiled in an unwelcome but kind of hilarious controversy, when it was discovered that the virtual citizens over whom the helicopter flew included crowds of strapping young men in Speedos who enjoyed kissing one another. The source of the transgression proved to be a gay Maxis artist named Jacques Servin, who wished to make a statement about the “aggressive heterosexuality” of game culture: “I’ve always wanted to be an activist, but activism is so moribund now. Do you think these heads of corporations are going to walk into an art gallery and say, ‘Oh, wow, I was wrong?’ Symbols are so much more powerful where you don’t expect them.” Servin was fired by Maxis after his handiwork hit the press — not, the company was careful to say, for being gay, but for inserting unapproved content into its game. Nonetheless, a brouhaha with the gay advocacy group ACT UP followed, despite Servin’s own statement that his former employer was actually “very enlightened” about sexual orientation. This was publicity that Maxis emphatically didn’t need, managing to cast them in an unfavorable light with the accepting and the bigoted alike. It became one more data point to contribute to the growing impression of a company for which absolutely nothing was going right anymore.

I never saw any men in Speedos during my time with SimCopter; Maxis patched them out of the game just as quickly as they could. But never fear, the Internet preserves all.

Just eighteen months after the IPO, Maxis was looking like a hot mess. From a high of $50, the stock price was down to barely $10. None of their Sim games since SimTower had done very well, nor any of the grab bag of non-Sim oddities they had published on behalf of other studios. Meanwhile the excitement over educational software that had done so much to fuel the IPO was starting to look misplaced, for the same reason that the buzz over non-game multimedia CD-ROMs was fizzling out: the World Wide Web was eating those offline software markets alive.

Development had begun on a new SimCity, traditionally the company’s savior, but that wasn’t going well either, having gotten bogged down in a bid to make the game fully 3D. A free-floating 3D camera, noted Will Wright,

takes what used to be a very simple, Lego-like thing and turns it into AutoCAD. “What am I looking at? Oh, I see, I’m facing the building two inches away.” It becomes that kind of experience. So that was part of it. The other part was the technology. Without going with really severe restrictions on what you could build, we just couldn’t have a decent frame rate and have the level of detail that we could have in an isometric viewpoint.

Both of these issues were exaggerated by the need to cater to the casual players who made up so much of the SimCity base, who were neither comfortable with the latest generation of 3D games nor had computers that were up to running them all that well.

Maxis’s investors pronounced themselves “concerned” by the direction of the company, which was on track to lose more than $2 million in 1996. What revenues there were were mostly still being generated by SimCity 2000, now a steady-selling staple of the world’s budget bins. But even that wouldn’t last forever.

Jeff Braun was forced out of the company he had co-founded by his own board of directors in August of 1996. His replacement in the CEO’s chair was one Sam Poole, a former Disney executive who had joined up as sales manager at the time of the IPO. His marching orders from the board were clear: find a buyer before the ship of Maxis went under.

In June of 1997, Electronic Arts stepped up. No cash changed hands, only stock, worth $125 million on paper. “I entered the deal with a lot of trepidation,” says Will Wright. “I knew there were going to be layoffs, but not doing the deal would have meant even more layoffs down the road.” For its part, EA made it clear that it was buying Maxis for just one reason: to get its hands on the SimCity brand, the one item in the Maxis portfolio that it considered to be worth anything at all. Whatever was left of the old dream of SimEverything died the moment the contract was signed. As far as EA was concerned, Maxis was a one-product studio.

The last couple of shambling embodiments of the old spirit, in the form of products that had been a little too far along in the pipeline to cancel, came out in the months just after the acquisition. Streets of SimCity operated on the same principle as SimCopter, letting you drive instead of fly around your handmade city. But for some reason — one has to suspect the intervention of someone at EA, looking jealously at Activision’s hit Interstate ’76 — your car had guns to shoot the other cars on the road with. The whole thing was as ineptly implemented as it was abjectly tone-deaf. Had EA actually understood the brand it was buying? You know, the one that was always described as not being violent in the way of all those other videogames?

SimSafari was more defensible, if not ultimately all that much more compelling. It was another game in the mold of SimTown and SimPark, a “kid-friendly,” simplified version of the SimCity approach. But one might be excused for asking how big a difference there really was between building a “wildlife” park, as in SimPark, and a “safari” park, as in SimSafari.

Meanwhile back at the Maxis ranch, major if fairly predictable changes were afoot. Relieving Sam Poole of his duties, EA parachuted in a manager named Luc Barthelet to fix whatever ailed its acquisition and to get that all-important third SimCity back on track. In a report back to the mother ship that surprised absolutely no one, Barthelet bemoaned a lack of “focus” and “structure.” He set out to re-instill these qualities. “I told people from day one: we are going to be a PC studio that does top-ten products, and if we did that we’d stay in business,” he says. “A lot of people weren’t interested in this. They wanted to do their own little projects, and didn’t care if they would sell. Those people left.”

Barthelet may not have been an overly cuddly or sentimental sort, but his tough love was probably just what the moribund next-generation SimCity needed. He could see no reason why that game needed to be 3D. Stop screwing around, go back to the tried-and-true isometric view, and just get it done, he said. Ocean Quigly, the project’s art director, compares his early days on it to “running through the fog not knowing where we were going.” The fog was now lifted and a compass provided.

SimCity 3000 would prove another picture-perfect sequel. By some metrics, it expanded dramatically on its predecessor, with the possibility to construct 400 instead of just 100 different types of buildings in a city up to four times as big, all with sharper graphics and a cleaner interface. But it was still the SimCity that its fans and the media knew and loved, exuding a “don’t fix what isn’t broke” philosophy from its every pixel. The additions were largely commonsensical and cosmetic. The larger selection of buildings let you give your city a distinctive American or European aesthetic; the likes of the Statue of Liberty and the Eiffel Tower were waiting for you right there in the menu. Borrowing a page from Civilization II, a franchise it had partially inspired (have I mentioned that game design is always a dialog?), SimCity 3000 had a panel of municipal department heads to offer their advice to mayors who happened to be in doubt. (Also like in Civilization II, some of their advice was wise, some of it self-servingly parochial.) The biggest conceptual leap was to turn your city from an isolated specimen under glass into a part of a larger nation, complete with other towns with which you could trade electricity, water, or sewage services and enter into other sorts of business arrangements. There was still no structure beyond abstract building for its own sake, no campaign or thread of narrative, placing it at odds with some other city-builders, such as the historical ones made by Impressions Games. That said, a substantial percentage of SimCity players probably wouldn’t want it any other way. This SimCity product was as in harmony with its brand as Streets of SimCity had been off-key.

Maxis learned some of the advantages of hooking up with a deeper-pocketed parent when the time came to demonstrate the game publicly for the first time at the 1998 E3 trade show: EA flew former New York City Mayor Ed Koch in to promote SimCity 3000 in person. Sure, as a spokesman Koch was better primed to entice the mainstream media than the E3 gamer crowd — but then, that was kind of the point, wasn’t it? EA knew well where its SimCity bread would be buttered.

The original targeted release date was the Christmas of 1998, but Maxis couldn’t quite get it done to their satisfaction. Rather than let this make-or-break game out in a less than ideal state, Luc Barthelet convinced the EA brass to let them hold onto it a few months more. SimCity 3000 finally shipped in February of 1999, whereupon it became clear that the franchise’s winning streak would continue, as it became Maxis’s first really successful game since SimTower four and a half years earlier. In fact, it became the best-selling single computer game of 1999. (Admittedly, its release date so early in the year did give it a longer runway than almost all of the competition.) Luc Barthelet had his top-ten game and then some; the old SimCity magic of mixing casual with hardcore appeal still worked. This was a game that your Starcraft-playing buddies could enjoy as much as your middle-aged dad. By the end of 1999, Maxis could boast of having sold more than 7 million copies of its three SimCity games over the past decade. In comparison to this figure, the sales of all of their other games combined constituted little more than a rounding error.

There was one thing that did separate SimCity 3000 from its two predecessors: Will Wright contributed practically nothing to it. “It was a running joke around Maxis that whenever the SimCity team would come to ask me for advice I would go running,” he says. “They finally gave up. You know, the day they shipped SimCity 3000 was one of the happiest days of my life. They proved that we have a team within Maxis that knows how to build a SimCity without my involvement.”

Yet Will Wright was still employed at Maxis. So, just what was he doing, wondered plenty of people at the time. The answer to that question would soon turn the games industry on its head, making all of SimCity’s success look like the merest prelude to this, the mad scientist’s main act.



Did you enjoy this article? If so, please think about pitching in to help me make many more like it. You can pledge any amount you like.


Sources: The books Games That Sell! by Mark H. Walker; Building SimCity: How to Put the World in a Machine by Chaim Gingold; Game Design Theory & Practice (2nd ed.) by Richard Rouse III; SimIsle: The Official Strategy Guide by Douglas R. Brumley, Phil Powell, and Selby Bateman; The Secret History of Mac Gaming (expanded edition) by Richard Moss. Computer Gaming World of June 1994, April 1995, December 1995, March 1997, and May 1999; Retro Gamer 115; The American Prospect of March/April 1994 and March/April 1997.

Online sources include the collection of Maxis articles by Phil Salvador at The Obscuritory, Tristan Donovan’s interview with Will Wright for the old Gamasutra site, and “SIMply Divine: The Story of Maxis Software” by Geoff Keighley for GameSpot.

Where to Get Them: Of the Maxis games described in this article, only SimCity 3000 is currently available as a digital purchase.

 

Tags: , , , , , , ,

Bullfrog in the Dungeon


This article tells part of the story of Bullfrog Productions.


In January of 1995, Electronic Arts bought the prestigious British games studio Bullfrog Productions for an undisclosed sum that was rumored to be in the neighborhood of $45 million. The lives of the 35 or so people who worked at Bullfrog were transformed overnight. Peter Molyneux, the man who had built the studio’s reputation on the back of his “god game” Populous six years earlier, woke up the morning after the contract was signed with real money to spend for the first time in his life. For him and his colleagues, the new situation was gratifying, but also vaguely unsettling.

The acquisition came with all of the expected rhetoric from EA about “letting Bullfrog be Bullfrog.” Inevitably, though, the nature of the studio’s output changed in the years that followed. EA believed — and not without evidence — that series and brands were the key to long-term, sustainable success in the games industry. So, it encouraged or pressured the folks at Bullfrog to connect what they were doing now with what they had done before, via titles like Magic Carpet 2Syndicate WarsPopulous: The Beginning, and Theme Park World. Iteration, in other words, became at least as prized as the spirit of innovation for which Bullfrog had always been celebrated.

And where the spirit of innovation refused to die, you could always do some hammering to make things fit. The most amusing example of this is one of the two best remembered and most beloved games to come out of this latter period of Bullfrog’s existence. What on earth could be meant by a “theme hospital?” EA couldn’t have told you any more than anyone else, but the name sounded pretty good to it when one considered how many copies Theme Park had sold the year before the acquisition.


Theme Hospital was born as one of about a dozen ideas that Peter Molyneux wrote on a blackboard during the heady days just after the acquisition, when Bullfrog had a mandate to expand quickly and to make more games than ever before. This meant that some of the Bullfrog old guard got the chance to move into new roles that included more creative responsibility. Programmer Mark Webley, who had been doing ports for the studio for a few years by that point, thought that Molyneux’s idea of a simulation of life at a hospital had a lot of potential. He plucked it off the list and got himself placed in charge of it.

In the beginning, he approached his brief with a measure of sobriety, not a quality for which Bullfrog was overly known. He even thought to do some field research at a real hospital. He and visual artist Gary Carr, the second person assigned to the team, talked the authorities at Royal Surrey County Hospital, conveniently located right next door to Bullfrog’s offices in Guildford, England, into giving them a behind-the-scenes tour. It proved rather more than the pair had bargained for. Videogame violence, which Bullfrog wasn’t noted for shying away from, was one thing, but this was something else entirely. The two young men were thrown out of at least one operating theater for retching. “I remember we watched a spinal operation one morning, which was bad enough,” says Carr. “Then the person who had been assigned as our guide said, ‘Alright, after lunch we can pop down to the morgue.'” It was right about then that they decided that an earnest simulation of life (and death) at a National Health Service hospital might not be the right way to go. They decided to Bullfrog the design up — to make it silly and irreverent, in Bullfrog’s trademark laddish, blackly humorous sort of way. “It’s not about how a hospital runs,” said Mark Webley sagely. “It is about how people think [it] runs.”

Webley played for a while with the possibility of making Theme Hospital a sort of satirical history of medicine, from the Middle Ages (“when curing people usually meant hacking their legs off with a bloody great saw, covered in leeches”), through the Victorian Age (“lots of mucking about with electric shocks and the like”), and on to the present and maybe even the future. In the end, though, that concept was judged too ambitious, so Theme Hospital returned to the here and now.

