By Bill Mahin
In mid-November 1998, Bungie Software had more than 140,000 orders for its latest computer game, Myth II: Soulblighter. The team that was developing it had been working for almost a year, and they were obsessed with making it a better game than Myth: The Fallen Lords, which Computer Gaming World and PC Gamer, the two biggest magazines in the gaming industry, had called the best strategy game of 1997.
They wanted to have Myth II in the stores before Christmas, and to do that they had to have it finished at least two weeks before. They’d already missed their first target shipping date, November 1.
On November 13 sheets of paper covered one wall of the room where they were working, each one listing bugs that had to be fixed and changes that had to be made. There were a lot of pages, but Jason Regier, the lead programmer, said, “It seemed there was an end in sight.”
A lot of the staff looked exhausted, especially Regier and Mark Bernal, one of the artists. “There was a point near the end of summer, early fall,” said Bernal, “where it was just very difficult–the point where you basically can’t get away from it, where you dream of it.” When he would come to work in the morning, he’d look up at the sky “knowing that those few minutes would be all I’d see that day.” He figured he was “running at 70 percent capability. You find yourself making mistakes–your mind is not thinking as it used to.” He made up for it by working even longer hours.
One of the newer designers had dreamed about the game when he’d gone home the previous afternoon and collapsed. He’d dreamed he was one of the characters in the game, one of the Ghasts–the undead peasants. What was horrible, he said, was having to pick maggots out of his skin.
Some of the later levels in the game were still being developed. David Bowman was still working on level 16, the Ibis Crown. He’d finished the color maps that indicated to the artists where things such as grass, rocks, and trees should be, and he’d finished the elevation maps, which delineated the contours of the terrain. But he was still writing code for the action that would take place in the environment.
The setting for the level was a series of catacombs, one of the few underground environments in the game. The consensus of those who’d played the level in its early configurations had been that it was slow to get through and boring, so the team had decided to cut the size of the playing area by a third and change the action. The enemy, said Bowman, had been transformed into “ghosts from a long time past that have basically been hiding in the catacombs forever.” He wanted to make them capable of destroying the player’s forces even when they couldn’t be seen. He also wanted to “establish a feel for the level that’s hopefully going to get the player nervous. ‘This is different. Didn’t happen anywhere else in the game. Didn’t see it in Myth I.'”
By November 16 he’d been working on the level for three days straight. He hadn’t seen his 13-month-old daughter in a week, though he had taken the previous Saturday off to baby-sit so his wife could get out. Now, he said, “I’m working till I’m done.”
On November 21 the staff spent a long afternoon in a meeting. One of the agenda items was evaluating the level known as Down a Broken Path, where the ostensibly simple goal was having the player’s troops escort Rurik, a village leader, safely into town. Jason Jones, cofounder of Bungie, wasn’t happy with the way Rurik moved through the terrain, because he looked like “a complete fucking moron.”
Eventually the discussion turned to whether they were making the game too easy in response to criticism that Bungie games in general were too difficult. “It’s not about making it easy,” insisted Jones. “It’s about not making it discouraging.”
Then they debated whether to keep to a new deadline or to incorporate any and all changes that might make the game better. Tuncer Deniz, the project manager, argued that at some point they’d have to agree that if there was a problem that made the game crash they’d fix it, but that not every detail that enhanced the game could be included.
Jim Ruiz, a member of Bungie’s marketing team, complained about what he saw as a gratuitous loss of troops as a player advanced up the levels of the game. “You get attached to your troops after seven levels,” he said. “By then they’re all vets–it’s painful to throw them away.”
Deniz left the meeting with the responsibility of going back to the artists, programmers, and designers on the team and telling them all the changes they would have to make. “The nit-picking is almost getting out of hand,” he said. “I feel bad.”
Six people worked straight through that night, including Bowman, who now held the Myth II work record–he hadn’t left the office for eight days in a row. “My wife and daughter visited yesterday,” he said. After she left he tried to send her flowers, but the order got screwed up.
Bungie Software had its genesis in 1990 in the basement apartment of Alex Seropian, a University of Chicago undergraduate math major from Westchester, New York. He’d developed a computer game he called Gnop!, a knockoff of Pong, the black-and-white game that had kicked off the electronic-gaming industry in the early 70s. Like Pong, it had two rectangles (the paddles) that moved vertically and a black dot (the ball) that moved horizontally as it was hit by a player back and forth across the screen. Its most notable feature was the sound of Seropian’s chuckle, which was triggered when a player lost.
Seropian put the game on the Internet as shareware under the label Bungie and never made a cent from it. By the time he graduated in May 1991 he’d developed Operation Desert Storm, a tank game based in part on action in the real war that he hawked at trade shows because he couldn’t get stores to pick it up. He sold more than 2,000 copies.
Seropian had met Jason Jones in an artificial-intelligence class at the U. of C., and one day he stopped by Jones’s dorm room and saw him playing a game he’d created just for fun. Seropian persuaded him to clean up the game and sell it.
Jones dropped out of college to become Seropian’s partner. He’d been obsessed with creating computer games since he was in fifth grade. When the recess bell would ring, he says, “I’d race the other four computer people to the three computers we had. All the other kids went outside.” He knew how to draw images on the screen but couldn’t figure out how to animate them. He finally discovered the secret–program the computer to draw a picture in one place, erase it, and then redraw it in a slightly different place. “I was psyched–I was shaking when I realized I’d figured it out.”
Jones’s cleaned-up game, Minotaur, which came out in February 1992, was Bungie’s first multiplayer game; it allowed several players to hook their computers together and play one another instead of the computer–a rarity at the time. They took the game to trade shows and figured out how to market it through mail-order catalogs, selling about 2,500 copies.
