Like most of us, Tim Sweeney has owned several Ferraris. He's also owned a few cell phones.
The Ferraris are the receipt for a life spent making computers do amazing things. They're cool. They tell their own story about a boy who grows up with a computer as his best friend, a teenager who finds his footing in life, builds his own company, transforms the world of video games and achieves most of his goals.
What he does with his cell phones, however, reveals what kind of man Tim Sweeney always was.
"Whenever I have an old cell phone that I need to get rid of," he told me during a lengthy phone interview from his office in North Carolina, "I take it apart and look at all the components and see what they're doing."
We chuck our phones or stuff them in a drawer. He dissects them. He's a tinkerer, a man who believes he was born five years too late to become an auto mechanic and so has spent more than two decades getting under the hood of the electronics that entertain us, making them better and, just as significantly, helping hordes of other people, professional and amateur, make them better too.
In February, Tim Sweeney will be inducted to the Hall of Fame of the Academy of Interactive Arts and Sciences. He will earn a spot alongside 15 other giants of video games, people such as Super Mario inventor Shigeru Miyamoto, legendary Doom programmer John Carmack and Sim City dreamer Will Wright.
Like the other Hall of Famers before him, Sweeney gets the nod for more than his work on a single game. He did create the company that made Unreal and Gears of War. He also created the first Unreal Engine, the graphics technology that has evolved to become the ubiquitous bedrock upon which Xbox 360, PlayStation 3, PC and even iPhone and iPad blockbusters are built. And he built a business, Epic Games, a company he started solo in his parent's house and has grown into one of the most successful independent video game development companies in the medium. He drew great talent to him, a skill that his long-time colleague at Epic Mark Rein still marvels at. He was a rare combo, says Rein: "People think of him as a technical genius but he is a business genius as well."
If you've admired the look of Gears of War on an Xbox 360 or Infinity Blade on an iPhone, or Unreal Tournament on a PC or Mass Effect on a PlayStation 3, that's all because of what Sweeney built.
Thanks to Tim Sweeney, another of his longtime Epic colleagues Cliff Bleszinski says, "graphics finally don't suck." The lead creator of Gears of War might be overstating Sweeney's influence, but his praise makes sense. Sweeney's Unreal Engine was so good that, nearly 15 years ago, a gaming magazine could put a screenshots from a game running it on their cover.
These days, if a company wants to show off how impressive their new computer, phone or console is, the easiest path is to get the latest Unreal Engine on it. They want Sweeney or one of his colleagues on stage showing how well it runs and how easily people can make games with it. "He enabled us to make fun games that I call ‘system justifiers,'" Bleszinski explained. "He made an engine that made it easy for developers to make good-looking things fast, which then let you get a great-looking game on an iPad 2 or through a 3D card and show your wife that 'this is why I got this.'"
Sweeney thinks he took apart his first lawnmower when he was five or six years old. He was a born tinkerer. "I was always interested in technical things," he said, "building go-karts out of engines, boards and whatever pieces I could find. The people with my skill-set... there was a transition point. If you were born five years earlier than me, you were an auto mechanic. You did souped-up cars and drag-racing and all of these other crazy things. If you were born five years later, you were into computers. It was the same audience."
Technology enchanted him early. He didn't just want to dismantle things. He wanted to make them better. He was taking apart radios and TVs to see how they worked.
Born in 1970, he grew up in suburban Potomac, Maryland. His father worked for the Department of Defense's mapping agency, climbing from high school graduate to a directorship at a place where the employees secretly analyzed the photos of spy planes that could snap shots of license plates. "Whenever there was a war or hostilities, he was very, very busy," Sweeney remembered.
His mom raised him and two older brothers, Pat, 10 years Tim's senior, and Steve, who was 16 years older.
Sweeney first saw video games when he was nine or 10, when he discovered a video game arcade that was on his way to school. He dumped quarters into Space Invaders and into a forgotten Nintendo shooter called Space Firebird. Those arcade machines fascinated him too. "I had taken apart electronics," he said. "I understood that these were relatively simple, high-tech devices that were programmed by people."
Like most kids circa 1980 he also got an Atari 2600, his first home console. He mostly hated the "lousy machine." It was inferior to the cabinets at his arcade. He liked one game, Adventure, "but for the most part the games were too unsophisticated to catch my interest."