That said, the longer the slowly growing team worked, the less their game seemed to have to do with the real world of medicine that could be visited just next door. A former games journalist named James Leach came up with a matrix of absurd maladies: Bloaty Head (caused by dirty rainwater and putrid cheeses), Broken Wind (caused by too much exercise after a big meal), Slack Tongue (caused by too much vapid celebrity gossip), Infectious Laughter (caused by hearing too many sitcom laugh tracks), King Complex (afflicting Elvis impersonators who spend too much time in character). The cures are as imaginative as the diseases, generally involving slicing and dicing the patients/victims with one or more horrifying-looking Rube Goldberg contraptions. Theme Hospital is healthcare as those with a deathly fear of doctors imagine it to be.

The finished game is deceptively complex — perhaps a little too much so in my opinion. You start each scenario of the campaign with nothing more than a plot of land and a sum of starting money that never seems to be enough. You have to build your hospital from scratch, deciding where to place each and every room, and then where to put everything that goes inside each room, from the reception desk to the toilets to the even more important soda and candy machines. (This is most definitely not socialized medicine: your hospital is expected to make a profit.) Then you have to hire doctors, nurses, janitors, and administrative personnel from a wide variety of applicants. Figuring out how best to fiddle the countless knobs that affect the simulation requires considerable dedication; every input you make seems to result in a cascade of advertent and inadvertent consequences. Once the action really starts to heat up, Theme Hospital can become as frenzied as any session of Quake or Starcraft. Keeping the pot from boiling over in the midst of an epidemic or a disaster — events that become more and more frequent as you progress further — requires constant manual intervention, no matter how efficiently you’ve laid out your hospital and how judiciously you’ve selected its staff.

Your reaction to all of this will depend on two factors: whether you’re someone who is inherently drawn to open-ended simulations of the SimCity and Theme Park stripe, and whether Bullfrog’s brand of humor causes you to come down with your own case of Infectious Laughter. My answer to both of these questions is a qualified no, which makes Theme Hospital not really a game for me. By the time I’d played through the first couple of scenarios, I could see that the ones to come weren’t going to be all that different, and I just didn’t have the motivation to climb further up the campaign’s rather steep ladder of difficulty. Speaking of which: the campaign is as rudimentary as the simulation is baroque. Each of its scenarios is the same as the one before, only with more: more diseases to cure, more requirements to meet, more pressure. It left me with the same set of complaints I recently aired about the campaign in Rollercoaster Tycoon: I just wish there was a bit more there there, an attempt to provide a more interesting and diverse set of challenges that entailed less building of the same things over and over from scratch.

But, just as was the case with Rollercoaster Tycoon, the game about which I’m complaining today did very well for itself in the marketplace, an indication that plenty of people out there don’t share my peculiar preferences. (Who would have thought it?) Released in the spring of 1997, Theme Hospital was by all indications Bullfrog’s biggest single latter-day commercial success, a game which continued to sell reasonably well for several years, sufficient to reach the vicinity of that magic number of 1 million units. In the process, it became sneakily influential. You don’t have to squint too hard to see some of the DNA of Theme Hospital in the more casual time-management games that became crazily popular about ten years on from it.


Dungeon Keeper’s box sports its mascot, the Horned Reaper. Or, as the lads at Bullfrog preferred to call him, Horny.

The other widely acknowledged classic from this period of Bullfrog’s history took even more time and effort to make than Theme Hospital. In fact, Dungeon Keeper — one does have to breathe a sigh of relief that EA’s marketers didn’t insist that it be called Theme Dungeon! — had the most extended and torturous development cycle of any game Bullfrog ever made.

The idea for Dungeon Keeper came to Peter Molyneux some time before the EA acquisition: in mid-1994, when he had just finished up work on Theme Park.

I was sitting in my car in the middle of a traffic jam. I was bored to tears, waiting for the cars in front to begin moving again. Then the idea of a reverse role-playing game popped into my head. Yes, I thought, this could be a good game. You could have loads of monsters crawling around deep, dark tunnels. You could have the power to control them directly, deal with all their problems and petty grievances. As your dungeon grew, your power would increase. You could mine and hoard gold and have to put down rebellions. On top of all this, you could have the traditional heroes invading the trap-laden dungeon you’d created. I was so deep in thought, I hadn’t realized the traffic had cleared.

Dungeon Keeper was a set of answers to questions that some of the more literal-minded players of Dungeons & Dragons and its ilk, whether played at the tabletop or on computers, had been asking for many years already. Just what was up with those nonsensical labyrinths filled with creatures, traps, and treasure that seemed to be everywhere in the worlds of Greyhawk and the Forgotten Realms? What was the point of them? For that matter, what did all those monsters eat while they were hanging out waiting for the next group of adventurers to come by? And what was it like to be the evil wizard with the thankless job of taking care of this lot?

Peter Molyneux, who called CRPGs “my favorite type of game” even though he had never made one, proposed to find some answers now by approaching the genre from the opposite point of view from the norm. It was one of the most brilliant conceptions he ever came up with: quick and easy to explain to just about anyone who knew anything about computer games, and at the same time something no one had ever attempted to do before. And it was as on-brand as could be, gently transgressive in that trademark Bullfrog way. Molyneux:

You have to hire monsters and keep them employed. How do you keep a monster employed? There are basically two ways to motivate your staff. The first is basically to pay them money, but you really don’t want to do that. The second and preferred method of influence is fear. Say you have a band of ten goblins and you want to maintain their loyalty. What’s the best way to do that? Kill off five of them. Ritual sacrificing is really important. You have to prove to the people below you that you’re evil. This is a dungeon, after all.

When the acquisition negotiations began, Molyneux was already tinkering with Dungeon Keeper alone. He was joined by a second programmer, Simon Carter, in November of 1994. EA loved the idea; in the wake of DOOM, dark and gritty violence was in in games. The two-man project may have been in its infancy, but it was a major selling point for Bullfrog during the talks with EA, who saw it as a game with the potential to become huge.

Alas, after the deal was done EA promptly started to pull Molyneux in way too many directions for his liking. He was given a set of executive titles and the executive suite to go along with it, removing him from the day-to-day work in the trenches.

There was me, this passionate sixteen-hour-a-day coder, who kind of lived with these other blokes who were in the office, and all of a sudden Electronic Arts came in and said, “Okay, right. We think you’re fantastic, and we want to expand your studio, and rather than one game [per year] we can get five games out of you.” The company literally went from 35 people to 150 people in nine months, and suddenly there were all these strangers there. Then they said to me, “We think you’re fantastic. We’re going to make you head of our studio in Europe,” and part of me was thinking this was amazing — these people really think I’m clever — but there was another part of me that felt awful, that this was alien and horrible to me. The foundation stones of my work completely changed…

A man with an obsessive-compulsive relationship to the games he made, who would increasingly acknowledge himself to be neurodivergent in later years, Molyneux was painfully ill-equipped for the role of a glad-handing EA vice president; it just made him uncomfortable and miserable. He was forced to oversee Dungeon Keeper — his own baby! — from afar. “I didn’t have time to look at games,” he says. “I was walking in, spending half an hour with them, and then walking out. That was the worst period of my life. Bullfrog was everything to me, and suddenly I was this character that would have to walk into a room and make instant judgments about things.”

From the beginning of 1995, Dungeon Keeper was extravagantly hyped in the gaming press, which ran preview after preview. This gives us an unusual insight into the stages of its evolution. For example, a preview published in the American magazine Computer Games Strategy Plus in the fall of that year describes a heavy focus on multi-player action. At this point, Dungeon Keeper sounds almost like a proto-Diablo, with the important twist that one player is allowed to be the mastermind of the dungeon itself. For those on the side of Good, “one strategy for building up your character would be to enter the dungeon, kill off a few creatures, and run away. Return later (fully healed, but at a cost), kill a few things, and run away again. Stats are kept for the usual physical attributes as well as experience and monsters killed. One interesting feature will allow players to take their saved [character] with them to other people’s dungeons as well.” A player character who conquered a dungeon’s overlord would have the option to come over to the side of Evil and take his place.

In truth, however, progress throughout 1995 was slow and halting, a matter of two steps forward and one step back — and sometimes vice versa. “We spent quite a bit of time just messing around with the idea, but not getting very far in terms of a design,” admits lead artist Mark Healey. Dungeon Keeper may have had a brilliant core concept, but there was no clear vision of how that concept would be implemented, of what kind of game it would ultimately be at the gameplay level. Was it primarily a puzzle game, a kind of Tower Defense exercise where you had to set up your monsters and traps just right to fend off a series of different heroes? Or was it an open-ended sandbox strategy game, a SimCity — or Theme Park — in a dungeon? Or was it, as Strategy Plus dared to write, a game “which may redefine the role-playing genre”? At one point or another, each of these identities was the paramount one, only to fall back down in the pecking order once again. Programmer Jonty Barnes, who was responsible for the Dungeon Keeper level editor, left Bullfrog in the summer of 1995 to finish up his degree at university. He claims that when he rejoined the team a year later the game had “reset. It was back to where we were when I’d left off.”

Be that as it may, the inflection point for the project had already come and gone by then. EA was extremely disappointed when Dungeon Keeper failed to ship in time for the Christmas of 1995, to supplement a Bullfrog lineup for the year that otherwise included only the under-performing Magic Carpet 2 and the thoroughly underwhelming racing game Hi-Octane — a poor early return indeed on a $45 million investment. Management started to make noises about cancelling Dungeon Keeper entirely if it didn’t show signs of coming together soon. Meanwhile Peter Molyneux was as miserable as ever: “I still really love games, and I hate what I’m doing at the moment.” He thought about resigning, then trying to find or found another company where he could make games like he had in the old days, but he couldn’t bear to lose all influence over Dungeon Keeper. And as for seeing his baby axed… well, that might just send him around the bend completely.

So, at the beginning of 1996, he went to the rung above him on the EA corporate ladder with a modest proposal. He’d had enough of meet-and-greets and wining-and-dining, he told his bosses forthrightly. He wanted to quit EA — but first he wanted to finish Dungeon Keeper. He wanted to climb back down into the trenches and make one last amazing game before he said farewell. EA scratched its collective head, then said okay — on two additional conditions. He couldn’t make his game from the Bullfrog offices, where he would be privy to inside information about the other projects going on and might, it was feared, try to poach Bullfrog employees for whatever he decided to do after Dungeon Keeper. He would have to find another spot for the team to work from. And naturally, he couldn’t tell the public that he was leaving the studio with which he was so closely identified until Dungeon Keeper was done. This sounded reasonable enough to Peter Molyneux.

He didn’t have to look far for a new workspace: he simply moved the core of the team — consisting of no more than half a dozen people — into a space above the garage at his house.1 Suddenly it was like the old days again, just a small group of friends working and playing together for a ridiculous number of hours each day to make The Best Game Ever. The team effectively became Molyneux’s roommates in addition to his colleagues. Jonty Barnes:

We would play the game at night drinking beers, arguing over who should have won based on how the game should be versus the way it was. If somebody won too easily, we could change the design the following day, and that really honed the intensity and the strategy of building Dungeon Keeper. We ended up working pretty much six or seven days a week for a long period of time, but we didn’t notice because we cared so much about what we were creating.

Over these endless hours of work and play, the what of the game slowly crystallized. Certain elements of the puzzler were retained: each scenario of the campaign would require you to fend off attacks by one or more avatars of Good. Ditto some elements of the CRPG, whose legacy would take the form of the ability to “possess” any of the monsters in your dungeon, running around and fighting from a first-person view. At heart, however, Dungeon Keeper would be a management strategy game, all about setting up and running the best dungeon you could with the resources to hand. It would borrow freely not only from the likes of Theme Park (and now Theme Hospital) but also from the more straightforwardly confrontational real-time-strategy genre that was exploding in popularity while it was in development.

Despite the jolt of energy that accompanied the move into Molyneux’s house — or perhaps because of it — work on the game dragged on for a long, long time thereafter. The Christmas of 1996 as well was missed, while EA’s bosses continued to shake their heads and make threatening noises. At last, just a few weeks after Theme Hospital had shipped, Molyneux and company declared Dungeon Keeper to be finished. True to his word, Bullfrog’s co-founder, leading light, and heart and soul officially resigned on that very day. “Dungeon Keeper is my final game for Bullfrog, and this is part of the reason I wanted to make it so good,” he told a shocked British press. “It’s a sort of goodbye and thanks for all the great times past.”


You start each scenario with only a Dungeon Heart. You must protect this nerve center and in particular the orb it contains at all costs, for if the orb is destroyed, you die instantly.

Keeping a dungeon turns out to be a bit like keeping a zoo. You need a Hatchery to provide food for your minions, in the form of hordes of woebegone chickens.

Each type of monster also needs its own Lair to retreat to. (It turns out that a lot of these creatures don’t like each other very much.)

Neither a zoo nor a dungeon can run for long without a steady supply of filthy lucre.

The second scenario introduces the Training Room, where you can “level up” your monsters in exchange for gold.

The third scenario introduces the Library, where you can research spells with which to smite your enemies (or your minions). The campaign continues in this vein, layering on the complexity piece by piece.



