RSS

Tag Archives: ultima online

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.” (His wife Kristen Koster may have been one of them; she has never worked on another game since Ultima Online, and never spoken publicly about her time at Origin.) “Many of the tactics we would use on MUDs just didn’t work at a large scale,” continues Raph. “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.
 
16 Comments

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

 

Tags: , , , , , , , ,

The Rise of POMG, Part 2: Multima

This sign, executed in throwback 1980s Ultima iconography, hung on the wall outside the elevator on the fifth floor of Origin Systems’s office building, pointing the way to the Multima team. If you neglected to follow the sign’s advice and turned left instead of right here, you would plunge five stories to your doom.

To carry our story forward into the next phase of persistent online multiplayer gaming, we first need to go backward again — all the way back to the days when tabletop Dungeons & Dragons was new on the scene and people were first beginning to imagine how computers might make it more accessible. Some dreamed of eliminating the pesky need for other humans to play with entirely; others wondered whether it might be possible to use networked computers in place of the tabletop, so that you could get together and play with your friends without all of the logistical complications of meeting up in person. Still others dreamed bigger, dreamed of things that would never be possible even with a tabletop big enough to make King Arthur blush. What if you could use computers to make a living virtual world with thousands of human inhabitants? In a column in a 1983 issue of Starlog magazine, Lenny Kaye mused about just such a thing.

When played among groups of people, [tabletop] RPGs foster a sense of cooperation toward a common goal, something videogames have hardly approached.

But with the computer’s aid, the idea of the game network can expand outward, interfacing with all sorts of societal drifts. The technology for both video[games] and RPGs is still in its infancy — one with graphics and memory far from ideal, the other still attached to its boards and figurines — and yet, it’s not hard to imagine a nationwide game, in which all citizens play their part. Perhaps it becomes an arena in which to exercise and exorcise ourselves, releasing our animal instincts through the power of the mind, understanding the uses and misuses of our humanity.

While they waited for the technology which could make that dream a reality to appear, people did what they could with what they had. The ones who came closest to the ideal of a “nationwide game” were those running MUDs, those “multi-user dungeons” that allowed up to 100 players to interact with one another by typing commands into a textual parser, with teletype-style streaming text as their eyes and ears into the world they all shared.

Island of Kesmai did soon come along on the big online service CompuServe, with an interface that looked more like Rogue than the original game of Adventure. But again, only 100 people could play together simultaneously there, about the same number as on the biggest MUDs. And in this case, each of them had to pay by the minute for the privilege. For all that it was amazing that it could exist at all in its time and place, Island of Kesmai was more like a wealthy gated village than a teeming virtual world.

Still, it must be acknowledged that the sellers of traditional boxed computer games were even farther away from that aspiration, being content to offer up single-player CRPGs where combat — that being the aspect of tabletop RPGs that was easiest to implement on a computer — tended to overwhelm everything else. The one obvious exception to this norm was Origin Systems of Austin, Texas. Especially after 1985’s landmark Ultima IV, Origin’s Ultima series became not only the commercial standard bearer for CRPGs but the best argument for the genre’s potential to be about something more than statistics and combat tactics. These games were rather about world-building and about the Virtues of the (player’s) Avatar, daring to introduce an ethical philosophy that was applicable to the real world — and then, in later installments, to muddy the waters by mercilessly probing the practical limitations and blind spots inherent in any such rigid ethical code.

Richard Garriott — “Lord British” to his legion of fans, the creator of Ultima and co-founder of Origin — had first conceived of the games as a re-creation of the tabletop Dungeons & Dragons sessions that had made him a minor celebrity in his Houston neighborhood well before he got his hands on his first Apple II. The setting of computerized Ultima was the world of Britannia, the same one he had invented for his high-school friends. The idea of adventuring together with others through the medium of the computer thus made a lot of sense to him. Throughout the 1980s and beyond, he repeatedly brought up the hypothetical game that he called Multima, a multiplayer version of Ultima. In 1987, an Origin programmer was actually assigned to work on it for a while. Garriott, from a contemporary interview:

James Van Artsdalen, who does our IBM and Macintosh translations, is working on a program that lets several people participate in the game. Two people can do this with different computers directly connected via modems, or even more can play via a system with multiple modems. We don’t know if we’ll be able to support packet networks like CompuServe because they may be too slow for this application. We’ll do it if we can.