They then began working on the game that would become Pathways Into Darkness, in which players, using a keyboard or mouse, rush down stone hallways shooting monsters and picking up additional weaponry, ammunition, and magic crystals as they try to stay alive long enough to get through all 25 environments of the game and destroy the “Dreaming God” buried at the bottom of a pyramid, thereby saving the world. Both men had used Macintosh computers for years, so that’s what they worked with as they painstakingly laid out the hallways, doors, pillars, walls, etc of the game environment and then wrote myriad lines of code that allowed the computer to display any action a player took as well as respond to those actions. For example, they decided to put a Banshee behind one pillar. If a player entered an invisible section of the floor, the Banshee would attack–which meant they had to write code so that it would attack the player wherever he entered the section, as well as code dictating how fast the Banshee would move, how it would attack, how far it could move in any direction. When the Banshee charged, the computer also had to be programmed to evaluate the strength of the player, then essentially flip a coin to determine whether the Banshee or the player would be wounded. And the computer had to be programmed to continuously redraw the game world–30 times per second–updating it in response to the player’s slightest move.
Pathways was released in August 1993. It proved addictive, and it got great reviews. Macworld gave it an award for best role-playing game of the year and said, “Welcome to the state of the art.” The Inside Mac Games reviewer wrote, “Many games boast that they actually get your heart pounding faster, only to fall short of expectations and marketing hype. Pathways delivers.”
But it was difficult to play. Its 25 levels–plus a rumored secret one–made it so complex that few people were able to finish it. In an attempt to help out frustrated gamers, Bungie published the Pathways Into Darkness Official Hint Book, which contained suggestions such as “Learn how to use the knife. In the deeper underground levels, using it will give you about the same chance as a dyspeptic, tone-deaf rhinoceros has of performing Carl Orff’s Carmina Burana on a kazoo. But in the pyramid [an early, easy level], the knife can save you quite a lot of ammunition.”
Sales, primarily through mail-order catalogs, were modest by industry standards–roughly 20,000 copies–but they persuaded Jones and Seropian to move Bungie into a real office in Pilsen and to begin hiring other artists and programmers.
In May 1994 Doug Zartman, a PR person for the American Bar Foundation and a percussionist-vocalist for the band Spiney Norman, showed up, portfolio in hand, wanting to be an artist. He’d been a gamer for much of his life, though he’d majored in sociology and anthropology at Carleton College. Seropian and Jones hired him–their first full-time employee–to do tech support for Pathways. But he soon became its PR person–its “mouthpiece,” according to the current in-house directory.
Zartman helped nurture the jaded, attitude-laden, often sophomoric humor that had been an integral component of Bungie’s corporate identity from the beginning. His response to a letter from a Pathways fan about the origin of the sound made by a Skitter, which Zartman describes as a “pretty terrifying creature that spits poison and has a unique cry,” is typical: “It was not easy to digitize the sound of the dying Skitter. We actually had to go out and find a Skitter, bring it to the office, kill it, and make a sample of its screams. In doing so we violated several state and federal laws, but we think it was worth it. (If you don’t buy that, how about this: ‘The sound is a combination of electronic and computer-generated waveform technologies.’ Take your pick.)”
Earlier that spring the Bungie workforce, including a handful of contract workers, had begun working on a successor to Pathways, Marathon. “We were all young,” says Greg Kirkpatrick, a U. of C. grad who designed some of the game maps and helped develop the story. “I was 22. Jason was maybe 23 or 24, Alex a couple years older. We were all kind of winging it at the time–a bunch of young guys who didn’t know what the hell they were doing.”
But they spent a lot of hours doing it. “A minute doing anything other than working on Marathon was considered a wasted minute,” Matt Soell, who was hired in 1995 to take over tech support, later wrote in the Marathon Scrapbook, the story of the game’s development. “So activities like sleep and lunch were carefully kept to a minimum.”
“It was just a lot of long hours–really dedicated, nonstop working,” says Kirkpatrick. “Six months, probably like 14 hours a day, seven days a week. Pretty much your stereotypical vision of a small computer-game company–eating a lot of pizza, drinking a lot of Coke.” He adds, “Working that much with a small group of people, we did drive each other crazy. But there was the idea of doing something that hadn’t been done before–that was really the excitement of it. In some ways it didn’t matter if we didn’t get along well.”
Like Pathways, Marathon was a first-person-shooter game, with a hand–ostensibly the player’s–visible at the bottom of the screen and everything else seen from that single perspective. In carrying out the player’s mission–conquering evil beings, rescuing civilians, saving the world–the hand fired a variety of powerful weapons, including a .44 magnum mega class A1 sidearm, an MA-75 battle rifle with integral 40mm grenade launcher, a Zeus-class fusion pistol, and a TOZT-7 backpack napalm unit. The hand could also punch, a low-tech, last-ditch option.
Yet Marathon was a much faster game than Pathways, and it had higher-resolution graphics and better effects. It even had a story line, involving an interstellar colony ship that had been taken over by aliens. The player was the colony’s only hope, but the odds weren’t good. At the beginning of the game you’d be told that your shuttle had been destroyed by a berserk computer before it landed and you’d barely managed to reach the ship. The rest of the story had to be figured out by gathering information from computer terminals scattered throughout the ship–though it was certainly possible for players to ignore the background story and simply try to move through the 26 levels of the game, doggedly punching, shooting, or otherwise eliminating any enemy who appeared.
Detailed background stories and characters were unusual at the time. Zartman says, “Id Software, the makers of [the incredibly successful] Doom, became famous for denigrating stories in their games. ‘Visceral adrenaline rush is all you need–stories are for saps.'”
Bungie’s game manual even told the story of the original Battle of Marathon in 490 BC, in which the Greeks’ General Multiades used speed and strategy to defeat the vastly larger Persian army. The manual stated, “Lessons applicable to the game: move fast, seize the initiative, wield superior firepower, dive into the melee, anticipate enemy movements, slaughter the defenseless, endure.”
In August 1994 Bungie took a demo version of Marathon to the Macworld Expo in Boston. Few people had heard of it, but word spread quickly. By the last day, says Zartman, the Bungie booth “was mobbed.” Stacks of order forms disappeared, while “journalists swarmed around like sharks in the chum.”