When he looks back upon his career, Peter Molyneux tends to be a harsh critic of his own games. Dungeon Keeper is no exception: “It had a lot of things wrong with it. Too many icons. Too many game mechanics that were just wrong. It ended up being too dependent on little bits that were supposed to be there as jokes. In the end, I didn’t feel as proud about it as I had hoped.”

Very few people have agreed with this assessment. In its heyday, Dungeon Keeper enjoyed the best reviews of any Bullfrog game in years, outdoing even Theme Hospital in that respect. (Combined, the two titles created a real if fairly brief-lived “Bullfrog is back!” frisson.) Computer Gaming World called Dungeon Keeper “a damned fine creation. Its utter uniqueness and sense of style alone are worth the price, especially in these days of recycled inspiration. It’s a true gamer’s game — tremendously deep, demanding, and open to exploration.” GameSpot deemed Dungeon Keeper “among the best games released so far this year, and any fan of real-time strategy or classic fantasy role-playing games should run right out and buy this game. Hopefully, this is the new face of Bullfrog…”

Sales were solid, although not quite as strong as the reviews — not to mention EA’s early expectations — might have led one to believe. The game’s Achilles heel, to whatever extent it had one, was a dated presentation, a byproduct of the long development cycle and the small development team. It was among the last slate of games to be targeted at MS-DOS rather than Windows, and its muddy, pixelated graphics, caked with enough brown sauce to recreate the Mona Lisa, looked to be of even lower resolution than they actually were. At a glance, Dungeon Keeper looked old and a bit dull, in a milieu obsessed with the new and shiny.  (Theme Hospital was another MS-DOS anachronism, but its brighter color palette and cleaner aesthetics left it looking a lot more welcoming.)

But if you can get beyond the visuals, Dungeon Keeper goes a long way toward living up to the brilliance of its central conceit. If the first half of its development cycle was one extended identity crisis, the creators more than figured out what identity their game should have by the time that all was said and done. Its every element, from the silkily malevolent voiceovers by the actor Richard Riding, who sounds like all of Shakespeare’s villains rolled into one, to the countless little sadistic touches, like the way you can slap your minions with your big old claw-like hand to get them working faster, serves to justify the marketing tagline of “Evil is Good.” Even if you’re someone like me who genuinely prefers to be the good guy in games, you can’t help but admire Dungeon Keeper’s absolute commitment to its premise. And for those who do like the idea of playing the dark side more than I do, this game is the perfect antidote to all of those CRPGs that claim to let you do so, only to force you to do the right thing in the end for the — horror of horrors! — wrong reason. (A whole Parisian salon worth of existentialist philosophers would like to have a word with those game designers about the real nature of Good and Evil…) There’s no such prevarication in Dungeon Keeper; your intentions and your deeds are in complete harmony.

For my part, I enjoyed the game more than I really expected to — in fact, more than any other Bullfrog game I’ve played in the course of writing these histories. It helped immensely that Dungeon Keeper marks the point where Bullfrog finally gave us a proper, bespoke campaign, complete with a modest but effective spine of story, which sees you corrupting a succession of goodie-goodie realms with sticky-sweet names like “Eversmile.” The first half in particular of the 20-scenario campaign is one of the best of this era of gaming. Unlike a lot of earlier Bullfrog games, which were content to throw you in at the deep end, Dungeon Keeper starts off with just a handful of elements, then gradually layers on the complexity scenario by scenario, introducing new room types, new monsters, and new spells in a way that feels fun rather than overwhelming. The game plays quickly enough, and the scenarios are varied enough, that having to start over and build a new dungeon from scratch each time feels tempting rather than punishing. You’re always thinking about how you can build the next one better, thinking about how you can integrate the new elements that are introduced into what you’ve already learned how to do. It’s enough to win over even a middle-aged skeptic like me, whose heart isn’t set all aflutter by the transgression of being Evil.

Unfortunately, the happy vibes came to a crashing halt when I hit the tenth scenario of the campaign. Here the game breaks with everything the previous nine scenarios have been painstakingly teaching you; suddenly it’s not about constructing an efficient and deadly dungeon, but about bum-rushing the domain of another dungeon keeper before he can do the same to you. This aberration might be acceptable, if the game clearly communicated to you that a dramatic change in tactics is needed. But instead of doing so, the introductory spiel actively misleads you into believing that this scenario should be approached the same way as the nine preceding ones. Even worse, if you fail to attack and kill the enemy quickly, he will soon make himself effectively invulnerable, leaving you stuck in a walking-dead situation; you can easily waste an hour or more before you realize it. Coming on the heels of so much good design, the scale of the design failure here is so enormous that I want to believe it’s some sort of technical mistake — surely nobody could have intended for the scenario to be like this, right? — but it’s hard for me to see how that could be the case either.

I did eventually figure out what the scenario wanted from me, but I was left feeling much the same as when I run into a really bad, unfair puzzle in an adventure game. At that point, the game has lost my trust, and it can never be regained. I played through a couple more scenarios — which, to be fair, returned to the old style that I had been enjoying — but somehow the magic was gone. I just wasn’t having fun anymore. So, I quit. Let this be a lesson to you, game designers: just like trust between two people in the real world, trust between a game and its player is slow and hard to build and terrifyingly quick and easy to destroy. For any of you readers who choose to play Dungeon Keeper — and there are still lots of good reasons to do so — I’d recommend that you go in forewarned and forearmed with a walkthrough or a cheat for the problem-child scenario.2


The Dungeon Keeper story didn’t end with the original game. Late in 1997, Bullfrog and EA released a rather lazy expansion pack called The Deeper Dungeons, a collection of leftover scenarios that hadn’t made the cut the first time around, with no new campaign to connect them. Far more impressive is Dungeon Keeper 2, which arrived in the middle of 1999.

The team that made Dungeon Keeper 2 was mostly new, although a few old hands, among them programmer Jonty Barnes, did return to the fold. The lead designer this time was Sean Cooper, the father of Bullfrog’s earlier Syndicate. He remembers the project as “a shift from writing games without many plans to getting really organized about what we were going to build. It was about keeping everyone in full view of what we were all trying to do together. That’s common practice today, but it was one of the first times for us.”

Less an evolution of the first game’s systems than a straight-up remake with modernized graphics and interface, the sequel’s overall presentation is so improved that this doesn’t bother me a bit. What was murky and pixelated has become crisp and clear, without losing any of its malevolent style; Richard Riding has even returned to louche it up again as the velvet-tongued master of ceremonies. I only played about five scenarios into this 22-scenario campaign due to time pressures — being a digital antiquarian with a syllabus full of games to get through sometimes forces me to make hard choices about where I spend my supposed leisure time — but what I saw was an even better version of the first campaign. Plus, I’m given to understand that there are no aberrations here like the tenth scenario of that campaign. Some purists will tell you that the sequel lacks the first game’s “soul.” Personally, though, I’ll take the second game’s refinement over any such nebulous quality. If I was coming to Dungeon Keeper cold today, this is definitely where I would start.

But, whether due to a lack of soul or just because it was too much too soon of the same idiosyncratic concept, Dungeon Keeper 2 didn’t sell as well as its predecessor, and so the series ended here. Sadly, this was a running theme of Bullfrog’s post-Molyneux period, during which the games became more polished and conventionally “professional” but markedly less inventive, and didn’t always do that well in the marketplace either. “I felt the heart of the place was missing,” says Mark Healey. “Bullfrog was no longer a creative haven for me. It felt more like a chicken factory.” And the chickens kept laying eggs. Populous: The Beginning, the big game for 1998, disappointed commercially, as did Dungeon Keeper 2 the following year. The same year’s Theme Park World3 did better, but had much of its thunder stolen by Rollercoaster Tycoon, an era-defining juggernaut which had been partially inspired by Bullfrog’s first Theme Park.

Early in 2000, EA closed Bullfrog’s Guildford offices, moving most of the staff into the mother ship’s sprawling new office complex in the London suburb of Chertsey, which lay only ten miles away. The Bullfrog brand was gradually phased out after that. Old games studios never die; they just fade away.

Setting Dungeon Keeper aside as a partial exception, I’ve made no secret of the fact that the actual artifacts of Bullfrog’s games have tended to strike me less positively than their modern reputation might suggest they should, that I consider the studio to have been better at blue-sky innovation than execution on the screen. Yet it cannot be denied that Bullfrog laid down a string of bold new templates that are still being followed today. So, credit where it’s due. Bullfrog’s place in gaming history is secure. It seems only fitting that Peter Molyneux have the last word on the studio that he co-founded and defined, whose ethic has to a large extent continued to guide him through his restlessly ambitious, controversy-fraught post-millennial career in games.

It [was] an amazing, joyful roller coaster which I wouldn’t have traded for anything. It was an obsession. We were just obsessed with doing stuff that other people hadn’t done before. It was working very late and very hard and smoking lots of cigarettes and eating lots of pizza and bringing together some crazy insane people. We just had the reason to do stuff. There wasn’t a lot of process involved. We didn’t have producers. There were some brutally tough times there, but it was amazing.



Did you enjoy this article? If so, please think about pitching in to help me make many more like it. You can pledge any amount you like.


Sources: The book Bullfrog’s Official Guide to Dungeon Keeper by Melissa Tyler and Shin Kanaoya; PC Zone of May 1996 and May 2000; Retro Gamer 43, 71, 104, 110, 113, 130, 143, and 268; Computer Gaming World of December 1995, August 1997, October 1997, and October 1999; Computer Games Strategy Plus of November 1995.

Online sources include Peter Molyneux’s brief-lived personal blog, “Peter Molyneux: A fallen god of game design seeking one more chance” by Tom Phillips at EuroGamerGameSpot’s vintage review of Dungeon Keeper, and “Legends of Game Design: Peter Molyneux” by Rob Dulin at the old GameSpot.

Where to Get Them: Theme HospitalDungeon Keeper Gold, and Dungeon Keeper 2 are all available as digital purchases at GOG.com.

 
 

Tags: , , , ,

The Rise of POMG, Part 4: A World for the Taking

Just as the Ultima Online beta test was beginning, Electronic Arts was initiating the final phase of its slow-motion takeover of Origin Systems. In June of 1997, the mother ship in California sent down two Vice Presidents to take over completely in Texas, integrate Origin well and truly into the EA machine, and end once and for all any semblance of independence for the studio. Neil Young became Origin’s new General Manager on behalf of EA, while Chris Yates became Chief Technical Officer. Both men were industry veterans.

Appropriately enough given that he was about to become the last word on virtual Britannia, Neil Young was himself British. He attributes his career choice to the infamously awful English weather. “There are a lot of people in the games industry that come from the UK,” he says. “I think it’s because the weather is so bad that you don’t have a lot to do, so you either go into a band or teach yourself to program.” He chose the latter course at a time when computer games in Britain were still being sold on cassette tape for a couple of quid. After deciding to forgo university in favor of a programming job at a tiny studio called Imagitec Design in 1988, he “quickly realized there were more gifted engineers,” as he puts it, and “moved into producing.” Having made a name for himself in that role, he was lured to the United States by Virgin Interactive in 1992, then moved on to EA five years later, which organization had hand-picked him for the task of whipping its sometimes wayward and lackadaisical stepchild Origin into fighting shape.

Chris Yates had grown up amidst the opposite of English rain, hailing as he did from the desert gambler’s paradise Las Vegas. He was hired by the hometown studio Westwood Associates in 1988, where he worked as a programmer on games like Eye of the Beholder, Dune II, and Lands of Lore. In 1994, two years after Virgin acquired Westwood, he moved to Los Angeles to join the parent company. There he and Young became close friends as well as colleagues, such that they chose to go to EA together as a unit.

The two were so attractive to EA thanks not least to an unusual project which had occupied some of their time during their last year and a half or so at Virgin. Inspired by Air Warrior, the pioneering massively-multiplayer online flight simulator that had been running on the GEnie commercial online service since the late 1980s, a Virgin programmer named Rod Humble proposed in 1995 that his company invest in something similar, but also a bit simpler and more accessible: a massively-multiplayer version of Asteroids, the 1979 arcade classic whose roots stretched all the way back to Spacewar!, that urtext of videogaming. Neil Young and his friend Chris Yates went to bat for the project: Young making the business case for it as an important experiment that could lead to big windfalls later on, Yates pitching in to offer his exceptional technical expertise whenever necessary. Humble and a colleague named Jeff Paterson completed an alpha version of the game they called SubSpace in time to put it up on the Internet for an invitation-only testing round in December of 1995. Three months later, the server was opened to anyone who cared to download the client — still officially described as a beta version — and have at it.