Even though much of his life would involve the creation of video games, Sweeney never became an avid gamer. Sure, he messed around with games, but, to this day, he thinks he has only finished two of them: Doom and Portal. He skipped Zelda and never touched a Final Fantasy. He tried Sonic the Hedgehog and Super Mario Bros., just for a few hours "I would play games long enough to discover what games were doing and how they were doing it. And then I'd spend the rest of my time building."
Playing games was never the draw. No, the draw for Sweeney was the technology that ran games. To a child in 1980, that tech appeared to be slowly but surely taking over the world. He was a computer kid.
"There is some tendency to anoint a few programmers in the game industry as superstars. It isn't really fair; there are surely superior programmers toiling in obscurity in various places, but Tim and I both get the treatment to some degree. It has been a long time since the 'heroic age' of game programming where a single programmer really did make most of the difference on a project, but Tim was there doing hard core low level programming with the best of them, and he has continued to string together wise decisions and good leadership over more than a decade of successful products since.
"In an industry well known for having its share of large egos, it was especially telling to see Tim once jokingly refer to himself as 'Carmack MX' when talking about hardware product branding."
He had a nerd's childhood, with all the good and bad that comes with that. At 8 or 9, he had friends who built toy spaceships in the woods with him. Then, as kids do, they decided what was cool. Tim learned that he wasn't.
"The last few years of elementary school to the first years of high school, I really didn't have a clear set of friends," he said. "I was kind of the awkward smart kid." He didn't try hard in school, so he didn't get good grades either. "I wasn't with the smart kids; I wasn't with the cool kids. I didn't really have a group to hang out with. Those years were very awkward. I don't look back on them very fondly."
Mark Rein will tell you that Sweeney has people skills, that Tim Sweeney's ability to talk with impressive intelligence and kindness to just about anyone, is the reason he's a great recruiter of talent. But those skills emerged slowly and not on purpose.
"Some people are really enthusiastic about social interaction," Sweeney said. "They go around and talk and collect things and negotiate with other people. I was never good at that. I always really gravitated toward the things that were technical in nature. I'm a smart, mathematical type of thinker. Whenever I had the opportunity to see how something worked and tried to improve on it, I had some good success with that, more-so than with social interactions. "
When he was about 11, Sweeney went to California to visit a company his oldest brother had started up. "He had a brand new IBM PC. It was one of the first several IBM PC's that they'd shipped when it had first come out. He showed me how to use it, how to program in BASIC. I spent a whole lot of the week that I was out in California with him just programming the computer, figuring things out. It was really incredible. Trying to build a go-kart, you can spend months on something and it never quite works right. But the computer would do exactly what you say. I could write, in just a few hours, a really impressive program. It was the ultimate machine to tinker with."
The IBM was the second computer Sweeney had seen. It was probably the one that determined the course his life would take. "It was astonishing seeing the IBM PC. It was such a crisp machine. I'd seen a Commodore PET computer before, and it was a crappy device that never really did what you wanted, I couldn't figure it out and nobody was there to show me how to use it. I sat down at this IBM and every key you pressed made this bright, quick sound. [It had] this nice clear screen, this very powerful basic programming language. It just took me a couple of days to figure out how to use that. It was totally love at first sight. From that point on, I tried to dedicate all of the time I had free to learning to do more with computers."
A few months after he learned the IBM, Sweeney got his first computer. Technically, it was his dad's. His oldest brother had bought his father an Apple II for a birthday, but young Tim was the one who pounced on it. The Apple II would turn Sweeney into a video game developer.
There was only one home video game that Sweeney loved. It was the one that didn't even try to have impressive graphics. It was the blocky but intriguingly mysterious Adventure. "That was the one game that really captured my sense of going out and exploring the woods," he said. "As soon as I got the Apple II it was my goal to figure out how to write that. And it took a while. Gosh, it probably took about a month to figure out the basics of how to write an Adventure-style game."
Sweeney's brain didn't work the way most gamers' did. He didn't play Adventure and then simply yearn to play Adventure II. He yearned to make an Adventure II. So, on his Apple, he figured out how to make a dot. Then he figured out how to make the dot move. He sorted out how to make a room and then another and then how to program collision detection for the objects in each room. He didn't know about subroutines, so he kept copying his code. He remembers discovering the GOSUB command that let him use the same code in multiple places. "There are several points in programming in those early days that I still remember," he said. They were revelations. There were his private discoveries of new gravities and physics. He was discovering the rules of another world and, soon, he'd be changing them.