They’d taken with them T-shirts that had a single line of copy on the back: “Don’t make us kick your ass.” Zartman explains, “Solitaire you might lose, but it won’t kick your ass. Whereas in Marathon the alien launches a rocket at your feet, and your body starts spewing blood everywhere, and you’re launched 50 feet into the air–you have a very tangible feeling that you’ve gotten your ass kicked.” They sold hundreds.
After they came back, interest in the game only grew more intense.That fall two early test versions of the game were stolen and posted illegally on the Internet.
People who’d filled out orders at the Boston expo had been told that the shipping date for the finished version would be in two weeks. But as the few remaining changes were incorporated and the last bugs eliminated, the Bungie staff discovered more bugs. And in correcting them, they found new ones. The game wasn’t finished until December 14. “Over the previous four days,” Soell wrote in the scrapbook, “the Bungie team had slept less than ten hours–their last push after months of 14-hour days.”
It was too late to ship most of the orders before Christmas, though on December 23 Seropian, Zartman, and one of the artists spent the day working on the assembly line at the small Aurora warehouse where most of the game components were manufactured and assembled so that at least 500 of the people who’d ordered overnight delivery would get the game in time.
At the Macworld Expo in San Francisco in January 1995, so many people wanted a copy that the Bungie booth, Zartman says, was “pandemonium, with Bungie employees scooping up a handful of cash, handing over a game box, stuffing the cash into their back pockets, then reaching back for another copy–rep after rep, a continuous rhythmic motion.” By midday “our back pockets were bulging with thousands of dollars in 20s, 50s, some crisp hundreds, and untold numbers of frayed fives and tens.”
A retailer from Malaysia bought ten copies for $400, but his check later bounced. “What would have been a painful financial loss before,” says Zartman, “was now shrugged off–a drop in the bucket.” The orders didn’t stop when they got back to Chicago. “Marathon was enormous,” he says, “a huge phenomenon for the Macintosh.”
The volume of orders forced Bungie to change the way it handled them. “We’d had this system,” says Zartman. “The customer would fax or mail or call in an order. I’d fill out a sheet by hand. Then I would laboriously type it into a database, one order at a time.” He would then fax the information to the company that was handling the packaging and distribution. “That worked fine for Pathways–you could handle maybe 15, 20 orders a day.” But not tens of thousands of preorders for Marathon. Bungie turned the entire ordering process over to a commercial company–a huge shift given their do-it-all-in-house philosophy. “Back when I started,” says Jason Jones, “our whole attitude was do it all ourselves–reinvent the wheel 10,000 times, discover fire every week. We had almost no respect for experience.”
The success of Marathon brought them even more attention. “Previously we were delighted to have a little paragraph in a little Mac newsletter,” says Zartman. Now Wired magazine was calling. In the February issue a reviewer wrote, “Marathon comes on like a relentless, fuel-injected nightmare. For better or worse, I’m hopelessly addicted to this game–and I’m an adult.” A MacUser reviewer wrote, “It’s sick, it’s twisted, it’s gory, and I love it to death. Marathon is the most spectacular Mac game in years….Marathon is nothing less than awesome.”
Marathon also won major awards. Macworld magazine enshrined it in its Game Hall of Fame. MacUser called it the best action game of 1995, and the Mac Home Journal called it one of the top ten Mac games of all time.
In May 1995 Mark Bernal was hired to work on the sequel, Marathon 2: Durandal. He’d heard of Bungie through a newspaper ad the year before. “Basically it was asking for an artist who wanted to make kick-ass games. ‘Yeah, that’s what I want to do.'” He was interviewed, but then heard nothing. Finally he ran into Seropian and asked if he still needed an artist. Seropian said he did.
Initially only Bernal and another artist, Robert McLees, worked on Marathon 2, even though they were supposed to finish in less than a year. In theory the game would take less time because it was a sequel. “The code was already worked out to a degree,” says Bernal, “there were less bugs, and the artwork was already set.”
But, Zartman says, “Sequels are different in computer gaming than in the movie business. Often the sequel is better than its predecessor, because the technology has improved and the company is able to put in a lot of elements they weren’t able to include in the original version. There were a lot of cool things we’d thought up that couldn’t go in the original version because we had to ship the damn thing.” Now they could add sound effects for wind and bubbling rivers, a broader range of lighting effects, a wider field of view, and several new friendly characters to help the player.
Hours of work were sometimes wasted. Zartman, who’d created game maps before, designed one of the levels, Ex Cathedra. He based the architecture on plans he’d found for Durham cathedral in England, and he flooded part of the level, which forced the player’s character to wade. Then he designed “these elaborate traps, like pits that you couldn’t see under the surface of the water that you’d fall into.” But by the time he’d perfected the level, the programmers had come up with a way to enable the character to swim. Zartman’s submerged traps were now irrelevant, though they weren’t removed from the final version of the game and are still there.
On November 24 Bungie released Marathon 2. A Computer Gaming World review said, “Far from treading over the same old ground, Marathon 2 blazes new trails, once again raising the standard of 3-D action games on the Macintosh.” The game won some awards, though not as many as Marathon. Zartman says that’s because editors tend not to give awards to sequels, regardless of their quality.
Marathon 2 sold many more copies than the original, though like Pathways, it was difficult to master, even for expert players. Bungie eventually decided it had to publish a strategy guide, which was written by Tuncer Deniz.
Deniz had entered the digital world relatively late, when his father bought him a Mac for college. After graduating with a marketing degree from Indiana University in 1991, he’d started his own digital magazine, Inside Mac Games, which is how he met Seropian and Jones. “They were just coming out with Pathways,” he says. “They came over to my house and showed it to me.” He previewed the game in a subsequent issue of his magazine. Three years later, “bored working at home,” he went to work for Bungie.
