By February 1999, the staff had had enough. Carmack clearly had no interest in managing
the day-to-day affairs of the game. They wanted a producer. So Carmack called Graeme
Devine. Graeme was a prodigy in gaming history. At sixteen he was expelled from high
school in England because he was spending too much time programming games for Atari.
He later moved to the United States to cofound Trilobyte, the company that produced
one of the bestselling and most technologically impressive CD-ROMs of the early nineties,
7th Guest. Graeme and Carmack had struck up a programmers’ relationship along the
way, corresponding often about the latest in coding. Now that it was time to bring
on help for Quake III, Carmack thought of Graeme, whose own company had recently gone
under. Graeme was more than happy to come on board, but he was surprised at what he
found.
When he asked the fourteen people what they thought the direction of the project was,
he got fourteen different answers. The night before his first day, three of the newer
guys took him out for a coffee at Starbucks in Mesquite to prepare him for the bloody
arena he was about to try to tame. “They are going to tell you that you have power,”
one explained, “but you will not have any power. They may say something’s okay, but
they will override your decisions.” They warned Graeme of what to watch out for: the
mind games, the politics, the people to distrust. Graeme wasn’t swayed. “Everything’s
going to be okay,” he assured them. “Don’t worry. Things will change.”
Nothing did. Graeme discovered that the egos at id were stronger than he’d surmised.
Though people felt like they were working without direction, at the same time no one
wanted really to be told what to do. To make matters worse, a fundamental ingredient
of the game—the bots—was missing. Bots were characters controlled by the computer.
A good bot would blend in with the action and flesh out the scene like a robotic extra,
as well as interact with the player. For Quake III, a deathmatch-only game, bots were
essential for single-player action. They were implicitly complex because they had
to behave like human beings.
Carmack had decided, for the first time, to delegate the job of creating these bots
to another programmer in the company. But he failed to follow up. Once again, Carmack
incorrectly assumed that everyone was as self-motivated and adept as he was. He was
wrong. When Graeme struggled to rein in the work, it was discovered that the bots
were completely ineffective. They didn’t behave at all like human beings. They behaved,
basically, like bots. The staff began to panic. By March 1999, they had reason to
be scared.
At the Game Developers Conference in San Jose, id employees got their first look at
Unreal Tournament, a new game by Epic, the creators of the 1998 shooter Unreal. Epic
had quietly become formidable competition. Tim Sweeney, Epic’s lead programmer, was
revered. The company had even employed two former id guys: Jay Wilbur and Mark Rein—the
“probationary president” from the Wolfenstein days—to handle business affairs. Unreal
was a surprise hit, bringing, like Half-Life, more of a cinematic story feel to the
genre. But their new game took id by even greater surprise. Unreal Tournament was
a deathmatch-only multiplayer game, just like Quake III.
Epic, some thought, had flat out stolen id’s idea. They resented the fact that Carmack,
as usual, had been so open about the company’s direction in his .plan file. But Epic
denied stealing anything, saying that they had been on that track long before Carmack
had announced it in his plan. The animosity and competition nevertheless remained.
And, with Quake III so disorganized, not to mention the heat of Columbine, there was
no reprieve.
Despite positive reviews of the game at E3, id began to fall apart. Two respected
young employees—Brandon James, a level designer, and Brian Hook, a programmer—quit
in frustration. Adrian separated from his wife, whom he had married during the height
of Doom’s success. Kevin, always the conciliatory owner, sensed that something had
to be done. He asked Carmack to move out of his own office and share space with him
and Adrian in hopes of improving communication. The move only intimidated the other
employees, who couldn’t help but wonder what was going on with their owners behind
closed doors. What was going on, in fact, was nothing much. Instead of talking, the
co-owners worked in silence. The only sound came when Carmack left and Adrian and
Kevin turned up their stereo.