What you’ll buy in the store will be a package containing all the core graphics routines and the game-development stuff (all the commands and so on), which you could even plug into your computer and play as a standalone. But with a modem you could tie a friend into the game, or up to somewhere between eight and sixteen other players, all within the same game.

We will most likely run a game of this out of our office. Basically, we can almost gamemaster it. There could be a similar setup in each town, and anybody could run one. Our intention is to let anyone capable of having multiple modems on their system have the network software. Anybody can be a node: two people can play if they each own a package, just by calling each other. But to be a base, a multiplayer node, you’ve got to have multiple modems and may need additional software. If the additional software is needed, we’ll let anyone who wants it have it, since we’re just supporting sales of our own products anyway.

Should be a lot of fun, we think. We hope to have it out by next summer [the summer of 1988], and it should cost between $30 and $60. This will be able to tie different kinds of computers together, since the information being sent back and forth doesn’t include any graphics. I could play on my Apple while you’re on your IBM, for instance. The graphics will be full state-of-the-art Ultima graphics, but they’ll already be on your computer.

Alas, this Multima was quietly abandoned soon after the interview, having been judged just too uncertain a project to invest significant resources into when there was guaranteed money to be made from each new single-player Ultima. As we’ll soon see, it was not the last time that argument was made against a Multima.

Circa 1990, Multima was revived for a time, this time as a three-way partnership among Origin, the commercial online service GEnie, and Kesmai, the maker of Island of Kesmai. The last would be given the source code to Ultima VI, the latest iteration of the single-player series, and would adapt it for online play, after which GEnie would deploy it on its central mainframe. As it happened, a similar deal was already taking SSI’s Gold Box engine for licensed Dungeons & Dragons CRPGs through the same process of transformation. It would result in something called Neverwinter Nights[1]Not the same game as the 2002 Bioware CRPG of the same name. going up on America Online in 1991 with an initial capacity of 100 simultaneous players, eventually to be raised by popular demand to 500, making it the most massively multiplayer online game of the early 1990s. But Multima was not so lucky; the deal fell through before it had gotten beyond the planning stages, perchance having proved a simple case of too many cooks in the kitchen.

Richard Garriott had largely ceased to involve himself in the day-to-day work of making new Ultima games by this point, but he continued to set the overall trajectory of the franchise. And as he did so, he never forgot those old hopes for a Multima. Not long after Electronic Arts (EA) purchased Origin Systems in 1992, he thought he saw the stars aligning at long last. For 1993 brought with it NCSA Mosaic, the first broadly popular multimedia Web browser, the harbinger of the home Internet boom. This newly accessible public Internet could replace fragile peer-to-peer modems connections as easily as it could the bespoke private networks of CompuServe and GEnie, making the likes of Multima seem far more practical, both to implement and to offer to customers at an affordable price.

Richard Garriott in 1995.

For better or for worse, though, Origin was now a part of EA, who had the final say on which projects got funded. Garriott tried repeatedly over a period of a year and a half or more to interest EA’s CEO Larry Probst in his Multima schemes, without success. As he tells the story, he and a couple of other true believers from Origin resorted to guerrilla tactics at the third formal pitch meeting.

We literally just refused to give up the floor. They said, “No. Get out.” And we stomped our feet and held our breath. “We are not leaving until you guys get a clue. As a developer, we go over-budget by 25 to 50 percent every year. We’re only asking for $250,000 to build a prototype.” I already had a piece of paper prepared: “You approve of us going $250,000 over-budget to prove that this can work.” And finally, after heated yelling and screaming, they signed the piece of paper and kicked us in the ass as we left the room.