The Marathon 2 guide was a fast-accessing digital document that gave a breakdown of each of the 28 levels in the game as well as playing tips. “While it will not walk you through each level step by detailed step,” the introduction states, “it will provide you with crucial information and hints that should get you past the trickiest spots.”
The guide makes it clear how difficult the game is. For example, on one level it’s important that the player locate a hidden shotgun. The guide explains how to gain access to the room where it’s concealed: “After turning right from the first terminal there’s a room to drop down into. This room is bounded by four raised pools. Stand on the ledge and look at the pool at the top left–it has a platform hidden within. On the far wall opposite there’s a secret door. The catch is to get up into the pool and onto the platform. Once on the platform the secret door can be opened and a passageway into the Shadow Room can be explored.”
At the end of March 1996 Bungie announced that it would release a PC version of Marathon 2. Many Mac gamers, who’d had the Bungie games to themselves, felt betrayed. Within a couple days Bungie received 400 angry E-mail messages. One of the worst stated, “You fucking assholes have gone and done it now. You said you wouldn’t create Marathon for PC. Fucking liars! What have I to hold above the PC gamer’s heads now, huh you fuckers?!?! Bitch trick whore slut ass fuckers!!!! BURN IN HELL ASSHOLES!!!!!!!! You know, I was actually gonna buy the Marathon/Marathon 2 bundle. But not now! You assholes ain’t getting any of my money. I’ll fucking use a copy from a friend.”
The Bungie staff knew they were risking alienating some of their loyal customers, though Zartman says they never promised not to go PC. But they also knew that the potential PC market was huge. Selling 40,000 copies of a Mac game would be considered a success. Selling 40,000 copies of a PC game would be considered a disappointment. Last July 340,000 Mac-format games were sold, but more than ten times that many PC games were sold; Mac-game sales were nearly $6 million, PC-game sales were just under $74 million. The disparity is so great that most retailers refuse to stock Mac games, forcing Mac developers to sell their games through mail-order houses or over the Internet.
Facing a potentially huge jump in sales, Seropian knew he needed to find distributors that could get Bungie games into the big chains. “They have clout where you don’t,” says Zartman. “If a little company calls Best Buy, nothing’s gonna happen.” But distribution, Seropian later noted on the Bungie Web page, is “a really sucky business.” He complained that distributors can take a long time to evaluate a game, and if they offer you a contract you have to guarantee that the distributor will get “a better price than anyone else in the world,” agree to take back any copies that don’t sell and provide a full refund, and pay a marketing fee, which ranges from 3 to 6 percent of sales. On top of that you might have to pay a “spiff,” an additional percentage of the sale price of every unit individual stores sell, which theoretically encourages salespeople to push your game instead of the competitors’. He concluded, “Bungie ends up getting the shaft like the rest of the software developers, and bends over for the contract.”
Moreover, he added, “They never pay until they need more product.” He likes to tell the story about the time Bungie sold a lot of copies of Marathon to a distributor, which sold several thousand of them to a retailer, who soon declared bankruptcy. “Our friendly partner in sales–our distributor–said to the now-bankrupt retailer: ‘Sell it back to us for $2 a unit, and we’ll take them off your hands.'” That’s a standard business practice, but then the distributor returned the copies to Bungie as unsold inventory and demanded a full refund. Only by coincidence did Bungie learn what had happened. They threatened never to sell any games to the distributor again, and it backed down.
Meanwhile part of the Bungie staff had been trying to devise a new game idea. Jones, who’d come up with most of the basic concepts for Bungie games, seemed to have some ideas. “We didn’t quite know what he planned,” says Bernal. “One day he mentioned, ‘What do you think about having this world with 100 guys fighting 100 other guys in 3-D?’ And we’re like, ‘Huh? Yeah, I guess so. That might be cool.'”
In the fall of 1996 Jason Regier joined the development team for the new game. He’d been writing software for Qualcomm, a cellular-phone company in San Diego, and designing computer games in his spare time, one of which, Amoebarena, had been picked up by a distributor. Qualcomm, he says, “was very corporate. I felt like a cog in a big wheel. It was a lot like the world of Dilbert.” In contrast, Bungie “was like a frat house. You felt part of something big–that the success of the company depended on how much work you put into the project.”
In November, after working in isolation for nearly a year, the team had a demo of Myth: The Fallen Lords ready. Zartman took it on the road and showed it to the major game magazines. “There wasn’t much game play in what we showed them–two small groups on opposite sides of a small map rushing at each other, becoming a bloody knot at the middle where they all collided and blew each other up.” And the AI–artificial intelligence, the specific code that controls the actions of the opposing characters–was crude. But the reviewers could see that it had “strategic combat on real 3-D terrain”–something no other game at the time had, something that radically changed how the game was played. “An archer could fire farther from the top of the hill than from in the valley,” he says. “It was easy to demonstrate to the press that an archer on high ground was going to defeat an archer down in the valley.”
Myth was radically different from anything Bungie had done before. The biggest change was that where Pathways and Marathon had given the player the point of view of and control over one character, Myth gave the player the ability to oversee a whole battlefield and control many different characters. On the player’s side–“the Light”–were characters such as Warriors, Berserkers, Archers, and Dwarves. “Dwarves throw small bottles filled with an unstable concoction which explodes powerfully upon impact,” said the game manual. “Most of the time, anyway.” Berserkers: “Their bare, battle-scarred flesh doesn’t protect them much, but the speed and ferocity of the Berserkers’ attack gives them a definite edge.”
The opponents–“the Dark”–included Ghols. “Although they are relatively weak, they can dash up a hill and hack a group of archers to pieces before they have a chance to react.” And Wights: “A stitched-up corpse, given new life by dark magic as a breeding ground for virulent disease and foul decay. The Wight shambles up to its target and plunges a dagger into its gas-filled body, causing it to explode, shaking the earth around it, destroying anything in its immediate vicinity and coating everything with a thin film of pus, which induces a brief paralysis.”