By the end of the project, Graeme wasn’t producing, he was programming. The bots were
farmed out to a well-known mod maker in the Netherlands, who heroically brought them
to life. The levels were stitched together in some sensible sequence. In November
1999, the game was close enough to completion that some members of the company went
on a promotional bus tour for the release. The fun was cut short, however, when they
discovered that their competitor, Epic, had one-upped them again, releasing Unreal
Tournament just a week before Quake III hit shelves in early December.
At the wire, the question remained: Would players go with Carmack’s—or Epic’s—vision
of a plot-free, deathmatch-only online world, or would they stand by Romero, the beleaguered
designer who was hoping to prove that Daikatana, in all its wild ambition, would save
the story after all? Romero weighed in on the matter in a business preview of Quake
III that appeared in
Forbes.
“Online gaming is still a small segment
of the market,” he told the magazine. “And the people who play over the Web are the
ones most likely to find sites where they can download the game for free.” The magazine
offered its own verdict of who would ultimately win the battle of the Two Johns. “It’s
quite possible,” the story concluded, “that id’s far less grandiose strategy is the
better one.”
“Aaaaarrggggggggh!”
Shawn Green screamed as he thrashed his computer keyboard against the floor. It was
midnight in the coders’ cove of Ion Storm, and the cubes were as dark as the city
below. Dressed in a black T-shirt, Romero’s old friend hunched like an ape at the
beginning of
2001: A Space Odyssey
as he whacked keys across the floor. A skinny programmer stretched his neck out of
a nearby cube to observe the tantrum, then nonchalantly returned to his work. Shawn
brushed his hair back as a smile crept across his face. “Nothing like a little stress
relief,” he said, tossing the battered keyboard down the hall.
Shawn, like the others on the Daikatana team, was deep into crunch mode. Despite Romero’s
pledge years before to Carmack that his death schedule days were over, he had upped
the team’s core hours to include weekends; the staff was now elbowing for bed space
in the lounge. Brian “Squirrel” Eiserloh had recently spent eighty-five out of ninety
days without leaving the office. Several others were crawling to sleep under their
black-shrouded cubicles, nestling on floors covered in loose M&M’s and pizza box pillows.
Stevie Case was stuck home sick with a kidney infection. Romero had even taped a sign
to the office’s most popular arcade game that said, “No More Tekken 3 Until Daikatana
Ships!”
Shawn himself was about to take his first and only break in weeks, heading off to
an abandoned abortion clinic to unwind with his death metal band, Last Chapter. After
staring at lines of code all day and sucking down a half case of Mountain Dew, he
was always looking for new ways to blow off the steam and caffeine. He and Romero
joked about making a life-size porcelain doll that would stand in the office holding
a baseball bat. The punch line was that it held its own demise.
So did Ion Storm by the fall of 1999. Romero’s ship wasn’t just off course, it was
perched on a rock in a violent sea with a steady flow of crew members leaping—or pushing
each other—off board. Reeling from the Ion Eight departures, the
Dallas Observer
story, and the Columbine controversy, the company had suffered yet another blow as
a result of that year’s E3 convention. The pressure going into the convention had
been enormous, especially with the game promised, this time, for a December release.
Todd confronted Romero before the event. “Look,” he said, “Eidos is significantly
concerned, and we need to have some sort of oversight; I need to make sure that things
are coming along the way they’re supposed to come along.” He would get the game ready
and send it to them at E3.
Romero and Stevie showed up at E3 in May 1999 looking the part of gaming’s rock-and-roll
royalty. Romero was in black leather pants, mesh black shirt, long silver chain. Stevie’s
bleached blond hair spilled over a skintight baby blue shirt and black pants. Despite
the bad press, they were besieged by the usual hordes of lustful boys and autograph-seeking
Doom fanatics. When the disk of Daikatana’s demo arrived, however, they didn’t like
what they saw. Bugs in the program were causing the game to chug slowly across the
screen. Romero flew back to Dallas and stormed into the office of John Kavanaugh,
the Eidos representative who had been stationed to keep an eye on the company.