SubSpace was obviously a very different proposition from the likes of Ultima Online, but it fits in perfectly with this series’s broader interest in persistent online multiplayer gaming (or POMG as I’ve perhaps not so helpfully shortened it). For, make no mistake, the quality of persistence was as key to its appeal as it was to that of such earlier featured players in this series as Kali or Battle.net. SubSpace spawned squads and leagues and zones; it became an entire subculture unto itself, one that lived in and around the actual battles in space. The distinction between it and the games of Kali and Battle.net was that SubSpace was massively — or at least bigly — multiplayer. Whereas an online Diablo session was limited to four participants, SubSpace supported battles involving up to 250 players, sometimes indulging in crazy free-for-alls, more often sorted into two or more teams, each of them flying and fighting in close coordination. It thus quickly transcended Asteroids in its tactical dimensions as well as its social aspects — transcended even other deceptively complex games with the same roots, such as Toys for Bobs’s cult classic Star Control. That it was playable at all over dial-up modem connections was remarkable; that it was so much fun to play and then to hang out in afterward, talking shop and taking stock, struck many of the thousands of players who stumbled across it as miraculous; that it was completely free for a good long time was the icing on the cake.

It remained that way because Virgin didn’t really know what else to do with it. When the few months that had been allocated to the beta test were about to run out, the fans raised such a hue and cry that Virgin gave in and left it up. And so the alleged beta test continued for more than a year, the happy beneficiary of corporate indecision. In one of his last acts before leaving Virgin, Neil Young managed to broker a sponsorship deal with Pepsi Cola, which gave SubSpace some actual advertising and another lease on life as a free-to-play game. During that memorable summer of the Ultima Online beta test, SubSpace was enjoying what one fan history calls its “greatest days” of all: “The population tripled in three months, and now there were easily 1500-plus people playing during peak times.”

With the Pepsi deal about to run out, Virgin finally took SubSpace fully commercial in October of 1997, again just as Ultima Online was doing the same. Alas, it didn’t go so well for SubSpace. Virgin released it as a boxed retail game, with the promise that, once customers had plunked down the cash to buy it, access would be free in perpetuity. This didn’t prevent half or more of the existing user base from leaving the community, even as nowhere near enough new players joined to replace them. Virgin shut down the server in November of 1998; “in perpetuity” had turned out to be a much shorter span of time than anyone had anticipated.

As we’ve seen before in this series, however, the remaining hardcore SubSpace fans simply refused to let their community die. They put up their own servers — Virgin had made the mistake of putting all the code you needed to do so on the same disc as the client — and kept right on space-warring. You can still play SubSpace today, just as you can Meridian 59 and The Realm. A website dedicated to tracking the game’s “population statistics” estimated in 2015 that the community still had between 2000 and 3000 active members, of whom around 300 might be online at any given time; assuming these numbers are to be trusted, a bit of math reveals that those who like the game must really like it, spending 10 percent or more of their lives in it. That same year, fans put their latest version of the game, now known as Subspace Continuum, onto Steam for free. Meanwhile its original father Rod Humble has gone on to a long and fruitful career in POMG, working on Everquest, The Sims Online, and Second Life among other projects.



But we should return now to the summer of 1997 and to Origin Systems, to which Neil Young and Chris Yates came as some of the few people in existence who could boast not only of ideas about POMG but of genuine commercial experience in the field, thanks to SubSpace. EA hoped this experience would serve them well when it came to Ultima Online.

Which isn’t to say that the latter was the only thing they had on their plates: the sheer diversity of Young’s portfolio as an EA general manager reflects the confusion about what Origin’s identity as a studio should be going forward. There were of course the two perennials, Ultima — meaning for the moment at least Ultima Online — and Wing Commander, which was, as Young says today, “a little lost as a product.” Wing Commander, the franchise in computer gaming during the years immediately prior to DOOM, was becoming a monstrous anachronism by 1997. Shortly after the arrival of Young and Yates, Origin would release Wing Commander: Prophecy, whose lack of the Roman numeral “V” that one expected to see in its name reflected a desire for a fresh start on a more sustainable model in this post-Chris Roberts era, with a more modest budget to go along with more modest cinematic ambitions. But instead of heralding the dawn of a new era, it would prove the franchise’s swan song; it and its 1998 expansion pack would be the last new Wing Commander computer games ever. Their intended follow-up, a third game in the Wing Commander: Privateer spinoff series of more free-form outer-space adventures, would be cancelled.

In addition to Ultima and Wing Commander, EA had chosen to bring under the Origin umbrella two product lines that were nothing like the games for which the studio had always been known. One was a line of military simulations that bore the imprimatur of “Jane’s,” a print publisher which had been the source since the turn of the twentieth century of the definitive encyclopedias of military hardware of all types. The Jane’s simulations were overseen by one Andy Hollis, who had begun making games of this type for MicroProse back in the early 1980s. The other line involved another MicroProse alum — in fact, none other than Sid Meier, whose name had entered the lexicon of many a gaming household by serving as the prefix before such titles as Pirates!, Railroad Tycoon, Civilization, and Colonization. Meier and two other MicroProse veterans had just set up a studio of their own, known as Firaxis Games, with a substantial investment from EA, who planned to release their products under the Origin Systems label. Origin was becoming, in other words, EA’s home for all of its games that were made first and usually exclusively for computers rather than for the consoles that now provided the large majority of EA’s revenues; the studio had, it seemed, more value in the eyes of the EA executive suite as a brand than as a working collective.

Still, this final stage of the transition from independent subsidiary to branch office certainly could have been even more painful than it was. Neil Young and Chris Yates were fully aware of how their arrival would be seen down in Austin, and did everything they could to be good sports and fit into the office culture. Brit-in-Texas Young was the first to come with the fish-out-of-water jokes at his own expense — “I was expecting a flat terrain with lots of cowboys, cacti, and horses, so I was pleasantly surprised,” he said of Austin — and both men rolled up their sleeves alongside Richard Garriott to serve the rest of the company a turkey dinner at Thanksgiving, a longtime Origin tradition.

Neil Young and Chris Yates on the Thanksgiving chow line.

Young and Yates had received instructions from above that Ultima Online absolutely had to ship by the end of September. Rather than cracking the whip, they tried to cajole and josh their way to that milestone as much as possible. They agreed to attend the release party in drag if the deadline was met; then Young went one step farther, promising Starr Long a kiss on the lips. Yates didn’t go that far, but he did agree to grow a beard to commemorate the occasion, even as Richard Garriott, whose upper lip hadn’t seen the sun since he’d graduated from high school, agreed to shave his.

Young and Yates got it done, earning for themselves the status of, if not the unsung heroes of Ultima Online, two among a larger group of same. The core group of ex-MUDders whose dream and love Ultima Online had always been could probably have kept running beta tests for years to come, had not these outsiders stepped in to set the technical agenda. “That meant trading off features with technology choices and decisions every minute of the day,” says Young. He brought in one Rich Vogel, who had set up and run the server infrastructure for Meridian 59 at The 3DO Company, to do the same for Ultima Online. In transforming Origin Systems into a maintainer of servers and a seller of subscriptions, he foreshadowed a transition that would eventually come to the games industry in general, from games as boxed products to gaming as a service. These tasks did not involve the sexy, philosophically stimulating ideas about virtual worlds and societies with which Raph Koster and his closest colleagues spent their time and which will always capture the lion’s share of the attention in articles like this one, but the work was no less essential for all that, and no less of a paradigm shift in its way.

So, the big day came and the deadline was met: Ultima Online shipped on September 24, 1997, three days before Meridian 59 would celebrate its first anniversary. The sleek black box was an end and a beginning at the same time. Young and Yates did their drag show, Starr Long got his kiss, and, most shockingly of all, Richard Garriott revealed his naked upper lip to all and sundry. (Opinions were divided as to whether the mangy stubble which Chris Yates deigned to grow before picking up his razor again really qualified as a beard or not.) And then everyone waited to see what would happen next.

A (semi-)bearded Chris Yates and a rare sight indeed: a clean-shaven Richard Garriott.

EA made and shipped to stores all over the country 50,000 copies of Ultima Online, accompanying it with a marketing campaign that was, as Wired magazine described it, of “Hollywood proportions.” The virtual world garnered attention everywhere, from CNN to The New York Times. These mainstream organs covered it breathlessly as the latest harbinger of humanity’s inevitable cyber-future, simultaneously bracing and unnerving. Flailing about for a way to convey some sense of the virtual world’s scope, The New York Times noted that it would take 38,000 computer monitors — enough to fill a football field — to display it in its entirety at one time. Needless to say, the William Gibson quotes, all “collective hallucinations” and the like, flew thick and fast, as they always did to mark events like this one.

Three weeks after the launch, 38,000 copies of Ultima Online had been sold and EA was spooling up the production line again to make another 65,000. Sales would hit the 100,000 mark within three months of the release. Such numbers were more than gratifying. EA knew that 100,000 copies sold of this game ought to be worth far more to its bottom line than 100,000 copies of any other game would have been, given that each retail sale hopefully represented only the down payment on a long-running subscription at $10 per month. For its publisher, Ultima Online would be the gift that kept on giving.

In another sense, however, the sales figures were a problem. When Ultima Online went officially live, it did so on just three shards: the Atlantic and Pacific shards from the beta test, plus a new Great Lakes one to handle the middle of the country. Origin was left scrambling to open more to meet the deluge of subscribers. Lake Superior came up on October 3, Baja on October 10, Chesapeake on October 16,  Napa Valley on November 14, Sonoma on December 13, Catskills on December 22. And still it wasn’t enough.

Origin’s estimates of how many players a single server could reliably support proved predictably overoptimistic. But rather than dial back on the number of players they allowed inside, thereby ensuring that each of them who did get in could have a reasonably enjoyable experience, they kept trying to cover the gap between technical theory and reality by hacking their code on the fly. As a result, Ultima Online became simultaneously the most loved and most hated game in the country. When it all came together, it was magic for many of its players. But truth be told, that didn’t happen anywhere near as often as one might have wished in that first year or so. Extreme lag, inexplicable glitches, dropped connections, and even total server crashes were the more typical order of the day. Of course, with almost everyone who surfed the Web still relying on dial-up modems running over wires that had been designed to carry voices rather than computer data, slowdowns and dropped connections were a reality of daily online life even for those who weren’t attempting to log onto virtual worlds. This created a veneer of plausible deniability, which Origin’s tech-support people, for lack of any other suggestions or excuses to offer, leaned on perhaps a bit too heavily. After all, who could say for sure that the problem any individual player might be having wasn’t downstream from Origin’s poor overtaxed server?

Weaselly excuses like these led to the first great act of civil disobedience by the residents of Britannia, just a few weeks after the launch, when hundreds of players gathered outside Lord British’s castle, stripped themselves naked, broke into the throne room, drank gallons of wine, and proceeded to disgorge all of it onto Richard Garriott’s virtual furniture, whilst chanting in unison their demands for a better, stabler virtual world. The world’s makers were appalled, but also weirdly gratified. What better sign of a budding civic life could there be than a full-on political protest? “We were all watching and thinking it was a grand statement about the project,” says Richard Garriott. “As unhappy as they were about the game, they voiced their unhappiness in the context of the game.” Much of what happened inside Ultima Online during the first year especially had the same quality of being amazing for philosophers of virtual worlds to witness, but stressful for the practical administrators who were trying to turn this one into a sustainable money tree. The rub was that the two categories were combined in the very same people, who were left feeling conflicted to say the least.

The journals of hardcore gaming, hardly known for their stoicism in the face of hype on most days, were ironically more reserved and skeptical than the mainstream press on the subject of Ultima Online, perchance because they were viewing the virtual world less as a harbinger of some collective cyber-future and more as a game that their readers might wish to, you know, actually play. Computer Gaming World wittily titled its scathing review, buried on page 162 and completely unmentioned on the cover of the issue in question, simply “Uh-Oh.” Among the litany of complaints were “numerous and never-ending bugs, horrible lag time, design issues [that] lead to repetitive and time-consuming activities, and [an] unbalanced economy.” The magazine did admit that “Ultima Online could become a truly great game. But we can’t review potential, we can only review concrete product.” Editor-in-chief Johnny L. Wilson, for his part, held out little hope for improvement. “Ultima Online begins with hubris and ends in Greek tragedy,” he said. “The hubris is a result of being unwilling to learn from others’ mistakes. The tragedy is that it could have been so much more.” Randy Farmer, co-creator of the earlier would-be virtual world Habitat, expressed a similar sentiment, saying that “Origin seems to have ignored many of the lessons that our industry has learned in the last ten years of building online worlds. They’re making the same mistakes that first-time virtual-world builders always make.”

The constant crashes and long periods of unexplained down time associated with a service for which people were paying good money constituted a corporate lawyer’s worst nightmare — or a different sort of lawyer’s wet dream. One of these latter named George Schultz began collecting signatures from Origin’s most disgruntled customers within weeks, filing a class-action lawsuit in San Diego at the beginning of March of 1998. Exhibit A was the copy right there on the back of the box, promising “a living, growing world where thousands of real people discover real fantasy and adventure, 24 hours a day, every day of the year,” with all of it taking place “in real time.” This was, claimed Schultz, a blatant case of false advertising. “We’re not trying to tell anyone how to design a good or a bad game,” he said. “What it’s about is holding Origin and EA to the promises they made on the box, in their advertising, and [in] the manual. It’s about the misrepresentations they’ve made. A big problem with the gaming industry is that they think there are some special rules that only apply to them.”