It must be said here that Garriott is generally not one to let an overly fussy allegiance to pedantic truth get in the way of a good story. Thus I suspect that the decision to green-light a Multima prototype may not have been arrived at in quite so dramatic a fashion as the tale above. (Garriott has in fact told a number of versions of the story over the years; in another of them, he’s alone in Larry Probst’s office, hectoring him one-and-one into signing the note granting him $250,000 to investigate the possibility.) Most important for our purposes, however, is that the decision was definitely made by early 1995.

The answer to another question is even more murky: that of just when the change was made from an Ultima that you could play online with a few of your buddies — a feature that would be seen well before the end of the 1990s in a number of otherwise more conceptually conventional CRPGs, such as Diablo and Baldur’s Gate — to a persistent virtual Britannia inhabited by dozens, hundreds, or even thousands of other players. The acronym by which we refer to such worlds today would be coined by Richard Garriott in 1997: the “massively multiplayer online role-playing game,” or MMORPG. (And you thought POMG was a mouthful!) Yet it appears that Multima may not have become an MMORPG in the minds of its creators until after Garriott secured his funding, and that that hugely important conceptual shift may not have originated with Garriott himself. One alternative candidate is the man to whom Garriott turned to become the nuts-and-bolts administrator of the project: Starr Long, a quality-assurance lead who had been one of his few Multima allies inside the company for quite some time. (“The original idea,” Long said in a 1996 interview — i.e., close to the events in question — “was to take an Ultima and just make it so you could have a party of people travel together.”) A third candidate is Rick Delashmit, the first and for a long time the only programmer assigned to the project, who unlike Garriott and Long had deep connections to the MUD scene, gaming’s closest extant equivalent to what Multima would become. In fact, Delashmit had actually co-founded a MUD of his own: LegendMUD, which was still running on a single 80486-based personal computer humming away under a desk in an Austin data center.

Regardless of who decided to do what, Long and Delashmit soon had cause to wonder whether signing onto Multima had been their worst career move ever. “We were kind of the bastard stepchild,” Long says. “No one got it and no one was really interested, because everyone wanted to build the next [single-player] Ultima or Wing Commander. Those were the sexy projects.” They found themselves relegated to the fifth floor of Origin’s Austin, Texas, headquarters, jammed into one corner of a space that was mostly being leased out to an unaffiliated advertising agency.

Making the best of it, Long and Delashmit hacked together a sort of prototype of a prototype. Stealing a trick and possibly some code from the last attempt at a Multima, they started with the now-archaic Ultima VI engine. “All you could do was run around and pick up a single object off of the ground,” says Long. “And then if another player ran into you, you would drop it. We had a scavenger hunt, where we hid a few objects around the map and then let the whole company loose to find them. Whoever was still holding them after an hour won.”

It was a start, but it still left a million questions unanswered. A game like this one — or rather a virtual world — must have a fundamentally different structure than a single-player Ultima, even for that matter than the tabletop RPG sessions that had inspired Richard Garriott’s most famous creations. Those games had all been predicated on you — or at most on you and a few of your best mates — being the unchallenged heroes of the piece, the sun around which everything else orbited. But you couldn’t fill an entire world with such heroes; somebody had to accept supporting roles. Some might even have to play evil rather than virtuous characters, upending the core message of Ultima since Ultima IV.

Garriott, Long, and Delashmit weren’t the only ones pondering these questions. In a 1993 issue of Sierra On-Line’s newsletter InterAction, that company’s head Ken Williams — a man who bore some surface similarities to Garriott, being another industry old-timer who had given up the details of game development for the big-picture view by this point in his career — asked how massively multi-player adventuring could possibly be made to function. He had good reason to ask: his own company’s Sierra Network was doing pioneering work in multiplayer gaming at the time, albeit as a closed dial-up service rather than on the open Internet.

I think a multiplayer adventure game is the next major step. Imagine a version of Police Quest, looking like it does now, except that your partner in the patrol car and the people in the street around you are real people. I think this would be cool.