“I’m fucking leaving,” he told Kavanaugh. “If Todd’s going to stay here, I’m fucking
gone. I can’t work with this guy, he’s ruining everything.” Kavanaugh told Romero
a meeting would be called with Charles Cornwall, the owner of Eidos. “All you have
to do,” Kavanaugh said, “is nod your head.”
In June, the Ion Storm partners—Romero, Tom Hall, Todd Porter, and Jerry O’Flaherty—flew
to Los Angeles, ostensibly to talk with Eidos about buying out the majority of the
company to relieve the owners’ $30 million debt. Todd and Jerry quickly learned otherwise.
“I’m sick of all this bullshit!” Romero burst out. “Either Todd goes or I go.”
Kavanaugh feigned incredulousness. “That’s bullshit, John,” he said. “You
are
Ion Storm. There’s no decision here.”
Jerry knew what this was all about: control. Romero wanted total control over his
game. He tried haplessly to offer a solution: “If your problem is Todd working on
your project,” he said, “can we give Todd a project? Does anyone have a problem with
that?”
“Look,” Romero said, “this is not a fun place to work anymore, and I think we need
to take a partnership and go in a different direction.”
Before Jerry could suggest another option, to his surprise, Todd agreed. “Look,” he
said quietly, “you’re right, John. It’s no fun for me either. It’s clearly no fun
for Jerry.” Jerry acquiesced. They would go. The four owners walked out into a flash
of lightbulbs. But this time it wasn’t for them. It was for actors Heather Graham
and Rob Lowe, who were there for a photo shoot.
Ultimately, Todd and Jerry were happy to be getting what would surely be a healthy
buyout for a company that seemed to be going down the tubes. Romero and Tom were relieved
to have a new beginning. Maybe the company would be saved by the games, but either
way they all still believed in their original vision: that design, that
the games
could be law. The problem, as Romero said, was that the design didn’t take into account
technology and it didn’t take into account that the designer doesn’t necessarily know
how to manage.
In October, after reporting a loss of $44.8 million, Eidos announced that it was purchasing
51 percent of Ion Storm. Romero spent the fall buckling down on the game. The interview
requests were turned away. Tekken 3 remained unplugged. Though the monsters, levels,
sound, and art were nearing completion, there was a formidable task ahead: burning
through the remaining bugs—all five hundred of them—in time for their promised December
17, 1999, release. But Eidos was confident enough to schedule a release party for
that day, despite Romero’s objections.
The party came, but without a finished Daikatana game. Not until April 21, 2000, would
Romero finally feel ready to release it. The next day he sat down at his computer
and typed a message for readers on the Internet. “My god, it’s finally finished,”
he wrote. “And I thought working on a game for 1.5 years was long. . . . Wow. I wish
everyone would take a nice, objective look at the game and not base their criticisms
on hype, but on play value and what we’ve worked to achieve: a really fun single-player
experience.” Romero tried to dissuade the inevitable comparisons with his former company.
“We did not,” he wrote, “develop Daikatana to take on Quake 3.” But the final score
was out of his hands.
SIXTEEN
Persistent Worlds
By the end of 2000,
Carmack didn’t just have a new game release to celebrate, he had a marriage: his own.
A couple of years earlier he had received an e-mail from a young businesswoman and
Quake fan in California named Anna Kang. She wanted to start an all-female Quake tournament.
Carmack said that’d be fine, but she’d probably get only twenty-five people. She got
fifteen hundred. He respected anyone who could prove him wrong.
Who was Anna Kang?
She was a strong-willed woman on a lifelong quest for respect. As an Asian American
growing up in Los Angeles,
Katherine Anna Kang was called a banana
, a slur given to Asian American women who were thought to be white on the inside
and yellow on the outside. Anna didn’t let the insults sway her beliefs that, as she
said, “women don’t need to be subservient to males, that marrying outside their own
race shouldn’t be a sin, and that in general, capitalism isn’t evil and socialism
isn’t ideal.” One of her greatest role models was the author Ayn Rand; Anna wanted
to be a powerful person like Gail Wynand, a character from
The Fountainhead.