Whatever the truth of that last claim, there was no denying that just about half of the learning curve of Ultima Online was learning to navigate around the countless bugs and technical quirks. For example, Origin took down each shard once per day for a backup and a “therapeutic” reboot that was itself a testament to just what a shaky edifice the software and hardware were. When the server came back up again, it restored the state of the world from the last backup. But said state was a snapshot in time from one hour before the server went down. There was, in other words, an hour every day during which everything you did in virtual Britannia was doomed to be lost; this was obviously not a time to go on any epic, treasure- and experience-point-rich adventures. Yet such things were documented nowhere; one learned them only through the proverbial school of hard knocks.

In their defense, Origin was sailing into completely uncharted waters with Ultima Online. Although there had been online virtual worlds before, dating all the way back to that first MUD of 1978 or 1979, none of them — no, not even Meridian 59 and The Realm — had been as expansive, sophisticated, and most of all popular as these shards of Britannia. Most of the hardware technologies that would give rise to the era of Web 2.0, from DSL in homes to VPS’s in data centers, existed only as blueprints; ditto most of the software. No one had ever made a computer game before that required this much care and feeding after the initial sale. And it wasn’t as if the group entrusted with maintaining the beast was a large one. Almost the entirety of the Ultima IX team which had been parachuted in six months before the launch to just get the world done already was pulled out just as abruptly as soon as it started accepting paying subscribers, leaving behind a crew of maintainers that was little bigger than the original team of ex-MUDders who had labored in obscurity for so long before catching the eye of EA’s management. The idea that maintaining a virtual world might require almost as much manpower and ongoing creative effort as making it in the first place was too high a mental hurdle for even otherwise clever folks like Neil Young and Chris Yates to clear at this point.

Overwhelmed as they were, the maintainers began to rely heavily on unpaid volunteers from the community of players to do much of the day-to-day work of administrating the world, just as was the practice on MUDs. But Ultima Online ran on a vastly larger scale than even the most elaborate MUDs, making it hard to keep tabs on these volunteer overseers. While some were godsends, putting in hours of labor every week to make Britannia a better place for their fellow players, others were corrupted by their powers, manipulating the levers they had to hand to benefit their friends and punish their enemies. Then, too, the volunteer system was another legal quagmire, one that would doubtless have sent EA’s lawyers running screaming from the room if anyone had bothered to ask them about it before it was rolled out; sure enough, it would eventually lead to another lawsuit, this one more extended, serious, and damaging than the first.

In the meanwhile, though, most players did not rally behind the first lawsuit to anything like the degree that George Schultz might have been hoping. The fact was that even the ones who had vomited all over Lord British’s throne had done so because they loved their virtual Britannia and wanted to see it fixed rather than destroyed, as it would likely be if Schultz won the day. The suit concluded in a settlement at the end of 1998. The biggest concession on the part of the defendants was a rather weird one that gave no recompense to any individual inhabitant of virtual Britannia: EA agreed to donate $15,000 to the San Jose Tech Museum of Innovation. Perhaps Schultz thought that it would be able to innovate up a more reliable virtual world.

While many of the technical problems that beset Ultima Online were only to be expected in the context of the times, some of the other obstacles to enjoying the virtual world were more puzzling. First and foremost among these was the ever-present issue of players killing other players, which created so much frustration that George Schultz felt compelled to explicitly wall it off from the breach-of-trust claims that were the basis of his lawsuit: “We’re not getting into whether there should be player-killing.” Given that it had been such a constant theme of life (and death) in virtual Britannia going all the way back to the alpha-testing phase, the MUDders might have taken more steps to address it before the launch. As it was, though, one senses that, having seen so many of their ideas about a virtual ecology and the like not survive contact with real players, having been forced to give up in so many ways on virtual Britannia as a truly self-sustaining, living world, they were determined to make this the scene of their last stand, the hill that they would either hold onto or die trying.

Their great white hope was still the one that Richard Garriott had been voicing in interviews since well before the world’s commercial debut: that purely social pressures would act as a constraint on player-killing — that, in short, their world would learn to police itself. In fact, the presence of player-killing might act as a spur to civilization — for, as Raph Koster said, “cultures define and refine themselves through conflict.” They kept trying to implement systems that would nudge this particular culture in the right direction. They decided that, after committing murder five times, a player would be branded with literal scarlet letters: the color of his onscreen name would change from blue to red. Hopefully this would make him a pariah among his peers, while also making it very dangerous for him to enter a town, whose invulnerable computer-controlled guards would attack him on sight. The designers didn’t reckon with the fact that a virtual life is, no matter how much they might wish otherwise, simply not the same as a real life. Some percentage of players, presumably perfectly mild-mannered and law-abiding in the real world, reveled in the role of murderous outlaws online, taking the red letters of their name as a badge of honor rather than shame, the dangers of the cities as a challenge rather than a deterrent. To sneak past the city gates, creep up behind an unsuspecting newbie and stab her in the back, then get out of Dodge before the city watch appeared… now, that was good times. The most-wanted rolls posted outside the guard stations of Britannia became, says Raph Koster, “a high-score table for player killers.”

The MUDders’ stubborn inflexibility on this issue — an issue that was by all indications soon costing Ultima Online large numbers of customers — was made all the more inexplicable in the opinion of many players by the fact that it was, in marked contrast to so many of the other problems, almost trivial to address in programming terms. An “invulnerability” flag had long existed, to be applied not only to computer-controlled city guards but to special human-controlled personages such as Lord British to whom the normal laws of virtual time and space did not apply. All Origin had to do was add a few lines of code to automatically turn the flag on when a player walked into designated “safe” spaces. That way, you could have places where those who had signed up mostly in order to socialize could hang out without having to constantly look over their backs, along with other places where the hardcore pugilists could pummel one another to their heart’s content. Everyone would be catered to. Problem solved.

But Raph Koster and company refused to take this blindingly obvious step, having gotten it into their heads that to do so would be to betray their most cherished ideals. They kept tinkering around the edges of the problem, looking for a subtler solution that would preserve their world’s simulational autonomy. For example, they implemented a sort of karmic justice system, which dictated that players who had been evil during life would be resurrected after death only after losing a portion of their stats and skills. Inevitably, the player killers just took this as another challenge. Just don’t get killed, and you would never have to worry about it.

The end result was to leave the experience of tens of thousands of players in the unworthy hands of a relatively small minority of “griefers,” people who thrived on causing others pain and distress. Like all bullies, they preyed on the weak; their typical victims were the newbies, unschooled in the ways of defense, guiding characters with underwhelming statistics and no arms or armor to speak of. Such new players were, of course, the ones whose level of engagement with the game was most tentative, who were the mostly likely to just throw up their hands and go find something else to play after they’d been victimized once or twice, depriving Origin of potentially hundreds of dollars in future subscription revenue.

In light of this, it’s strange that no one from EA or Origin overrode the MUDders on this point. For his part, Richard Garriott was adamantly on their side, insisting that Ultima Online simply had to allow player-killing if it wasn’t to become a mockery of itself. It was up to the dissatisfied and victimized residents themselves to band together and turn Britannia into the type of world they wanted to live in; it wasn’t up to Origin to step in and fix their problems for them with a deus ex machina. “When we first launched Ultima Online, we set out to create a world that supported the evil player as a legitimate role,” said Garriott in his rather high-handed way. “Those who have truly learned the lessons of the [single-player] Ultima games should cease their complaining, rise to the challenge, and make Britannia into the place they want it to be.” He liked to tell a story on this subject. (Knowing Garriott’s penchant for embellishment, it probably didn’t happen, or at least didn’t happen quite like this. But that’s not relevant to its importance as allegory.)

One evening, he was wandering the streets of the capital in his Lord British persona, when he heard a woman screaming. Rushing over to her, he was told that a thief had stolen all of her possessions. His spirit of chivalry was awoken; he told her that he would get her things back for her. Together they tracked down the thief and cornered him in a back alley. Lord British demanded that the thief return the stolen goods, and the thief complied. They all went their separate ways. A moment later, the woman cried out again; the thief had done it again.

This time, Lord British froze the thief with a spell before he could leave the scene of the crime. “I told you not to do that,” he scolded. “What are you doing?”

“Sorry, I won’t do it again,” said the thief as he turned over the goods for a second time.

“If you do that again, I’m going to ban you from the game,” said Lord British.

You might be able to guess what happened next: the thief did it yet again. “I said I was going to ban you, and now I have to,” shouted Lord British, now well and truly incensed. “What’s wrong with you? I told you not to steal from this woman!”

The thief’s answer stopped Garriott in his tracks. “Listen. You created this world, and I’m a thief,” he said, breaking character for the first time. “I steal. That’s what I do. And now you’re going to ban me from the game for playing the role I’m supposed to play? I lied to you before because I’m a thief. The king caught me and told me not to steal. What am I going to do, tell you that as soon as you turn around I’m going to steal again? No! I’m going to lie.”

And Garriott realized that the thief was right. Garriott could do whatever he wished to him as Lord British, the reigning monarch of this world. But if he wished to stay true to all the things he had said in the past about what virtual Britannia was and ought to be, he couldn’t go outside the world to punish him as Richard Garriott, the god of the server looking down from on-high.

Some of the questions with which Origin was wrestling resonate all too well today: questions involving the appropriate limits of online free speech — or rather free action, in this case. They are questions with which everyone who has ever opened an Internet discussion up to the public, myself included, have had to engage. When does strongly felt disagreement spill over into bad faith, counterpoint into disruption for the sake of it? And what should we do about it when it does? In Origin’s case, the pivotal philosophical question at hand was where the boundary lay between playing an evil character in good faith in a fantasy world and purposely, willfully trying to cause real pain to other real people sitting behind other real computers. Origin had chosen to embrace a position close to the ground staked out by our self-described “free-speech maximalists” of today. And like them, Origin was learning that the issue is more dangerously nuanced than they had wished to believe.

But there were others sorts of disconnect at play here as well. Garriott’s stern commandment that his world’s inhabitants should “cease their complaining, rise to the challenge, and make Britannia into the place they want it to be” becomes more than a bit rich when we remember that it was being directed toward Origin’s paying customers. Many of them might have replied that it was up to Origin rather than they themselves to make Britannia a place they wanted to be, lest they choose to spend their $10 per month on something else. The living-world dynamic held “as long as everyone is playing the same game,” wrote Amy Jo Kim in an article about Ultima Online and its increasingly vocalized discontents that appeared in Wired magazine in the spring of 1998. “But what happens when players who think they’re attending an online Renaissance Faire find themselves at the mercy of a violent, abusive gang of thugs? In today’s Britannia, it’s not uncommon to stumble across groups of evil players who talk like Snoop Doggy Dogg, dress like gangstas, and act like rampaging punks.” To be sure, some players were fully onboard with the “living-world” policy of (non-)administration. Others, however, had thought, reasonably enough given what they had read on the back of the game’s box, that they were just buying an entertainment product, a place to hang out in a few hours per day or week and have fun, chatting and exploring and killing monsters. They hadn’t signed up to organize police forces or lead political rallies. Nor had they signed up to be the guinea pigs in some highfalutin social experiment. No; they had signed up to play a game.

As it was, Ultima Online was all but impossible to play casually, thanks not only to the murderers skulking in its every nook and cranny but to core systems of the simulation itself. For example, if you saved up until you could afford to build yourself a nice little house, made it just like you wanted it, then failed to log on for a few days, when you did return you’d find that your home had disappeared, razed to make room for some other, more active player to build something. Systems like these pushed players to spend more time online as a prerequisite to having fun when they were there. Some left when the demands of the game conflicted with those of real life, which was certainly the wisest choice. But some others began to spend far more time in virtual Britannia than was really good for them, raising the specter of gaming addiction, a psychological and sociological problem that would only become more prevalent in the post-millennial age.

Origin estimated that the median hardcore player spent a stunning if not vaguely horrifying total of six hours per day in the virtual world. And if the truth be told, many of the non-murderous things with which they were expected to fill those hours do seem kind of boring on the face of it. This is the flip side of making a virtual world that is more “realistic”: most people play games to escape from reality for a while, not to reenact it. With all due respect to our dedicated and talented real-world tailors and bakers, most people don’t dream of spending their free time doing such jobs online. Small wonder so many became player killers instead; at least doing that was exciting and, for some people at any rate, fun. From Amy Jo Kim’s article:

There’s no shortage of realism in this game — the trouble is, many of the nonviolent activities in Ultima Online are realistic to the point of numbingly lifelike boredom. If you choose to be a tailor, you can make a passable living at it, but only after untold hours of repetitive sewing. And there’s no moral incentive for choosing tailoring — or any honorable, upstanding vocation, for that matter. So why be a tailor? In fact, why not prey on the tailors?

True, Ultima Online is many things to many people. Habitués of online salons come looking for intellectual sparring and verbal repartee. Some other people log on in search of intimate but anonymous social relationships. Still others play the game with cunning yet also a discernible amount of self-restraint, getting rich while staying pretty honest. But there’s no avoiding where the real action is: an ever-growing number are playing Ultima Online to kill everything that moves.