The story driving the action had become an integral part of the game, rather than merely background information to be assimilated or ignored at the player’s pleasure. “We wanted to avoid a lot of the cliches that are rife in these types of stories, most of which are derivative of the Tolkien story line,” says Zartman. “We wanted to create something that wasn’t as fuzzy and cuddly–something dark, a story we’d be proud of.”
One of their models was a noirish soldier-of-fortune fantasy novel by Glen Cook called The Black Company. The narrator, the company’s doctor and archivist, has lines like “The company is our family….It is the company against the world. Thus it has been and ever will be….I am haunted by the clear knowledge that, in the end, evil always triumphs.”
“One of the things we liked was that Cook presents this pretty amoral world,” says Zartman. “One side is sort of the good guys, because the narrator is on that side, but they’re not on a moral high ground over their opponents. It’s not a simple good-evil dichotomy. We’re dealing with a sophisticated world here, with politics and treachery and betrayal from both sides–as much conflict from within the ranks as from the enemy.”
The prologue to the game manual states, “The war in the North is in its seventh year, and I grow tired of writing this record. Force of habit counts for something, but I’ve written of so many half-hearted assaults, so many retreats–why do I go on? Writing down every detail I could remember–the names of dead men and burning cities and the feeling of heat at our backs as we ran away, again–used to help me sleep at night. Now it’s just something to do between fighting and sleeping.”
Like many computer games, Myth includes blood and gore–and is therefore enmeshed in the controversy surrounding violent entertainment. After four young girls and a teacher were killed by two young boys in Jonesboro, Arkansas, in 1998, Time magazine said, “Few researchers bother any longer to dispute that bloodshed on TV and in the movies has an effect on the kids who witness it. Added to the mix now are video games, at least the ones built around the model of hunt and kill. Captivated by effects that are ever more graphic, game boys learn to associate gusts of ‘Blood’ with the primal gratification of scoring.” Shortly after that, Wired did a story on senators Joseph Lieberman of Connecticut and Herb Kohl of Wisconsin, who’d written a report in conjunction with the National Institute on Media and the Family that concluded, “Gory video games contribute to an atmosphere that desensitizes children to violence, leading to increasingly violent behavior and horrific events, like the recent rash of schoolyard shootings.” Lieberman stated, “Killing and carnage is not enough anymore. To torture and maim is often the name of these games now.”
There are some pretty gory games. Blood II: The Chosen, for example, is a first-person-shooter game in which the player basically fires at anything that materializes on-screen–scientists, lab techs, mutant enemies. If the player keeps firing into already dead victims, the corpses eventually explode, raining body parts and blood everywhere. The player then has the option of zooming in on excruciatingly rendered close-ups of legs, livers, and eyeballs.
But Zartman insists Bungie games are different. “The majority of our players are 18 to 26. Our games are more sophisticated than arcade games and most video games, where the way you attack is to get all your guys in a group and run into the enemy and hack ’em all to pieces. If you took that approach in Myth you’d lose quickly.” The key to winning in Myth, he says, is “using your troops strategically on the battlefield by moving them in formations, holding the high ground, flanking an enemy formation, covering the advance of your swordsmen with archer fire, and so on.”
He also says, “Like any type of entertainment, a successful game depends on the user suspending their disbelief. You want to draw the player into the game world, and one of the ways you do that is with realistic effects. In the case of a game of battle like Myth, blood on the battlefield is a part of the realism.”
And he adds, “Doom and Quake [two best-selling violent games] are huge games in Japan and all over Europe–worldwide–and yet these school shootings do not happen in other cultures like they do here. All this American media flowing into all these other countries around the world–the same games, the same movies–and you don’t have kids going out and shooting each other up. What does that tell you? It ain’t the games.”
Bungie released Myth in November 1997. In addition to the praise from Computer Gaming World and PC Gamer, it got accolades from Macworld magazine and Computer Games Strategy Plus, both of which called it the game of the year. Online Game Review said it was one of the top 50 games of all time. “Just coming from the Mac world into the PC world, that’s huge,” says Bernal. “Company comes out of nowhere and takes over that year–pretty amazing feat.”
Myth became Bungie’s best seller ever; 300,000 copies, at $40 each, have sold worldwide in nine languages–English, German, Japanese, French, Spanish, Portuguese, Chinese, Korean, and Italian. One reason for its success was that it was the first Bungie game that could be played over the Internet. People could now play with other–many other–gamers anywhere in the world, and that changed the whole nature of gaming. “Your opponents are humans, other players,” says Zartman. “They are much less predictable than AIs–hopefully. No matter how complex an AI is programmed, you’ll eventually find patterns in what it does and the limitations in how it reacts to you. Once you’ve found that out, it’s pretty easy to defeat them. That’s certainly not the case with a human player. There are still people playing networked games of Marathon 1 five years after it came out.”
Bungie has 100,000 registered Internet players and maintains a Web site–bungie.net–to support them. “Bungie.net is a perfect example of a thriving Internet community,” says Zartman. “There are lots of people there who are very good friends, and communicate with each other in real time every day, yet have never met in real life–or ‘meat space,’ as some excessively wired people call it.”
After five years in Pilsen, Bungie moved to a building on West Ontario in March 1998. The new space included two large, open rooms for the artists and programmers working on two new games, Myth II: Soulblighter and another game that still hasn’t been released. Futons and foldaway cots were tucked away in the rooms, and there was a kitchen with two refrigerators, one of which was crammed with soft drinks, a variety of juices, and occasionally beer. There was also a shower with soap, shaving gear, toothpaste, shampoo, towels, and a yellow rubber duck.
By this point Bungie had brought the ordering process back in-house and had set up a second office, in San Jose, California, to develop more games. Seropian and Jones would have preferred to expand the Chicago office, but the programmers they wanted were in California and didn’t want to move. Only a couple other computer game companies are headquartered here; most are in California or Texas.