She never felt as fierce as when she played Quake.
After successfully hosting the tournament, she stayed in correspondence with Carmack.
He was intriguingly selfless—the way he shared his code and his knowledge. Even though
she sometimes teasingly called him Spock, she believed he had a deep and generous
soul. Carmack was equally impressed by her, talking at great length about Ayn Rand,
philosophy, and games. He liked the way she challenged him.
They began a long-distance romance. Carmack ended up, with his staff’s approval, offering
Anna a job in business development so that she could have a reason to move to Dallas.
She came, but the days at id wouldn’t last. Her relationship with Carmack would. They
wed in Hawaii in front of a small crowd of family and friends. It was one of the only
vacations Carmack had taken in his life. And, like the other times, he brought his
laptop. There was work to be done.
In Carmack’s opinion, Quake III was—like all his other games—ancient history compared
with what he was ready for now. Online games had become successful. The most ambitious
existed as persistent worlds, which remained accessible over the Net around the clock
for players to visit and explore. Medieval-themed titles like EverQuest and Ultima
Online, based on Richard Garriott’s early hits, had sold millions of copies, forging
a genre called “massively multiplayer online role-playing games,” which could support
thousands of international players at a time. Players
lived
in these digital landscapes, spending dozens of hours per week exploring, battling,
and building their game characters; EverQuest was nicknamed Evercrack. Some had even
taken to selling coveted virtual objects—weapons and accessories—accumulated in the
game for
real
money on auction sites like eBay.
For Carmack, this phenomenon fulfilled an ideology that harked back to the populist
dreams of the Hacker Ethic. “It allows us to have virtual resource,” he said. “Any
of these digital resources allows us to create wealth from nothing, to be able to
replicate wealth freely. . . . Unlike most fundamental physical objects, with the
digital stuff there really is this possibility of wealth replication. The world can
be a richer place.”
Upon returning to Dallas, he decided to unveil his new direction to the rest of the
team. “We should focus on doing a generalized infrastructure,” he told them, “and
doing a game as one element of this generalized infrastructure which can have a lot
of the 3-D web environment that people always are thinking about and wishing about.
We can do it now.” This was it—the culmination of his work, his engineering, the dreams
of science-fiction writers from Aldous Huxley to William Gibson. The Holodeck, Cyberspace,
the Metaverse, the Virtual World, it had been called by many names, but the technology
was never ready to bring a true glimpse of that place—however primordial—to life.
That time, Carmack concluded, had come.
He looked around the conference room and waited for the response. All he got were
blank stares. “But we’re a game company,” Adrian said, “we make
games.
” Carmack sighed. He knew that despite his power and prestige, he couldn’t do this
alone. He needed an experimenter who could use his technology to paint the new world.
He needed a person who was so blown away he couldn’t speak, someone who committed
every cell of his body to bringing Carmack’s visions to life, someone who understood
that this was
the coolest fucking thing planet Earth had ever seen!
He needed Romero. The meeting was done.
The question of what id would do next remained unresolved. Though no one wanted to
do the Metaverse, there was growing feeling in the company that they wanted to do
something different. Graeme Devine came up with a proposal for a game called Quest:
a multiplayer role-playing game, a far cry from a first-person shooter. Adrian and
Kevin were excited about the possibility of a totally new kind of game. They weren’t
the only ones. With the community accusing id of going back to the well one too many
times, this was a chance, it seemed, for a break. As Graeme proclaimed, “No more rocket
launchers!” It was agreed. Quest would be id’s next game.