Sweeney thinks he was "alright" as a programmer by age 11. By age 15, he wagers, he was really good. "I could look at any program that anybody else had written and, with enough effort, build something cooler." He would go onto online bulletin boards and download documents about programming that other people had done. He figured out their techniques and then would do them himself.
"For my first 10,000 hours of programming it was entirely on my own and entirely self-directed," he said. His triumphs were private. "In those early days I didn't really have anybody to share my programming experiences with. I was the weird computer guy and nobody really cared."
Sweeney became a better student in the latter half of his high school years. He also played the saxophone, but not well. "Not even Bill Clinton-well," he clarified in a reference to the former President's warbling on an early-'90s TV talk show. He also ran cross-country and developed a love of hiking that is responsible, even to this day, for mud that occasionally gets tracked onto the floor in Epic HQ.
He graduated and enrolled at the University of Maryland to study mechanical engineering, but computers were still the draw. His fascination was helped when his father bought him a 286-class IBM computer. Sweeney never released the games he made on the Apple II. There wasn't much of a market for Apple II software, he recalled, but the IBM world was booming. "By the time I got an IBM, I started thinking, ‘Hey, I could build a business around this.'"
The first thing Sweeney had to do with his IBM was create a text editor, a fundamental tool for inserting text and manipulating it. He needed it so he could start programming games for the machine. "And then I realized, 'Hey you know, I could make each character on the screen have collision.' I could have the cursor be a game character. I could turn this text editor into a little game. Similar to Atari Adventure, I based it on rooms. Each screen full of text became a room. I had different graphical characters represent different gameplay behaviors. Suddenly you build a text document and hit the ‘play' button in this editor and now you're up and running with your game." He had just made the first game that would earn him recognition. He called it ZZT.
That basic idea of ZZT—the concept of having the tools that are used to create a game also be the game—was a winner. That idea, tied with the idea of letting anyone work on those editing tools to make their own games, became Epic's core philosophy.
"We always approached game development from the tools perspective first," he said. "We build the tools we need. We build a user-friendly set of tools and carry them forward. The fact that ZZT, bizarrely, began as an offshoot of a text editor actually had pretty profound implications on our company's strategy."
"In addition to Tim's direct contributions to various important game technologies, he's always been really open and sharing with information, which has helped other programmers move the technical ball forward on
numerous occasions. He's vehemently anti-software-patent, which I think is important not only because he's in charge of a large-scale and influential technology licensing operation, but also because he's a role model for a younger generation of game programmers who will hopefully think twice (or 10^100 times) before patenting what they're working on, and they will realize that unencumbered open access to technical information helps the entire industry advance.
"His current work on programming languages shows he's willing to spend a lot of time and effort persuing ideas he thinks will change the way games are made for the better, and I respect that a lot.
"He also picked me up at the airport in some kind of Ferrari one time. Luckily I didn't have much
Making ZZT opened Sweeney up. He started inviting people from school to check it out. He got kids in the neighborhood to swing by. He'd watch how people played his game and take notes. "Suddenly, having built these programs and seeing other people enjoying them—and being really excited about them—really changed my perspective on building software. Suddenly it was no longer just an intellectual exercise but a way of sharing something with the world."
Rein marvels that Sweeney was as excellent at programming as he was at business. Sweeney had his older brothers' business skills to learn from, but he also had some private revelations. He learned to be a businessman when he was 15. He worked at a hardware store and noticed that everyone who worked there got $4 an hour no matter how hard they worked. "It seemed like, financially, it was non-optimal."
He decided to borrow a tractor from his dad and instead started mowing people's lawns. Mowing companies were charging owners of Potomac's new mansions $120 for a lawn. He'd charge about $60 and calculated that he was pulling in $25 an hour. "That's when I came to a really clear realization that, by trying harder, and striving to find cool business opportunities, you could do far, far better than the wage earners, who I was when I was working at the hardware store. At that point it became really clear to me that there were big opportunities in the world."
His mind turned to making money by programming, of course. "I started thinking, ‘If I made a game and it sold this many copies, I would make this many dollars.'" He realized he could make a living making games. He didn't create ZZT on purpose, but as the neighborhood kids got into it, he realized he should sell it.