Bungie was still both a developer and publisher–something few other computer game companies could claim. “In the rest of the industry,” says Zartman, “the situation is, small development company, say 30 people, makes the game and then cuts a deal with a large publishing house with thousands of employees. Once the game is handed off to the publisher it’s the publisher’s property. The developer gets maybe 20 percent of the profits, and they typically have no say in how their game is advertised, packaged, priced, translated into other languages, etc. It’s a frequently antagonistic relationship. For comparison, another company that both develops and publishes internally is LucasArts. They have the talent and the wealth of the Lucas empire backing them. When you go to their complex they have a stable of 50 to 70 artists working on one game, to our 4 or 5. Their tech-support department is the size of our whole office. The fact that our games are competitive with those of a company like LucasArts is amazing.” (In 1999 Seropian and Jones would sell 19.9 percent of Bungie to a major publisher, Take 2 Interactive, a percentage that allowed them to retain control of their product.)
In May 1998 Bungie announced that Myth II: Soulblighter would be released later that year. “Last year Bungie’s inaugural PC game crashed into the gaming industry like a tidal wave, washing aside traditional strategy fare,” said a preview in the October issue of Computer Gaming World. “From what we’ve seen and heard, Bungie’s work is going to significantly improve the gameplay of the original Myth. If so, look for Myth II to be one of the biggest fantasy strategy games in recent years.”
For the first six months of the game’s development Jason Regier had been the sole programmer. “It felt so good in some respects,” he says. “It was like a return to the original days of programming for me.” But the hours soon became horrendous. “Mark [Bernal] and I were in here all the time for as long as I can remember.”
Bernal says he’d decided in February that he’d have to practically live in the office. “This had to be my new home, basically 10 to 12 hours a day.” By April they were working seven days a week. “Myth II,” he says, “was my whole life.”
In trying to make Myth II a great game, says Bernal, “We added more types of animation to the characters, added fire to the game–a whole bunch of new enhancements. Three-dimensional models became important to game play instead of just scenery.” He proudly describes a drawbridge he designed that requires the player to sneak a soldier inside the castle walls, find the winch that controls the drawbridge, and destroy it–causing the bridge to lower and allowing the besieging army into the castle.
They hadn’t known how to make a moving model like that before. They were also making the AI much more sophisticated. “There are a whole list of actions that will allow you to create a more interesting response than ‘Do this, and this always happens,'” says David Bowman. “You can call for random action from those triggers, so the player gets a different response randomly from the list.”
When they’d missed their November 1 shipping date no one seemed especially upset. Project manager Tuncer Deniz says, “A lot of it had to do with the time frame given to do the game, which was very small as opposed to most major titles. We added so many new things–ultimately that ends up extending the development cycle.”
But missing that deadline meant they might have a cash flow problem. “The later we ship,” says Seropian, “the later we start getting our investment back.” At that point they were living primarily on the earnings from sales of Myth I, and computer games have a fairly short shelf life (though all of Bungie’s games are still available).
Worse, if Bungie couldn’t get Myth II in stores when it had promised, it might have to pay penalties to dealers and distributors. According to Seropian, those penalties can run as high as $250,000 for a retailer the size of CompUSA. And everyone knew that it would be harder to negotiate terms with a big retailer or distributor in the future if they’d missed a deadline.
Missing a deadline might also lessen the impact of all the advertising that targeted the release date. “The way to make something a hit,” says Seropian, “is to make that day as big as possible.” Bungie planned to spend 70 percent of its advertising and marketing budget on making the release day big. “I’ve got this theory,” he says. “I call it core marketing–treat the customer base as vectors for disease, the disease in this case being something good, a positive impression of the game. You’re trying to spread it to everybody in the world. The larger the contagion is the first day, the wider and deeper it’s going to spread. The goal is to infect everybody. If we get them all primed and ready to explode, and we don’t get the product delivered, not working right, that kills the momentum.”
The advertising they’d scheduled was going to cost them a lot. “We actually spend more money rolling out a product than we do developing it,” says Seropian. “That’s a fairly consistent trend in this industry. It’s different from the movie industry, where they’ll spend 100 million making a movie, then 10 to 20 million on marketing. We’ll spend one million making a game, 1 to 1.5 times that much to launch it.” That’s not a large amount of money in the gaming industry. Eidos Interactive, the makers of Tomb Raider, reportedly budgeted $5 million for the ad campaign for just one of the games in that series.
In the second week of November, Jason Jones and Nathan Bitner, an old U. of C. friend of Jones’s who’d been hired in August to do story development and design game levels, were arguing over the title for one of the Myth II levels. Bitner wanted to keep the working name, “A Murder of Crows.” Jones said no one would understand the phrase. Bitner said they would, because it was used on a Counting Crows album. “I’d hate to dumb down for the herd,” he insisted. Eventually he won.
That same week Bungie got an order for 7,000 units from Kmart, which had never ordered a Bungie game before. The chain had decided to carry Myth II because it would feature a “no blood” option, which would allow parents to lock out the more violent blood-and-gore aspects–when the lock was on, deaths would be accompanied not by buckets of blood but showers of stars. That option would also be available in Germany and Korea. “Both countries,” says Zartman, “have very strict laws about the amount of violence you can have in a computer game.” The feature was critical to moving Bungie games into the mass market. “We pretty much got all the hard-core gamers with the first Myth,” says Zartman. “We really wanted mass market–because it’s a lot more sales.”
On November 18 the Bungie testers–most of the other staff–finished the first round of playing and commenting on the early levels of the game. The process proved easier than it had in the past, because a lot more people were doing the testing. Zartman would often play the game at the office, then take it home and play it another three or four hours, trying to find the best things to show the media as well as searching for problems like everyone else.