But before long Carmack grew to hate it. It seemed murky, at best, like a three-year
miasma waiting to happen. There was another idea, one that had surfaced again and
again over the years: a new Doom. Carmack didn’t exactly love the thought, but he
didn’t hate it either. Under id’s supervision, another company, Gray Matter, was already
working on a new PC game based on Wolfenstein 3-D called Return to Castle Wolfenstein,
and the anticipation among gamers was high. For Doom III, Carmack could incorporate
ideas he’d been kicking around for a next-generation graphics engine, something that
could dynamically exploit the world of lights and shadows. Second-generation id guys
like Tim Willits and Paul Steed, who got into the game business
because
of the original Doom, frothed at the opportunity to do a new installment. Carmack
even kicked the idea around with Trent Reznor of Nine Inch Nails, who had done the
Quake audio, to see if he would be interested in doing the sound for Doom III. Trent
said yes. But others were less than intrigued.
Kevin, Graeme, and Adrian, in particular, disliked the idea. “It’s like some old band
going back and remaking their first album because they can do a better job now,” Adrian
said. “What the hell’s the point in that? It’s, like, make something new with your
time! Instead of dedicating two years to making something you’ve already done, why
not try and push this genre that we created further?”
To bide time, it was decided that the company would work on an add-on pack for Quake
III called Team Arena. This game was a clear reaction to the success of Unreal Tournament
and the criticism that id had not provided enough team play in their own title. The
work on the mission pack, however, began to idle as people wondered what direction
the company was going to take. While the prospect of Quest became all the more grim
in Carmack’s mind, he hatched a plan to get his way once and for all.
He walked into Paul’s office one night and said, “Trent Reznor wants to do sounds
for Doom.”
“Doom?” Paul said. “We’re not doing Doom, we’re doing Quest.”
“Well, I decided that I want to do Doom. Are you on board?”
“Fuck yeah!” Paul said. Tim agreed.
The next day Carmack walked into Kevin and Adrian’s office and said, “I want to do
Doom. Paul wants to do Doom. Tim wants to do Doom. If we don’t do Doom, I’m leaving.”
Then he turned and walked out the door.
Kevin and Adrian couldn’t believe it. But what could they do, fire Carmack? What was
id without the Whiz Kid? They discussed the possibility of splitting the company in
two teams or maybe, as Adrian thought, just throwing in the towel altogether. Carmack
had threatened to quit before. And Adrian was beginning to feel like maybe the time
had really come, maybe Carmack was through. Later he approached him and asked what
was preventing him from just walking out the door once they started working on a new
project. Carmack said, “Nothing.” A decision was made. Carmack sat down at his computer
the next day and told the world his plan.
“How’s the Hummer?”
How’s the Hummer? How’s the Hummer?
That question again. At this moment in the summer of 2001, it was coming from a crew-cut
teenage boy with a greasy face and a handful of Mexican menus at Temerararia’s Restaurante
and Club, the self-proclaimed “finest in Mexican Dining” in Lake Tawakoni, Texas;
then again, it was the
only
Mexican dining here in this tiny strip of country thirty miles east of Dallas. Romero
and Stevie Case had become regulars since escaping to the country a few months earlier.
Now they were known around town less for their rock-star looks than for their fleet
of expensive sports cars. After the hell of Daikatana, it was a welcome change.
The game was brutalized—critically and commercially,
selling only 41,000 copies
in the United States. With the exception of a few favorable reviews, the fans and
media tore the game apart.
Entertainment Weekly
called it
“a disaster
of
Waterworld-
ian proportions,” referring to the epic failure of Kevin Costner’s notorious film.
PC Gamer
said the game
“signals nothing more remarkable
than the end of an era in fandom.”
Computer Gaming World
was more succinct.
“Yep,” the headline declared, “it stinks.”
Romero thought anyone who actually played the game would be hard pressed to not have
a good time. But most people were hard pressed to play the game much at all. The opening
level, with its swarm of mosquitoes and squishy stampede of robotic frogs, struck
players as annoying. Many would not get past the pests. Romero was the first to say
in interviews how much he enjoyed the game—despite its grueling development. He insisted
that it broke even through all its license deals and foreign sales. In addition, Warren
Spector’s Deus Ex had proven a great success for Ion Storm, being voted computer game
of the year by several publications. In his mind, no matter Daikatana’s sales, Deus
Ex had proven Romero’s vision of a gaming empire with multiple titles right.