All of this had an effect: all signs are that, after the first rush of sales and subscriptions, Ultima Online began to stagnate, mired in bad reviews, ongoing technical problems, and a growing disenchantment with the player-killing and the other barriers to casual fun. Raph Koster admits that “our subscriber numbers, while stratospheric for the day, weren’t keeping up” with sales of the boxed game, because “the losses [of frustrated newbies] were so high.”

Although Origin and EA never published official sales or subscriber numbers, I have found one useful data point from the early days of Ultima Online, in an internal Origin newsletter dated October 30, 1998. As of this date, just after its first anniversary, the game had 90,000 registered users, of whom approximately half logged on on any given day. These numbers are depicted in the article in question as very impressive, as indeed they were in comparison to the likes of Meridian 59 and The Realm. Still, a bit of context never hurts. Ultima Online had sold 100,000 boxed copies in its first three months, yet it didn’t have even that many subscribers after thirteen months, when its total boxed sales were rounding the 200,000 mark. The subscriber-retention rate, in other words, was not great; a lot of those purchased CDs had become coasters in fairly short order.

Nine shards were up in North America at this time, a number that had stayed the same since the previous December. And it’s this number that may be the most telling one of all. It’s true that, since demand was concentrated at certain times of day, Ultima Online was hosting just about all the players it could handle with its current server infrastructure as of October of 1998. But then again, this was by no means all the players it should be able to handle in the abstract: new shards were generally brought into being in response to increasing numbers of subscribers rather than vice versa. The fact that no new North American shards had been opened since December of 1997 becomes very interesting in this light.

I don’t want to overstate my case here: Ultima Online was extremely successful on its own, somewhat experimental terms. We just need to be sure that we understand what those terms were. By no means were its numbers up there with the industry’s biggest hits. As a point of comparison, let’s take Riven, the long-awaited sequel to the mega-hit adventure game Myst. It was released two months after Ultima Online and went on to sell 1 million units in its first year — at least five times the number of boxed entrées to Origin’s virtual world over the same time period, despite being in a genre that was in marked decline in commercial terms. Another, arguably more pertinent point of comparison is Age of Empires, a new entry in the red-hot real-time-strategy genre. Released just one month after Ultima Online, it outsold Origin’s virtual world by more than ten to one over its first year. Judged as a boxed retail game, Ultima Online was a middling performer at best.

Of course, Ultima Online was not just another boxed retail game; the unique thing about it was that each of the 90,000 subscribers it had retained was paying $10 every month, yielding a steady revenue of almost $11 million per year, with none of it having to be shared with any distributor or retailer. That was really, really nice — nice enough to keep Origin’s head above water at a time when the studio didn’t have a whole lot else to point to by way of justifying its ongoing existence to EA. And yet the reality remained that Ultima Online was a niche obsession rather than a mass-market sensation. As so often happens in life, taking the next step forward in commercial terms, not to mention fending off the competition that was soon to appear with budgets and publisher support of which Meridian 59 and The Realm couldn’t have dreamed, would require a degree of compromise with its founding ideals.

Be that as it may, however, one thing at least was now clear: there was real money to be made in the MMORPG space. Shared virtual worlds would soon learn to prioritize entertainment over experimentation. Going forward, there would be less talk about virtual ecologies and societies, and more focus on delivering slickly packaged fun, of the sort that would keep all kinds of players coming back for more — and, most importantly of all, get those subscriber counts rising once more.

I’ll continue to follow the evolution of PMOG, MMORPGs, and Ultima Online in future articles, and maybe see if I can’t invent some more confusing acronyms while I’m at it. But not right away… other subjects beg for attention in the more immediate future.



Did you enjoy this article? If so, please think about pitching in to help me make many more like it. You can pledge any amount you like.


Sources: the books Braving Britannia: Tales of Life, Love, and Adventure in Ultima Online by Wes Locher, Postmortems: Selected Essays, Volume One by Raph Koster, Online Game Pioneers at Work by Morgan Ramsay, Through the Moongate, Part II by Andrea Contato, Explore/Create by Richard Garriott, and MMOs from the Inside Out by Richard Bartle, and Dungeons and Dreamers by Bard King and John Borland. Origin Systems’s internal newsletter Point of Origin of February 20 1998 and October 30 1998; Computer Gaming World of February 1998 and November 1998; New York Times of October 20 1997; Wired of May 1998.

Web sources include a 2018 Game Developers Conference talk by some of the Ultima Online principals, an Ultima Online timeline at UOGuide, and GameSpot‘s vintage reviews of Ultima Online and its first expansion, The Second Age. On the subject of SubSpace, we have histories by Rod Humble and Epinephrine, another vintage GameSpot review, and a Vice article by Emanuel Maiberg.

 

Tags: , , , , , , , , , ,

The Rise of POMG, Part 3: Competition and Conflict

While the broth of Ultima Online was slowly thickening, not one but two other publishers beat EA and Origin Systems to the punch by releasing graphical persistent virtual worlds of their own. We owe it to them and to ourselves to have a look at these other POMG pioneers before we return to the more widely lauded one that was being built down in Texas. They were known as Meridian 59 and The Realm.


Meridian 59 was inspired by Scepter of Goth,[1]The first word in the name is often spelled Sceptre as well. a rare attempt to commercialize the text-only MUD outside of the walled gardens of online services such as CompuServe and GEnie. After a long gestation period on a mainframe of the Minnesota Educational Computer Consortium, it was ported in 1983 to an IBM PC/XT, to which were cabled sixteen modems and sixteen phone lines, one for each of the players who could be online at any given time. A company called InterPlay — no, not that Interplay — franchised the software out to operators in at least seven American cities. These franchisees then charged their customers an hourly fee to roam around inside the world. The business model worked surprisingly well for a couple of years, until InterPlay’s founder was sent to prison for tax evasion and his company went down with him.

During the fairly brief window of time that Scepter of Goth remained a going concern, a pair of brothers named Andrew and Chris Kirmse fell in love with the incarnation of it that was run out of their hometown of Fairfax, Virginia. Not yet teenagers when they discovered it, they never forgot it after it disappeared. In the summer of 1994, when Andrew had just earned his bachelor’s degree from MIT and Chris had just finished his junior year at Virginia Tech, they set about bringing something similar to life, albeit this time with a top-down graphical view of the world rather than scrolling text. By the end of the year, they felt they had “the foundation of a game,” as Andrew puts it.

A very early version of the game that would evolve into Meridian 59. At this point, it was known as Blackstone.

Then, like so many other young men of their generation and disposition, they found their productivity derailed by a little game called DOOM. “I spent the early part of 1995 playing DOOM II to the exclusion of all else,” admits Andrew. As soon as he had finished all of the single-player levels, he and a friend started to make a DOOM-like engine of their own — again, just as about a million other young programmers were doing at the time. But there was a key difference in Andrew’s case: he didn’t want to make a single-player game, nor even one oriented toward the one-and-done online deathmatches that were all the rage at university campuses all over the country. He rather wanted to combine DOOM with the persistent online game which he and his brother had already begun — that is to say, to make a DOOM that took place in a persistent world.

Andrew and Chris Kirmse cleared their schedules so that they could spend the summer of 1995 in their parents’ basement, figuring out whether it was possible and practical to make the unholy union a reality. With the Internet now entering the public consciousness in a big way, it was a no-brainer to move the game there, where it would be able to welcome far more than sixteen players without requiring a warehouse worth of modems. A handful of other young dreamers joined them as partners in a would-be company called Archetype Interactive, contributing art, world designs, and even a modicum of business acumen from locations all over the country. Like Kali and for that matter DOOM itself, it was the very definition of an underground project, springing to life far from the bright lights of the major publishers, with their slick “interactive movies” and their fixed — and, it would turn out, comprehensively wrong — ideas of the direction mainstream gaming was destined to go. At first the Archetypers wanted to call their game Meridian, simply because they thought the word sounded cool. But they found that the name was already trademarked, so they stuck an arbitrary number at the end of it to wind up with Meridian 59.

By December, they had a bare-bones world with, as Andrew Kirmse says, “no character advancement, no spells, no guilds, no ranged weapons, just the novelty of seeing other people walking around in 3D and talking to them.” Nevertheless, they decided they were ready for an alpha test, several months before Ultima Online would reach the same milestone. They fired up the server late one evening and went to bed, and were thrilled to wake up the next morning and find four people — out of a maximum of 35 — poking around in their world at the same time. Andrew still calls the excitement of that moment “the high point of the entire project.” They redoubled their efforts, roping in more interested observers to provide more art and expand upon the world and its systems, pushing out major updates every few weeks.

In an testament to the endearingly ramshackle nature of the whole project, the world of Meridian 59 was built using a hacked DOOM level editor. Likewise, much of the early art was blatantly stolen from DOOM.

The world went into beta testing in April of 1996. The maximum number of concurrent players had by now been raised by an order of magnitude, but Meridian 59 had become popular enough that the Archetypers still had to kick people out when they needed to log on themselves to check out their handiwork. Among the curious tire-kickers who visited was Kevin Hester, a programmer with The 3DO Company. Founded by Trip Hawkins five years earlier with the intention of bringing a “multimedia console” — don’t call it a games console! — to living rooms everywhere, 3DO was rather at loose ends by this point, having banked on a future of digital entertainment that was badly at odds with the encroaching reality. But Hawkins’s latest instincts were sounder than those of a half-decade previous: he had now decided that online play rather than single-player multimedia extravaganzas was the future. He jumped on Meridian 59 as soon as Hester brought it to his attention, putting together in a matter of days a deal to acquire the budding virtual world and its far-flung network of creators for $5 million in 3DO stock. The Archetypers all signed on the dotted line and moved to Silicon Valley, most of them meeting one another face to face for the first time on their first day in their new office, where they were thrilled to find five servers — enough for five separate instances of their virtual world! — just waiting for them to continue with the beta test.

It had started off like a hacker fairy tale, but the shine wore off quickly enough. Inspired by the shareware example of DOOM, the Kirmse brothers had expected to offer the game client as a free download, with the necessity to pay subscription fees kicking in only after players had been given a few hours to try it out. 3DO vetoed all of this, insisting that the client be made available only as a boxed product with a $50 initial price tag, plus a $15 monthly subscription fee. And instead of being given as much time as they needed to make their new world fit for permanent habitation, as they had been promised they would, the Archetypers were now told that they had to begin welcoming paying customers into Meridian 59 in less than three months. Damion Schubert, Meridian 59‘s world-design lead, claims that “3DO was using us to learn about the business of online gaming,” seeing their very first virtual world as a stepping-stone rather than a destination unto itself. Whatever the truth of that assertion, it is a matter of record that, while the Archetypers were trying to meet 3DO’s deadline, the stock they had been given was in free fall, losing 75 percent of its value in those first three months, thereby doing that much more to convince the accountants that Meridian 59 absolutely, positively had to ship before 3DO’s next fiscal year began on October 1.

An aesthetic triumph Meridian 59 was not.

So, the game that was officially released on September 27, 1996, was not quite the one the Kirmses had envisioned when they signed the contract with 3DO. To call it little more than a massively-multiplayer DOOM deathmatch with a chat system grafted on would be unkind but not totally unfair. Its pseudo-3D engine would have looked badly outdated in 1996, the year of Quake, even if the art hadn’t been such a mismatched grab bag of aesthetics and resolutions. Meridian 59 evinced none of the simulational aspirations of Ultima Online; this was not a world in which anyone was going to pass the time baking bread or chopping lumber. For lack of much else to do, people mostly occupied themselves by killing one another. Like Ultima Online, the software permitted player-versus-player combat anywhere and everywhere; unlike Ultima Online, there were no guards patrolling any of the world’s spaces to disincentivize it. A Meridian 59 server was a purely kill-or-be-killed sort of world, host to a new war every single day. Because there was no budget to add much other content to the world, this was just as well with its creators; indeed, they soon learned to lean into it hard. Activities in the world came to revolve around the possession of guild halls, of which each server boasted ten of varying degrees of splendor for the disparate factions to fight over. If you didn’t like to fight with your fellow players more or less constantly, Meridian 59 probably wasn’t the game for you.

Handed the first-ever full-fledged massively-multiplayer online role-playing game, 3DO’s marketers chose to… write non-sequiters about latex. This might be the worst advertisement I’ve ever seen; I literally have no idea what joke it’s trying and failing to land. Something about condoms, I presume?

Luckily, there were plenty of gamers who really, really did like to fight, as the popularity of DOOM deathmatches illustrated. Despite its dated graphics and despite promotional efforts from 3DO that were bizarrely inept when they weren’t nonexistent, Meridian 59 managed to attract 20,000 or more subscribers and to retain them for a good while, keeping all ten of the servers that were given over to it after the beta test humming along at near capacity most of the time. 3DO even approved a couple of boxed expansion packs that added a modicum of additional content.