For three months, Roberta [Williams, Ken’s wife], Chris [Williams, the couple’s son], and I have been arguing over how this would work. The problem is that most adventure games have some central quest story. Generally speaking, once you’ve solved the quest the game is over. You are there as the central character, and all of the other characters are there primarily to help move you towards completing your quest (or to get in your way).

A multiplayer adventure game would be a completely different animal. If 500 people were playing multiplayer King’s Quest at the same time, would there have to be 500 separate quests? There are also problems having to do with the fact that people aren’t always connected to the network. If my goal is to save you from an evil wizard, what do I do if neither you nor the evil wizard happen to sign on?

Here’s the thoughts we’ve had so far. What if we create a world that just contains nothing but forest as far as you can see? When you enter the game, you can do things like explore, or even build yourself a house. There’ll be stores where you can buy supplies. Soon, cities will form. People may want to build walls around their cities. Cities may want to bargain with each other for food. Or, for protection against common enemies. There needs to be some sense of purpose to the game. What if, after some amount of time in the land, the game “promotes” you to some status where your goals become to create the problems which affect the city, such as plagues, war, rampaging dragons, etc. In other words, some of the players are solving quests while others are creating them. Sooner or later, it becomes your turn to complicate the lives of others.

Either this is a remarkable case of parallel invention or someone was whispering in Ken Williams’s ear about MUDs. For the last paragraph above is an uncannily accurate description of how many of the latter were run, right down to dedicated and accomplished players being elevated to “wizard” status, with powers over the very nature of the virtual world itself. The general MUD design philosophy, which had been thoroughly tested and proven solid over the years, held that the creators of multiplayer worlds didn’t have to worry overmuch about the stories and quests and goals that were needed for compelling single-player games. It was enough simply to put a bunch of people together in an environment that contained the raw building blocks of such things. They’d do the rest for themselves; they’d find their own ways to have fun.

In our next article, we’ll see what resulted from Ken Williams’s ideas about massively-multiplayer adventure. For now, though, let us return to our friends at Origin Systems and find out how they attempted to resolve the quandary of what people should actually be doing in their virtual world.



In the immediate aftermath of that first online scavenger hunt, nobody there knew precisely what they ought to do next. It was at this point that Rick Delashmit offered up the best idea anyone had yet had. Multima would share many commonalities with textual MUDs, he said. So why not hire some more folks who knew those virtual worlds really, really well to help the Origin folks make theirs? He had two friends in mind in particular: a young couple named Raph and Kristen Koster, who had taken over the running of LegendMUD after he had stepped down.

The Kosters were graduate students in Alabama, meaning that much of the negotiation had to be conducted by telephone and computer. Raph was finishing up his program in poetry at the University of Alabama, while Kristen studied economics — a perfect pair of perspectives from which to tackle the art and science of building a new virtual world, as it happened. Raph Koster:

We started to get interview questions remotely as they worked to figure out whether we were qualified. All of the questions came from what was very much a single-player slant. We were sent examples of code and asked if we could write code like that and find the bugs that had been intentionally inserted into the code. It was something like a misplaced closing brace.

Finally, we were asked for a sample quest. I sent them the entire code and text of my quest for the Beowulf zone in LegendMUD, but then also said, “But we wouldn’t do quests this way at all now.” Kristen and I had been working on a simulation system based on artificial life and economic theory. We basically sent in a description of our design work on that. It was a system that was supposed to simulate the behavior of creatures and NPCs [non-player characters] using a simplistic version of Maslow’s Hierarchy of Needs, with abstract properties behind everything.

The Kosters were hired already in the spring of 1995, but couldn’t start on a full-time basis until September 1. Over the course of several drives between Tuscaloosa and Austin and back in the interim, they hashed out their vision for Multima.

They wanted to lean heavily into simulation, trusting that, if they did it right, quests and all the rest would arise organically from the state of the world rather than needing to be hard-coded. Objects in the world would be bundles of abstract qualities which interacted with one another in pre-defined ways; take an object with the quality of “wood” and touch it to one with the quality of “fire,” and exactly what you expected to happen would occur. This reliance on qualities would make it quick and easy for people — designers or just ordinary players — to create new objects on the fly and have them act believably.