By this point Bernal, Regier, and Bowman weren’t the only ones practically living at the office. On November 22 Diane Donohue, director of operations, was working even though it was Sunday and her birthday. She and Zartman were putting together 140 Myth II boxes that would be sent to the major game reviewers in the U.S. and Canada as soon as the game was finished. They were using big boxes–18 by 20 by 5 inches–even though each would contain only the book-sized game, a tiny metal figurine of the Soulblighter, and a press kit. Zartman says that during the Christmas season reviewers are swamped with games to be reviewed, and he didn’t want Myth II to be overlooked.
Despite all the cutting-edge digital technology used to produce the game inside the box, Zartman and Donohue were laboriously hand inking each of the three-inch letters that spelled out “This Game Is Huge,” then pressing them onto the box.
On November 24 the staff finally acknowledged that they weren’t going to finish in time to make Christmas. They decided to shoot for December 13, which would get the game on store shelves in time for the January 3 advertising blitz. “It’s not as catastrophic as it sounds,” said Zartman. “The second biggest time to buy games is immediately after Christmas. Gamers are taking back the game they didn’t want and getting the game they did want–and they’ve also got all that holiday money.”
But even December 13 seemed dicey. “There’s definitely some nervousness,” said Deniz. “Some levels still need a lot of work.”
Bowman was still working on level 16. It was now structured, he explained, so that “when you enter the catacombs, if you go to the left there’s a room with some ghost archers that fade in as you come in and shoot at you. As they’re shooting at you, some ghost Myrkridia, who are their enemies, appear and attack the ghost archers.” The player winds up in the middle. “You’re gonna take some damage.”
The next night at a staff meeting Jones said, “Catacombs have a lot of problems, but the new ghost effect is more believable. The ending’s horrible. Have to fix it.”
They then discussed the Forge, the last level in the game. The consensus was that it might be too easy, though they agreed that it did provide a nice closure to the game.
That week the Parade magazine reviewer put Myth II on his list of hottest new games, even though he’d seen only a demo. He called it “a bloody but unbelievably enjoyable real-time strategy game.”
By November 30 the team had 17 of 25 levels (and a secret one) ready for final testing. Deniz gave CDs to each tester. “We’ll all just treat it like we got it out of the box,” Zartman said. “Load it up, follow the instructions, play the game straight through–trying to play as stupidly as possible.” Everyone had agreed to report only critical errors, such as glitches that caused the game to crash. When a major problem was discovered, testing stopped, and the problem was fixed. Then new game CDs were handed out.
On December 4, Zartman wrote in an E-mail, “Alex buys dinner for everyone who stays late, and has been for a couple of months now. We’ve worked our way through the menus of all the delivery places nearby, so we decided to go to Gaylord. Well, it was delicious, but cost $200, which precipitated a serious discussion about whether we should get a cook. No decision was arrived at. This just goes along with the showers and the beds and the impromptu movie nights, another aspect of this place being much more of a home to the young single guys who work here than their empty, unused apartments. Tyson here is Canadian and will be going home soon when his visa expires, so he’s ditched his apartment and is living here full-time until we ship. He hasn’t gone outside in four days, and frankly has no reason to until it’s over. Dave Bowman has stayed here 8 days at a time, went home for an hour and a half, and came back for another 4 days.”
By December 7 anybody not fixing bugs was playing the game all day and all night. At some point one evening the testers slipped from testing to playing and started competing with one another, just playing for fun.
At 3:50 AM on December 10 the ninth version ran without a hitch. Myth II was “gold mastered,” “GM’d,” ready to go. There was a lot of cheering, but nobody broke out the champagne that had been stashed in the refrigerator for the occasion.
“When you work so hard for so long and you finally get to take your foot off the gas pedal, it seems anticlimactic,” says Regier. “I felt really drained by the end. It was nothing like Myth: The Fallen Lords, where we stuck around after GM’ing and played Myth on bungie.net until 6 AM.”
“I hadn’t been doing anything for three or four months except working at Bungie,” says Matt Soell, “sleeping and eating lots of mediocre Mediterranean food from the place down the block. Be here 18 hours. Go home–it takes one hour to get home. Sleep for four hours. I got an E-mail from this girl I knew the day after we went gold. She wondered if we could get together before she went to visit her family for Christmas. I wrote back, ‘You know, I don’t really have the mental wherewithal to decide anything right now. If you could come up with a date, a location, and an activity, and communicate those to me, then I can follow instructions.’ I didn’t even have the ability to make a simple sort of decision like going out with someone.”
Bernal says he was “relieved, but very tired. Happy that we accomplished it. And then you wonder what the next game is gonna be–pretty much right away.”
Later that morning Zartman sent an E-mail to the people on his press list. “Bungie Software tiredly but happily announces that Myth II: Soulblighter, sequel to 1997’s seminal 3D strategy title Myth: The Fallen Lords, has gone to replication. As you read this, a plant somewhere in Atlanta is replicating and boxing hundreds of thousands of copies and preparing to ship them to the far corners of North America. Elsewhere in the world, GT Interactive, Pacific Software Publishing and other publishing partners are doing the same….This sequel has been turned in a mere seven months.”
“It didn’t hit me until Doug’s news release went out,” says Bowman. “Right up to that moment it’d been such a nonstop look for the next problem that it seemed like it’d never be done. It was just like being a little kid–maybe five–and it’s Christmastime again.”
Three days later Jim Ruiz started thinking there might be a way to get the thousands of Myth II orders that individuals had placed delivered in time for Christmas. Bungie wasn’t under any obligation to do so; after the November 1 shipping deadline had passed, people who ordered weren’t promised a specific delivery date. But Ruiz thought that if he could bring it off, it would make a lot of people happy and it would make Bungie look good.
He knew the games would have to be shipped by December 16 if they were to arrive by Christmas. He ordered some credit-card-processing software, since there wouldn’t be time to process the orders by hand anymore. The software arrived, but it wouldn’t work. “I spent the next 15 hours with technical support at four different companies,” he says. “Spent all night.”