But then, in late 1997, 3DO all but killed the virtual world dead at a stroke. Deciding it was unjust that casual players who logged on only occasionally paid the same subscription fee as heavy users who spent many hours per day online, they rejiggered the pricing formula into a tangle of numbers that would have baffled an income-tax accountant: $2.49 per day that one logged on, capped at $9.99 per week, with total fees also capped at $29.99 per month. But never mind the details. Since the largest chunk of subscribers by far belonged to the heavy-user category, it boiled down to a doubling of the subscription price, from $15 to $30 per month. The populations on the servers cratered as a result. Meridian 59‘s best days — or at least its most populous ones — thus passed into history.


The other graphical MMORPG to beat Ultima Online to market had a very different personality. Sierra’s The Realm was the direct result of Ken Williams’s musings about what an “online adventure game” might be like, the same ones that I quoted at some length in my last article. After trying and failing to convince Roberta Williams to add a multiplayer option to King’s Quest VII, he went to a programmer named David Slayback, saying, “Wouldn’t it be cool if we could do something like our adventure games, that was Medieval themed, and allowed players to swap items with each other, buy weapons, and attack monsters?” Slayback then took the ball and ran with it; as Ken himself acknowledges, that initial conversation was “the limit of my involvement creatively.”

The original plan was for The Realm to become a part of America Online, the great survivor of the pre-Web era of commercial online services. That deal, however, fell through. Meanwhile Sierra was itself acquired by an e-commerce firm called CUC International, and The Realm seemed to fall between two stools amidst the reshuffling of deck chairs that followed. A beta test in the summer of 1996 did lead to the acceptance of the first paying subscribers in December of that year, but Sierra never did any real promotion beyond its own customer magazine, making the client software available only via mail order. Still, by all indications this virtual world attracted a number of players comparable to that of Meridian 59, perhaps not least because in its case buying the boxed client entitled the customer to a full year of free online play.

The Realm stands today as a rather fascinating artifact, being the road largely not taken in the MMORPG space. In presentation, aesthetics, and culture, it has more in common with Habitat, an amazingly early attempt by Lucasfilm Games and America Online’s direct predecessor Quantum Link to build a non-competitive graphical space for online socializing, than it does with either Meridian 59 or Ultima Online. This world was very clear about where its priorities lay: “The Realm offers you a unique environment in which to socialize with online friends (or make some new ones) and also gives you something fun to do while you’re socializing.” It was, in other words, a case of social space first, game second. As such, it might be better read as a progenitor to the likes of Second Life or The Sims Online than something like World of Warcraft.

Each player started in her own house, which she had to fight neither to acquire nor to defend. The interface was set up like one of the point-and-click graphic adventures that had been Sierra’s bread and butter since the mid-1980s, with the player guiding her avatar in the third person across a map made up of “rooms” that filled exactly one screen each. The graphical style too was right out of King’s Quest. None of this is terribly surprising, given that The Realm was built using SCI, Sierra’s venerable adventure-game engine.

Although there were monsters to fight and treasure to collect, player-versus-player combat was impossible. Even profanity was expressly forbidden. (“This includes ‘masking’ by using asterisks as part of the word,” noted the FAQ carefully.) The combat was also unusual in that it was turn-based. This choice, combined with the way that The Realm off-loaded an unusual amount of work to the player’s local client, meant that Sierra didn’t have to spread it across multiple servers; uniquely for this era, there really was just one Realm.

All of this attracted a dramatically different clientele from that of Meridian 59; many more women hung out in The Realm, for one thing. Interior decoration and fashion trumped murder and theft in the typical range of pursuits. Beth Demetrescu wrote in Sierra’s magazine InterAction about her own first days there:

As with all newbies, I started in my house. I was a poor, hungry, fashion faux pas. After I got out of my house, moved about six screens, and was lost in my hometown, I encountered HorseWoman, whose biography said she was an eleven-year-old. She took me to her home, gave me decent clothes, and taught me about basic communication, navigation, and combat. This was my first experience with the warm, welcoming community of The Realm.

I soon found myself outside of the town fighting rats. There are plenty of large, ferocious beasts to fight, but for the time being, all I could handle were rats. I was really worried the first time one of these rats killed me, thinking I was going to get kicked out of the game and would have to log back on. Instead, I lost everything I was carrying, but I was found by wanderers who dragged me home to heal…

I learned of Realm weddings. BlueRose, the Justice of the Peace, often called the Lady of Love, conducts over half of the Realm weddings…

I have picked up several valuable things from the many Realmers I have encountered. Not only did I get important information on The Realm’s features and inhabitants, but I also learned from their example about The Realm’s vast, multinational community. These people are friendly and helpful.

The contrast with Meridian 59, where a bewildered newbie was more likely to be given a broadsword to the back of the neck than navigational and sartorial assistance, could hardly have been greater.

A wedding in The Realm.

All told, then, Meridian 59 and The Realm provided the early MMORPG space with its yang and its yin: the one being a hyper-violent, hyper-competitive free-for-all where pretty much anything went, the other a friendly social space that was kept that way by tight moderation. Nevertheless, the two did have some things in common. Neither ever became more than moderately popular, for one — and that according to a pretty generous interpretation of “moderately” in a fast-expanding games industry. And yet both proved weirdly hard to kill. In fact, both are still alive to this day, abandoned decades ago by their original publishers but kept online by hook or by crook by folks who simply refuse to let them go away — certainly not now, when the aged code that makes their worlds come alive can be run for a pittance on a low-end server tucked away in some back corner of an office or data center somewhere. Their populations on any given evening may now be in the dozens rather than the hundreds or thousands, but these virtual worlds abide. In this too, they’ve set a precedent for their posterity; the Internet of today is fairly littered with online games whose heyday of press notices and mainstream popularity are well behind them, but that seem determined to soldier on until the last grizzled graybeard who cut his teeth on them in his formative years shuffles off this mortal coil. MMORPGs especially are a bit like cockroaches in this respect — with no insult to either the worlds or the insects in question intended. Suffice to say that community can be a disarmingly resilient thing.



But we return now to the story of Ultima Online, whose makers viewed the less than overwhelming commercial acceptance of Meridian 59 and The Realm with some ambivalence. On the one hand, Ultima Online had avoided having its own thunder stolen by another MMORPG sensation. On the other, these other virtual worlds’ middling trajectories gave no obvious reason to feel hugely confident in Ultima Online‘s own commercial prospects.

This was a problem not least because, as 1996 turned the corner into 1997, the project’s financial well had just about run dry, just as this virtual Britannia was ready to go from the alpha to the beta stage of testing, with ten to twenty times the number of participants of earlier testing rounds. It wasn’t clear how this next step could be managed under the circumstances; the client software was by now too big to ask prospective testers to download it in its entirety in this era of dial-up connections, yet there simply wasn’t sufficient money in the budget to stamp and ship 20,000 or more CDs out to them. The team decided there was only one option, cheeky though it seemed: to ask each participant in effect to pay Origin for the privilege of testing their game for them, by sending in $5 to cover the cost of the CD. The principals claim today that 50,000 people did so as soon as the test was announced online, burying Origin in incoming mail; I suspect this number may be inflated somewhat, as many of those associated with Ultima Online tend to be in the memories of those who made it. But regardless of the exact figure, the response definitely was considerable, not to mention gratifying for the little team of ex-MUDders who had been laboring in disrespected obscurity up there on a gutted fifth floor. It was the first piece of incontrovertible evidence that there were significant numbers of people out there who were really, really excited by the idea of living out an Ultima game with thousands of others.

The original Ultima Online beta CDs have become coveted collectors’ items.

As the creators tell the story, the massive popular reaction to the call for beta testers was solely responsible for changing the hearts and minds of their managers at EA and Origin. Realizing suddenly that Ultima Online had serious moneymaking potential, they went overnight from passive-aggressively trying to kill it to being all-in with bells on. In March of 1997, they moved the MUDders from their barren exile down to the scene of the most important action at Origin, where a much larger team had been working on Ultima IX, the latest iteration in the single-player series. Yet it was the latter project that was now to go on hiatus, not Ultima Online. This new amalgamation of developers, five or six times the size of the team of the day before, had but one mandate: get the virtual world done already. After two years of living hand to mouth, the original world-builders had merely to state their wishes in terms of resources in order to see them granted.

Most of the conceptual work of building this new online world had already been done by the time the team was so dramatically expanded. Still, we shouldn’t dismiss the importance of this sudden influx of sometimes unwilling bandwagon jumpers. For they made Ultima Online look like at least a passable imitation of a AAA prestige project, in a way that Meridian 59 and The Realm did not. A high design standard combined with a relatively high audiovisual one would prove a potent combination.

With its isometric perspective, Ultima Online most resembled Ultima VII in terms of presentation. The graphics were by no means cutting-edge — Ultima VII had come out back in 1992, after all — but they were bright and attractive, without going full-on cartoon like The Realm.

Did all of this really happen simply because the response to the call for beta testers was better than expected? I have no smoking gun either way, but I must say that I tend to doubt it. Just about everyone loves a good creatives-versus-suits story, such that we seldom question them. Yet the reasoning that went on in the executive suites prior to this turnaround in Ultima Online‘s fortunes was perhaps a little more complex than that of a pack of ravenous wolves chasing a tasty rabbit that had finally been revealed to their unimaginative minds. Whatever else one can say about them, most of the suits didn’t get where they were by being stupid. So, maybe we should try to see the situation from their perspective — try to see what Origin looked like to the outsiders at EA’s California headquarters.

Throughout the 1990s, Origin lived on two franchises: Richard Garriott’s Ultima and Chris Roberts’s Wing Commander. To be sure, there were other games here and there, some of which even turned modest profits, but it was these two series that kept the lights on. When EA acquired Origin in September of 1992, both franchises were by all indications in rude health. Wing Commander I and II and a string of mission packs for each were doing tremendous numbers. Ultima VII, the latest release in Richard Garriott’s mainline series, had put up more middling sales figures, but it had been rescued by the spinoff Ultima Underworld, which had come out of nowhere — or more specifically out of the Boston-based studio Blue Sky Productions, soon to be rebranded as Looking Glass — to become another of the year’s biggest hits.

Understandably under the circumstances, EA overlooked what a dysfunctional workplace Origin was already becoming by the time of the acquisition, divided as it was between two camps: the “Friends of Richard” and the “Friends of Chris.” Those two personifications of Origin’s split identity were equally mercurial and equally prone to unrealistic flights of fancy; one can’t help but sense that both of their perceptions of the real world and their place in it had been to one degree or another warped by their having become icons of worship for a cult of adoring gamers at an improbably young age. Small wonder that EA grew concerned that there weren’t enough grounded adults in the room down in Austin, and, after first promising a hands-off approach, showed more and more of a tendency to micro-manage as time went on — so much so that, as we learned in the last article, Garriott was soon reduced to begging for money to start his online passion project.

Wing Commander maintained its momentum for quite some time after the acquisition, even after DOOM came along to upend much of the industry’s conventional wisdom with its focus on pure action at the expense of story and world-building, the things for which both Garriott and Roberts were most known. Wing Commander III was released almost a year after DOOM in late 1994 with a cast of real actors headed by Mark Hamill of Star Wars fame, and became another huge success. Ultima, however, started to lose its way almost as soon as the ink was dry on the acquisition contract. Ultima VIII, which was also released in 1994, chased the latest trends by introducing a strong action element and simplifying most other aspects of its gameplay. This was not done, as some fan narratives wish to state, at the behest of EA’s management, but rather at that of Richard Garriott himself, who feared that his signature franchise was at risk of becoming irrelevant. That said, EA can and should largely take the blame for the game being released too early, in a woefully buggy and unpolished state. The critical and commercial response was nothing short of disastrous, leaving plenty of blame to go around. Fans complained that Ultima VIII had more in common with Super Mario Bros. than the storied Ultima games of the past, bestowing upon it the nickname Super Avatar Bros. in a backhanded homage to the series’s most hallowed incarnation, 1985’s Ultima IV: Quest of the Avatar, whose unabashed idealism now seemed like something from a lifetime ago in a parallel universe.

Then, in early 1996, Origin’s other franchise went squishy as well. As the studio’s own press releases breathlessly trumpeted, Wing Commander IV was the most expensive digital game ever made to that date, with a claimed production budget of $12 million. The vast majority of that money went into a real Hollywood film shoot, directed by Chris Roberts himself and starring a returning Mark Hamill among a number of other recognizable faces from the silver screen. Wing Commander IV wound up costing four times as much as its predecessor and selling half as many copies, taking months of huffing and puffing to just about reach the break-even point. The interactive-movie era had reached the phase of diminishing returns; under no circumstances was EA going to let Origin make a game like this one again.

But what kinds of games should Origin be making? That was the million-dollar question in the aftermath of Wing Commander IV. After Chris Roberts left the studio to pursue his dream of becoming the latest George Lucas in Hollywood, Origin announced that his series was to be continued on a less grandiose scale, moving some of the focus away from the cut scenes and back to the gameplay. Yet there was no reason to believe such games would make many inroads beyond the hardcore Wing Commander faithful. Meanwhile Richard Garriott had pledged to repair the damage done by Ultima VIII, by making the next single-player Ultima the biggest, best one ever. But epic CRPGs in general had been in the doldrums for years, and the Ultima IX project was already showing signs of becoming another over-hyped, over-expensive boondoggle like Wing Commander IV. Exacerbating the situation was the loss of two of the only people at Origin who had shown themselves to be capable of restraining and channeling Garriott’s flights of fancy. Origin and EA alike felt keenly the loss of the diplomatic and self-effacing designer and producer Warren Spector, the first everyday project lead on Ultima IX, who decamped to Looking Glass in 1995 when that project was still in its infancy. Ditto the production manager Dallas Snell, a less cuddly character whose talent for Just Getting Things Done — by cracking heads if necessary — was almost equally invaluable.