But the system could go much further. The non-player creatures which populated the world would have different sorts of qualities attached, derived, as Raph Koster notes above, from Maslow’s Hierarchy of Needs. Each creature would have its own types of Food and Shelter that it sought to attain and retain, alongside a more elevated set of Desires to pursue and Aversions to avoid once it had served for its most basic needs.

An early preview of the game, written by Paul Schuytema for Computer Gaming World magazine, describes how the system would ideally work in action, how it would obviate the need for a human designer to generate thousands of set-piece quests to keep players busy.

Designing a dynamic world is a tricky business. How do you create enough quests to interest 2000 people? Origin’s solution: don’t. Create a world with enough logical conditions that it will generate its own quests. For example, consider a cave in the virtual world. Any self-respecting cave needs a monster, so you assign the cave a “need for monster” request of the game-world engine. Poof! A monster, let’s say a dragon, is then spontaneously created in the cave. Dragons are big eaters, so the dragon sends out a “need for meat” request. Meat, in the form of deer, roams the forest outside the cave, so the dragon’s life consists of leaving the cave to consume deer. If something happens that lowers the deer population (bad weather, over-hunting, or a game administrator strategically killing off deer), the dragon will have to widen its search for meat, which might lead it to the sheep pastures outside town.

At this point, NPCs can be useful to tell real players about a ravenous dragon roaming the countryside. NPCs can also sweeten the pot by offering rewards to anyone who can slay it. But what are they going to say?

“All of the [NPCs’] conversations come from a dynamic conversation pool which is linked to the world state,” explains Long. This means that if you, as a player, were to walk into this town, any one of the villagers will say something like, “Hey, we need your help. This dragon is eating all our sheep.” Thus, the “Kill the Dragon” quest is underway — succeed, and the villagers will handsomely reward you.

If you put it all together correctly, you would end up with a living world, to an extent that even the most ambitious textual MUDs had barely approached to that point. Everything that was needed to make it a game as well as a virtual space would come about of its own accord once you let real humans start to run around in it and interfere with the lives of the algorithmically guided dragons and guards and shopkeepers. It would be all the more fun for being true to itself, a complex dynamic system responding to itself as well as to thousands of human inputs that were all happening at once, a far cry from the typical single-player Ultima where nothing much happened unless you made it happen. Incidentally but not insignificantly, it would also be a sociologist’s dream, a fascinating study in mass human psychology, perhaps even a laboratory for insights into the evolution of real-world human civilizations.

When they arrived for their first day on the job, the Kosters were rather shocked to find that their new colleagues had already begun to implement the concepts which the couple had informally shared with them during the interview process and afterward. It was disconcerting on the one hand — “It freaked us out because we had assumed there was already a game design,” says Raph — but bracing on the other. This pair of unproven industry neophytes, with little on their résumés in games beyond hobbyist experimentation in the obscure culture of MUDs, was in a position to decide exactly what Multima would become. It would take a year or more for Raph Koster to be given the official title of “Creative Lead”; when he arrived, that title belonged to one Andrew Morris, a veteran of Ultima VII and VIII who had recently been transferred to the project, where he would struggle, like so many others at Origin, to wrap his head around the paradigm shift from single-player to multiplayer gaming. But for all intents and purposes, Raph and Kristen Koster filled that role from the start. Each was paid $25,000 per year for building this brave new world. They wouldn’t have traded it for a job that paid ten times the salary. At their urging, Origin hired several other MUDders to work with them on what was now to be called Ultima On-Line; the game would soon loose the hyphen to arrive at its final name.