But he’d missed the December 16 deadline. Processing all the orders and getting the credit-card companies to accept the charges before shipping the games was no longer an option. So he persuaded Seropian to let him immediately send the orders to the warehouse for shipping, then process the orders, and have any that were rejected by the credit-card companies pulled off the shipping dock–a risky strategy, because there was no way to guarantee that the rejected orders could be pulled in time.
The next day the credit-card companies rejected only 100 orders. “I called the warehouse and canceled those,” says Ruiz, “and they pulled them off the truck.”
On December 22 the reviewer for the on-line All Games Network wrote, “If you only buy one game this Christmas, do yourself a big favor and buy Myth II: Soulblighter. Games this good only come by once in a long while, and trust me, this is one of them.” The next day the reviewer for the on-line Happy Puppy wrote, “From start to finish, Myth II: Soulblighter is a textbook example of how to make a great game. It’s not revolutionary the way the original was, but I wasn’t expecting that. Myth II is the jewel to Myth’s diamond-in-the-rough.”
Myth II was scheduled to start hitting the shelves of game stores on December 28, and that morning 200,000 copies were on skids or on trucks that were carrying them across the country. That same morning one of Bungie’s overseas publishers called to say that a woman had played the game and then tried to uninstall it. The game software had uninstalled Myth II as it was supposed to–but it had also erased some of the other files on her hard drive.
The first reaction was panic, says Zartman. “You tend to overreact, tend to think, ‘This is the end of the company.'”
Jones and another programmer tried to replicate the problem. They quickly found it and figured out how to correct the faulty lines of code. Fixing the problem would be easy–except for the 200,000 copies that were already packaged or on the road.
Seropian, Jones, Zartman, Donohue, and sales and marketing director David Joost met to decide what to do. They had two options. The first was to say nothing. The argument in favor of this alternative was that the problem occurred only when the program was uninstalled and then only when someone had installed the game in an unconventional way to begin with. Since the chances were slim that anyone would install the game in this weird fashion and uninstall it right away, the risk to the company’s reputation was minuscule, particularly since they could post an announcement on their Web site along with a free patch that could be downloaded to correct the problem.
The second option was to recall the game. That would force them to trace every copy that had already gone out, scrap every finished copy, and start all over again.
“The thing that made the decision easy,” says Jones, “was that if we were to ship the game anyway and try to fix the problem later, some people were gonna get screwed. And that was wrong. It might not have been very many people–maybe one or two. But it would have bothered us the rest of our lives. Maybe not–maybe just two years. We’d be sitting around today: ‘Damn, wonder when the next person’s gonna call?’ It was so clear that there was one decision that led down the road of eternal damnation. The other was to spend a lot of money and do the right thing–and never make the same mistake again.”
They agreed to recall every copy. “I’m very proud of that,” says Donohue. “Some companies you’re embarrassed to work for. Not here.” Seropian later calculated that that decision and the fines they had to pay for missing their deadlines cost them $800,000.
Some parts of the recall were relatively simple. Zartman had to call the reviewers on his press list, notify them of the problem, and tell them that an update, version 1.1, would soon be coming.
Joost and Donohue had the more difficult task of getting the 200,000 copies back, replacing them, and then sending out new copies. Donohue immediately called the factory in Atlanta and told the managers to stop printing copies of the game and hold any shipments that hadn’t already gone out. Joost began calling the hundreds of stores that were awaiting shipment, telling them to refuse any orders that arrived at their docks.
As the games that had been sent out were recovered, each individual package had to be remade. Joost explains, “You open the carton, then the box for each individual unit, take out the jewel case, unwrap the shrink-wrap, take out the old CD and destroy it, put in the new CD, re-shrink-wrap the jewel case, repack it in the box, put on a new UPC label that would electronically read this as version 1.1, and put a red sticker on the box saying that this is the new updated version 1.1.”
Joost also had to get every purchase order he’d received reissued. “It’d taken months to compile all these POs,” he says. “Now, in a couple of days, I had to get all the POs recut.”
The next day the on-line media responded to the recall decision. The statement of the reviewer at Mac Observer was typical. “Bungie has proved once again that it is not an ordinary game company. While it is common practice these days to release a product that is known to be defective while working on a ‘patch’…Bungie stands above the crowd.”
Donohue had sent a copy of the corrected disk to the factory in Atlanta by overnight courier and, as a precaution, had sent duplicates by a second delivery service. Then she booked a flight to Atlanta on Sunday, January 3, so she could be at the factory Monday morning to supervise the reworking and reshipping.
That was the weekend Chicago got hit by a blizzard–most flights out of the city were being canceled. “I tried to get a private pilot to fly me down to Georgia,” says Donohue. But no one would sign up. “Early, early Sunday morning, a car and driver took me to O’Hare for my flight at 2 PM. I left my house at 6 AM. It took two hours to get to the airport. The roads were a mess. As I walked up to the counter they were canceling my flight.” But an earlier Atlanta flight had been delayed and was leaving in 15 minutes. “They looked at my frequent-flyer miles, patted me on the shoulder, said, ‘It’s OK.’ I took off for Concourse C, made it to the gate.”
At 7 AM on Monday the factory began producing Myth II, version 1.1. Donohue had brought with her 16,000 stickers that said this was the updated version. She’d shipped the rest of the 200,000 stickers by FedEx. They hadn’t shown up, and FedEx couldn’t find them. She called around frantically until she found someone who could print 25,000, which was enough to complete the CompUSA order.
“Then there was a large explosion and the power went out,” she says. “The entire factory was down. So here I am standing with my mouth open, and my customer-service rep says, ‘I want everyone to meet our client–Job.'” She laughs. “The power came back on in 20 minutes.”
Shortly before midnight on Tuesday, Donohue and the loading-dock crew watched the truck with the last part of the reshipment leave the factory. She brought out the cheesecake she’d ordered and threw a small party.
On Thursday, January 7, version 1.1 was on store shelves. As it turned out, day-one sales equaled the total sales of Myth I.
Art accompanying story in printed newspaper (not available in this archive): photos/Eugene Zakusilo.