Of course, one can still ding EA for failing to see that Richard Garriott was onto something with Ultima Online long before they did. In their partial defense, though, Garriott tended to propose a lot of crazy stuff. As his checkered post-millennial career in game development illustrates all too clearly, he has not been a detail-oriented creator since his days of conceiving and coding the early Ultima games all by himself. This has made his ideas — even his good ones, which Ultima Online certainly was — all too easy to dismiss.

Nonetheless, the potential of persistent online multiplayer gaming was becoming impossible to deny by early 1997, what with the vibrant virtual communities being built on the likes of Kali and Battle.net, in addition to the smaller but no less dedicated ones that had sprung up in Meridian 59 and The Realm. You’d have to be a fool not to be intrigued by the potential of Ultima Online in a milieu such as this one — and, again, EA’s executives most definitely weren’t fools. They wanted to keep Origin alive and viable and relevant as badly as anyone else. Suddenly this seemed the best way to do so. Thus the mass personnel transfer from Ultima IX, which was increasingly smelling like gaming’s past, into Ultima Online, which had the distinct whiff of its future.

It was a difficult transition for everyone, made that much more difficult by the fact that most of the people involved were still in their twenties, with all of the arrogant absolutism of youth. Both the project’s old-timers and its newcomers had plenty of perfectly valid complaints to hurl at their counterparts. Raph Koster, who had been told that he was the design lead, was ignored by more experienced developers who thought they knew better. And yet he did little for his cause by, as he admits today, “sulking and being very rude” and “behaving badly and improperly” even to Richard Garriott himself. From his point of view, the newcomers showed that they fundamentally didn’t understand online games when they wasted their time on fluff that players who needed to be captured for months or years would burn through in a matter of hours, such as lengthy, single-player-Ultima-style conversation trees for the non-player characters. Yet the newcomers were right to express shock and horror when they found that, amidst all the loving attention that had been given to simulating Britannia’s ecology and the like, no one on the original team had thought up a consistent system for casting spells, a bedrock of Ultima‘s appeal since the very beginning. Even today, one Ultima IX refugee accuses the MUDders of being “focused on minutia, what I would call silly little details that really added nothing to the game.”

When the two-month-long beta test finally began after repeated delays in June of 1997, the dogged simulation-first mentality of Koster and company faced a harsh reckoning with reality. Many of the systems that had seemed wonderful in theory didn’t work in practice, or displayed side effects that they’d never anticipated. Here as in many digital games, attempting to push the simulation too far just plunged the whole thing into a sort of uncanny valley, making it feel more rather than less artificial. For instance, the MUDders had made it possible for you to learn or improve skills simply by standing in close proximity to someone who was using the skill in question at a high level, on the assumption that your character was observing and internalizing this example of a master at work. But they’d also instituted a cap on the total pool of skill points a character could possess across all disciplines, on the assumption that no Jack of all trades could be a master of them all; just as is the case for most of us in real life, in Ultima Online you could be really good at a few things, or fair at a lot of them, but not really good at a lot of them. When a character hit her skill-point cap, learning new things would cause some of her other skills to decline to stay under it. In practice, this caused players to desperately try to avoid seeing what that baker or weaver was doing, for fear of losing their ability to hunt or cast spells as a result. Problems like these hammered home again and again the fact that any digital simulation is only the crudest approximation of a lived existence; in the real world, matters are not quite so zero-sum as instantly losing the ability to catch a fish because one has learned to cook a fish.

But the most extreme case of unforeseen consequences involved the aforementioned lovingly crafted ecology of virtual Britannia. To put it bluntly, the players destroyed it — all of it, within days if not hours. The population of deer and rabbits, the food sources of apex predators like dragons, were slaughtered to extinction by players instead. This was not done out of sheer bloody-mindedness alone, although that was undoubtedly a part of the equation. The truth was that deer and rabbits had value, in the form of meat and pelts. In a sense, then, virtual Britannia was becoming a real economy, just as its creators had always hoped it would. But it was an economy without real-world limits or controls, unimpeded by consequences which were themselves only virtual, never real; no one was going to go hungry in real life for over-hunting the forests and fields of Britannia. The same went for trees and fish and a hundred other precious resources that we of the real world usually make some effort to conserve, however imperfectly. With the simulation spinning wildly out of control, Origin had to start putting its thumb on the scales, applying external remedies such as magically re-spawning rabbits and trees, lest the world degenerate into a post-apocalyptic wasteland, a deserted moonscape where roving bands of starving players were chased hither and yon by equally hungry dragons. People came to an Ultima game expecting a Renaissance Faire version of Merry Olde England, not a Harlan Ellison story.

Of course, the external corrections themselves had further knock-on consequences. By creating an endless supply of animals to hunt and trees to fell, Origin was in effect giving the economy a massive, perpetual external stimulus. The overseers were therefore always on the lookout for ways to suck gold back out of the world. Ironically, one of the best was to let players get killed a lot, since between death and resurrection they lost whatever money they’d been carrying with them. Thus Origin had a perverse incentive not to try too hard to make Britannia a safer, more friendly place.

Such collisions between idealism and reality were scarring for the MUDders. “This was a wake-up call for me,” says Raph Koster. “The limits on what we can get an audience to go along with, and how much we can affect the bottom line. A lot of people [on the development team] were emotionally hurt by the player killing. Many of the tactics we would use on MUDs just didn’t work at a large scale. Players behaved differently. They were ruder to one another.” All of which is to say that Richard Garriott’s fondly expressed wish that the persistent quality of Ultima Online would serve to put a brake on the more toxic ways of acting out on the anonymous Internet did not come to pass to anything like the extent he had imagined.

At the same time, though, it wasn’t all destruction and disillusionment during that summer of the beta test. Some players proved less interested in killing than they were in crafting, becoming armorers and blacksmiths, jewelers and merchants, chefs and bankers and real-estate agents. Players cooked food and sold it from booths in the center of the cities or earned a (virtual) living as tour guides, leading groups of people on treks to scenic but dangerous corners of the world. Enterprising wizards set up a sort of long-distance bus service, opening up magical portals to shuttle their fellow players instantly from one side of the world to another for a fee. Many of the surprises of the beta period were just the kind the MUDders had been hoping to see, emerging from the raw simulational affordances of the environment. “[Players] used the ability to dye clothing to make uniforms for their guilds,” says Raph Koster, “and they [held] weddings with coordinated bridesmaids dresses. They started holding sporting events. They founded theater troupes and taverns and police forces.” The agents of chaos may have been perpetually beating at the door, but there was a measure of civilization appearing in virtual Britannia as well.

Or rather in the virtual Britannias. One of the most frustrating compromises the creators had to make was necessitated by, as compromises usually are in game development, the practical limitations of the technology they had to hand. There was no way that any one of the servers they possessed could contain the number of players the beta test had attracted. So, there had to be two virtual Britannias rather than just one, the precursors to many more that would follow. Both Garriott and Koster have claimed to be the one who came up with the word “shards” as a name for these separate servers, each housing its own initially identical but quickly diverging version of Britannia. The name was grounded in the lore of the very early days of Ultima. In Ultima I back in 1981, the player had shattered the Gem of Immortality, the key to the power of that game’s villain, the evil wizard Mondain. It was claimed now that each of the jewel’s shards had contained a copy of the world of Britannia, and that these were the duplicate worlds inhabited by the players of Ultima Online. Rather amusingly, the word “shard” has since become a generalized term for separate but equal server instances, co-opted not only by other MMORPGs but by administrators of large de-centralized online databases of many stripes, most of which have nothing to do with games.

Each shard could host about 2500 players at once. In these days when the nation’s Internet infrastructure was still in a relatively unrefined state, such that latency tended to increase almost linearly with distance, the shards were named after their real-world locations — there was one on each coast in the beginning, named “Atlantic” and “Pacific” — and players were encouraged to choose the server closest to them if at all possible. (Such concerns would become less pressing as the years went by, but to this day Ultima Online has continued the practice of naming its virtual Britannias after the locations of the servers in the real world.)

On the last day of the beta test, there occurred one of the more famous events in the history of Ultima Online, one with the flavor of a Biblical allegory if not a premonition. Richard Garriott, playing in-character as Lord British, made a farewell tour of the shards in the final hours, to thank everyone for participating before the servers were shut down, not to be booted up again until Ultima Online went live as a paid commercial service. Among fans of the single-player Ultima games, there was a longstanding tradition of finding ways to kill Lord British, who always appeared as a character in them as well. People had transplanted the tradition into Ultima Online with a vengeance, but to no avail; acknowledging that even the most stalwart commitment to simulation must have its limits when it comes to the person who signs your paycheck, the MUDders had agreed to provide Lord British with an “invulnerability” flag. As he stood up now before a crowd on the Pacific shard to deliver his valediction, someone threw a fireball spell at him. No matter; Lord British stepped confidently right into the flames. Whereupon he fell over and died. Someone had forgotten to set the invulnerability flag.

If Lord British couldn’t be protected, decided the folks at Origin on the spur of the moment, he must be avenged; in so deciding, they demonstrated how alluring virtual violence could be even to those most dedicated to creating a virtual civilization. Garriott:

It’s amazing how quickly the cloak of civilization can disappear. The word spread verbally throughout the office: let us unleash hell! My staff summoned demons and devils and dragons and all of the nightmarish creatures of the game, and they cast spells and created dark clouds and lightning that struck and killed people. The gamemasters had special powers, and once they realized I had been killed, they were able to almost instantly resurrect Lord British. And I gleefully joined in the revelry. Kill me, will you? Be gone, mortals! It was a slaughter of thousands of players in the courtyard.

It definitely was not the noble ending we had intended.

And while some players enjoyed the spontaneity of this event, others were saddened or hurt by it. When most characters die they turn into a ghost and are transported to a distant place on the map. Then they have to go find their body. So the cost of being killed is a temporary existence as a ghost. In the last three minutes of these characters’ existence, they suddenly found themselves alone, deep in the woods, unable to speak or interact with anyone else. The net result of this mass killing in retaliation for the assassination of Lord British was that not only were all of these innocent people slaughtered, they were also cast out of the presence of the creators at the final moment. As the final seconds trickled down, they desperately tried to get back, but most often failed. The fact that all of us, the creators and the players, were able to turn the last few moments of the beta test into this completely unplanned and even unimagined chaos was proof that we had built something unique, a platform that would allow players to do pretty much whatever they pleased, and that it was about to take on a life — and many deaths — of its own.

After more than two and a half years, during which the face of the games industry around it had changed dramatically and its own importance to its parent company had been elevated incalculably, Ultima Online was about to greet the real world as a commercial product. Whether the last minutes of its existence while it was still officially an experiment boded well or ill for its future depended on your point of view. But, as Richard Garriott says, the one certain thing was uncertainty: nobody knew quite what would happen next. Would Ultima Online be another Meridian 59 or The Realm, or would this be the virtual world that finally broke through? And what would it mean for gaming — and, for that matter, for the real world beyond gaming — if it did?



Did you enjoy this article? If so, please think about pitching in to help me make many more like it. You can pledge any amount you like.


Sources: the books Braving Britannia: Tales of Life, Love, and Adventure in Ultima Online by Wes Locher, Postmortems: Selected Essays, Volume One by Raph Koster, Online Game Pioneers at Work by Morgan Ramsay, Through the Moongate, Part II by Andrea Contato, Explore/Create by Richard Garriott, MMOs from the Inside Out by Richard Bartle, and Not All Fairy Tales Have Happy Endings by Ken Williams; Sierra’s customer newsletter InterAction of Spring 1996, Summer 1996, Spring 1997, Summer 1997, Fall 1997, Summer 1998, and Fall 1998; PC Powerplay of November 1996; Next Generation of March 1997.

Web sources include a 2018 Game Developers Conference talk by some of the Ultima Online principals, an Ultima Online timeline at UOGuide, “How Scepter of Goth Shaped the MMO Industry” by Justin Olivetti at Massively Overpowered, David A. Wheeler’s history of Scepter of Goth, “How the World’s Oldest 3D MMO Keeps Cheating Death” by Samuel Axon at Vice, Andrew Kirmse’s own early history of Meridian 59, Damion Schubert’s Meridian 59 postmortem and its accompanying slides from the 2012 Game Developers Conference, and Gavin Annand’s video interview with the Kirmse brothers.

Footnotes

Footnotes
1 The first word in the name is often spelled Sceptre as well.
 
18 Comments

Posted by on February 16, 2024 in Digital Antiquaria, Interactive Fiction

 

Tags: , , , , , , , , , , ,