I can’t emphasize enough what a leap into uncharted territory this was for Origin, as indeed it would have been for any other games studio of the time. To put matters in perspective, consider that Origin began working on Ultima Online before it even had a static website up on the Internet. The company had always been in the business of making packaged goods, games on disc that were sold once for a one-time price. Barring a patch or two in the worst cases, these games’ developers could wash their hands of them and move on once they were finished. But Ultima Online was to be something entirely different, a “game” — if that word even still applied — that was run as a service. How should people pay for the privilege of playing it? (Origin had no experience with billing its customers monthly.) And how much should they pay? (While everyone on the team could agree that Ultima Online needed to be appreciably cheaper than the games on CompuServe and its ilk, no one could say for sure where the sweet spot lay.) In short, the practical logistics of Ultima Online would have been a major challenge even if the core team of less than a dozen mostly green youngsters — median age about 22 — hadn’t also been trying to create a new world out of whole cloth, with only partially applicable precedents.

The early Multima team, or the “MUDders of Invention,” as they sometimes liked to call themselves. Clockwise from top left: Rick Delashmit, Edmond Mainfelder, Clay Hoffman, Starr Long, Micael Priest, Andrew Morris, Scott Phillips, Kristen Koster, and Raph Koster.

When the advertising agency that had been the team’s neighbor moved out, Origin decided to gut the now almost empty fifth floor where the latter was still ensconced. “They literally knocked out all the walls and the windows,” says Raph Koster. “So, the elevator surfaced to bare concrete. If you turned right, you went to see our team in a converted hallway, the only part left standing. If you turned left, you fell five floors to your death.” It was freezing cold up there in the winter, and the dust was thick in the air all year round, such that they learned to wrap their computers in plastic when they left for the day. “This team was given the least possible support you can imagine,” says Richard Garriott.

Undaunted, the little group did what the best game developers have always done when cast onto uncharted waters: they made something to start with, then tested it, then iterated, then tested some more, ad nauseum. Garriott came by “once a month or so,” in the recollection of Raph Koster, to observe and comment on their progress. Otherwise, their ostensible colleagues at Origin seemed hardly to know that they existed. “We were punk kids doing stuff in the attic,” says Raph, “and our parents had no idea what we were up to.”

In January of 1996, they put a message up on the Origin Systems website (yes, one had been built by this point): “Ultima Online (working title, subject to change) is now taking applications from people interested in play-testing a pre-alpha version, starting on March 1, 1996.” Almost 1500 people asked to take part in a test which only had space for a few dozen.

The test began a month behind schedule, but it proved that the core technology could work when it was finally run from April 1 until April 8. “This will be damn awesome when it comes!” enthused one breathless participant over chat. Richard Garriott thrilled the group by appearing in the world as his alter ego one evening. “We ran into LB! The REAL THING!” crowed a starstruck tester. “So what? I went adventuring with him to kill monsters!” scoffed another.

A more ambitious alpha test began about a month later, with participants logging in from as far away as Brazil and Taiwan. The developers waited with bated breath to see what they would do in this latest, far more feature-complete version of the virtual world. What they actually did do was an eye-opener, an early warning about some of the problems that would continue to dog the game and eventually, in the estimation of Raph Koster, cost it half its user base. Koster:

On the very first day of the alpha, players appeared in the tavern in Britain [the capital city of Ultima’s world of Britannia]. They waited for other newbies to walk out the doors, then they stood at the windows and shot arrows at them until they died. The newbies couldn’t shoot back because the archers were hidden behind the wall inside the building. This is, of course, exactly what archery slits were invented for, but it made for a truly lousy first impression of the game.

That was the first time that players took something inherent in the simulation and turned it against their fellow players. Not long after, someone spelled out “fuck” on the main bridge into Britain using fish. Not long after that, players figured out that explosive potions detonated other potions. It was illegal to kill someone in town; it would get you instantly killed by the guards. But if you ran a chain of potions a long distance, like a fuse, you could avoid that.