But Romero’s enthusiasm, for once, couldn’t stave off the inevitable. After Daikatana,
he began sketching out ideas for a prototype of a follow-up game. Before long he had
to help Tom Hall complete Anachronox after many on the team had departed in frustration.
Tom’s sci-fi role-playing game, which followed the trail of a down-and-out space detective
who must thwart a mysterious alien oppressor, had become, like Daikatana, truly epic:
employing hundreds of creatures, an arsenal of customizable weapons, and numerous
games within games. But by early 2001, the work was near completion.
It wasn’t the only thing coming to an end. One day, word spread that Eidos was going
to be terminating some Ion Storm employees after Anachronox was complete. Curious,
Romero walked over to the CFO’s office to see who was on the list. He picked up the
sheet of paper. It listed everyone’s name in the Dallas office, including Tom’s and
his own. Warren Spector’s team in Austin, however, would remain. Romero returned to
his desk, sat in his chair with the highway behind him, and dialed Stevie. “That’s
it,” he told her. “It’s done.” Romero was never one to mull, but this time it did
actually stir his feelings. It was fucked up, he thought, that the vision, the dream
design, didn’t pan out like he had hoped. It was even . . . sad. But the sadness,
as always, wouldn’t stay long. He called Tom. It was like the clocks were rolling
back and Romero was back at id, calling Tom at the end of one chapter and the beginning
of the next.
Before long Romero, Tom, and Stevie hatched plans to start a new company. They sketched
out ideas for games: maybe one about a ten-year-old kid, not unlike Commander Keen,
who has to get through life, do his chores, and get along with his family, maybe an
Old West shooter, or a game based on Madonna. They also talked about doing games specifically
for mobile platforms—Pocket PCs, cell phones—
an estimated $6 billion history
by the year 2006. And Romero had an idea for a new game based on the Quake franchise.
One afternoon he took the long drive to Mesquite to talk it over with id.
Id was no longer in its notorious black cube. The company had moved offices the year
before to afford everyone more space. The new digs, Romero found, were even more plain
than the last: just an ordinary office building across from a Hooters and an Olive
Garden. Though Romero had seen Carmack around town occasionally, it had been years
since he had come to him with a business idea, a creative idea, a
game,
not since Quake, five years before. Carmack was at his computer, optimizing his new
Doom engine, when Romero walked in. The office was bigger, cleaner, but minimalist
as ever. Carmack sat before a large monitor angled in the corner. Through the blinds,
he could keep a watchful eye on his Ferrari out back.
“Hey,” Romero said.
“Hey,” Carmack said.
Romero told him the reason for his visit and gave a brief pitch: “How would you feel
about me licensing the Quake name to develop a game based on the franchise but set
in a persistent world?”
Carmack nodded. “Sure, why not.” He had already been kicking around the idea of getting
the original team—Romero, Tom, Adrian, and himself—back together to do a version of
Commander Keen for the Nintendo Game Boy, the handheld gaming device. Though he and
Romero both knew they would probably not be in the same company again, that didn’t
mean they had to work apart.
Romero left and was on his way, on to the next thing, the flipped bit, the new vision.
It all started with a house. Romero knew that he wanted a company like the early id,
something intimate, something communal. And for that he needed just the right environment,
like the lake house in Shreveport. Stevie saw something advertised online that she
thought could fit the bill. They hopped in the Hummer and drove out past id, past
Rockwall, down a long country road with strange country-road establishments: trading
post, abandoned school buses, and a rusty flying saucer the size of a bus in the middle
of a field. When they asked about the spaceship, locals joked that it had just appeared
there a few years ago; now someone was trying to turn it into a hot-dog stand.