Chalk it up as another lesson, one to which the team perhaps should have paid more attention. The hope was always that the community would learn to police itself much like those of the real world, that the risks and the stigma of being a troublemaker would come to outweigh the gains and the jollies for all but the most hardened members of the criminal element. Richard Garriott seems to have been in an unusually candid mood when he expressed the hope in a contemporary interview:

One of the unfortunate side effects of computer gaming is that we have a whole generation of kids who have no social graces whatsoever. And this is exemplified in my mind by how much I hate going online for discussions I am invited to do once a month or so. And I truly abhor going to do that for a variety of reasons, one because it’s such a slow experience, but also because everyone has a level of anonymity behind their online persona, [and] they lose their normal and proper etiquette. So you have people screaming over each other to get their questions in, people screaming expletives, people popping in a chat room and making some dumb comment and then popping out again — the kind of behavior that you would never get away with in the real world. So one of the things that I’m really keen to introduce with Ultima Online is making people responsible for their actions, and this will happen as people are recognized by their online persona within the game. They won’t be so anonymous anymore.

This hope would be imperfectly realized at best. “I used to think that you could reform bad apples and argue with hard cases,” says an older and presumably wiser Raph Koster. “I’m more cynical these days.”

So, the alpha testers ran around breaking stuff and making stuff, while the gods of the new fantasy world watched closely from their perch on a gutted floor of an anonymous-looking office building in flesh-and-blood late-twentieth-century Texas to see what it was they had wrought. A bug in the game caused killed characters to respawn naked. (No, not even these simulation-committed developers were cruel enough to make death permanent…) Some players decided they liked it that way, leading to much of Britannia looking like a nudist colony. The guards in the cities had to be instructed to crack down on these victimless criminals as well as thieves and murderers.

But the most important finding was that most of the players loved it. All of it. “We had people who literally did not log off for a week,” says Starr Long. “Groups of players formed tribes, and at the end of the test there was a huge battle between the two largest tribes. None of that we set up. We gave them the world to play in, and human nature took over.” From one of the first substantial previews of the work-in-progress, published in Next Generation magazine:

In a very real sense, the world is what you make of it. One of the more interesting results of Ultima’s alpha testing is that when you have several hundred people in one place at one time, they tend to form their own micro-societies. There are already some two dozen player-created “guilds.” For example, when Richard Garriott signed on as his alter-ego Lord British, two groups sprang up: the Dragon Liberation Front, which pledged itself to destroying him, and the aptly named Protectors of Virtue & Lord British. Threats were made, battles were joined, and a fine time was had by all.

Just as gratifyingly, reports of the alpha test, in the form of previews like the above in many of the glossy gaming magazines, were received very positively by ordinary gamers who hadn’t known that a massively multiplayer Ultima was in the offing prior to this point. Some went so far as to form and run guilds through email, so that they would be ready to go when the day came that they could actually log in. There was still much work to be done before Ultima Online would be ready to welcome the unwashed masses inside, but the MUDders who were making it believed that all of the arrows were pointing in the right direction.

Until, that is, September 27, 1996, when it suddenly seemed that its thunder might have been well and truly stolen. On that date, The 3DO Company published something called Meridian 59, which purported to already be what Ultima Online intended to be: a massively-multiplayer persistent virtual fantasy world. Its box copy trumpeted that hundreds of people would be able to adventure together at the same time, killing monsters and occasionally each other, whilst also chatting and socializing. It sounded great — or terrible, if you happened to be working on Ultima Online.



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; Questbusters of July 1987; Sierra’s customer newsletter InterAction of Summer 1993; Computer Gaming World of October 1995 and October 1996; Starlog of December 1983; PC Powerplay of November 1996; Next Generation of June 1996, September 1996, and March 1997; Origin’s internal newsletter Point of Origin of November 3 1995, January 12 1996, and April 5 1996.

Web sources include a 2018 Game Developers Conference talk by some of the Ultima Online principals, an Ultima Online timeline at UOGuide, and some details about LegendMUD, the winner of MUD Connector‘s coveted “MUD of the Month” prize for October 1995.

Footnotes

Footnotes
1 Not the same game as the 2002 Bioware CRPG of the same name.
 
 

Tags: , , , , ,