In the early '90s, you could buy games at game shops, but you could also make, sell or simply play shareware. Shareware games were often distributed in pieces. The first piece was free and was designed to hook you to pay for the rest. That was the model for one of shareware's biggest successes, Scott Miller of Apogee Software. Sweeney admired Miller's approach and his business sense. The games were marketed well. They were sent to reviewers with a well-written form letter attached. This was how a professional game company worked, Sweeney thought. In 1990, Sweeney even sent Miller a note, seeking advice about a game he was making. Miller wrote back. Everything Miller wrote was polished, Sweeney noticed. He was embarrassed that his own writing had been sloppy. "I spent a whole lot of time polishing, and it worked."
In 1991, Sweeney released ZZT to the world as a shareware game. He was 21. He called his company, which he ran out of his parent's house, Epic MegaGames. He figured that sounded impressive. Orders came in right away, he recalled. He got three or four a day. That netted him about $100 per day. He liked that, but he knew that game's reach was limited. "The audience for that game was people who appreciated gameplay and puzzles but didn't care about graphics," he said, possibly citing the last game he ever worked on that didn't emphasize how good it looked. Sweeney read computer magazine articles about better-looking games like id Software's Commander Keen. Those articles said games like that sold 10 to 100 copies a day. Sweeney wanted to make a game that sold like that.
Sweeney's next game, and his first with any graphics meant to impress, was a sidescroller called Jill of the Jungle. It starred a female character, because Sweeney felt that would be an interesting twist. He realized that his art and music skills weren't very good, so he started recruiting people to help him through Internet bulletin boards. He put together a team of four people. "Jill of the Jungle was the game that really put Epic on the map," he said. "ZZT was played by a few enthusiasts. But once I released Jill of the Jungle, we were getting 20 or 30 orders a day. It was a massive amount of money, for a kid in college."
(Sweeney was an active poster on Internet bulletin boards, where he promoted Epic's games and tried to recruit talent. What follows is a message he posted on November 22, 1991:
Are there any game programmers out there interested in working on the next generation of shareware games for the PC? Epic MegaGames develops and publishes them, and we're looking for one or two programmers who have the guts to write super-fast EGA & VGA graphics code and object-oriented game engines. We have two teams working on Nintendo-style action games now. We have several college students involved in our game design teams; all you need is access to a PC and lots & lots of talent! If you have the ambition to work at the state of the art, why not get in touch? You need to mail a sample of your PC programming work then we can talk (ask for a NDA if you need one.) Thanks, and please help get the word out — Epic MegaGames is here!
The success of Jill emboldened Sweeney to call for talented artists and programmers to submit their work to Epic. One of those people was Cliff Bleszinski, future lead designer for Gears of War. Meanwhile, Sweeney was talking to Mark Rein, then at id Software, about id publishing his next game. Rein parted ways with id and tried to team up with Sweeney. On the day Rein flew in to Maryland to Sweeney's parent's house, Bleszinski's submission had just arrived. Rein picked up the Sweeney home phone and called Bleszinski and talked him into signing on to make his game for Epic. Rein promised Bleszinskii that Epic would be a success.
"The thing that sticks into my mind wasn't so much what he was doing on Jill of the Jungle," Rein recalled. "That was, of course, very cool. He's a smart guy and that was technically ahead of everyone else. The more impressive thing was all these other developers who were also really good and making really cool games who had gravitated toward him to be their publisher. James Schmalz, who made Epic Pinball [went on to found Digital Extremes, and then partnered with Epic on multiple Unreal games] There was Cliff Bleszinski. The first day I met Tim he said, we should call this guy. He's making this really interesting game, and he seems like a smart young fella. He had some cool other games like Zone 66. It was just clear he was able to attract other talented people to work with him."
What surprised Rein most was that this programming expert cared about the business side of the business. "When I met him he was reading the Harvard Business Review," Rein recalled. Sweeney had a few people in his basement working a simple assembly line to pack copies of his games. One person made a floppy disc, another made a mailer and passed things along.
Epic made a bunch of games, titles like the non-Sweeney Jazz Jackrabbit and Solar Winds. College and the challenges of starting a growing company pulled Sweeney away from serious programming for a few years. He missed it but eventually rediscovered his talents: "I surprised myself. I thought, ‘I'll never be able to compete with these guys like [id's] John Carmack, who are building these amazing graphics algorithms. But once I started building the Unreal editor and realized I needed to write a little renderer so I could get 3D images on the screen, it turned out I was actually pretty good at that."
The success of id's Wolfenstein 3D and the first glimpses of Doom impressed Sweeney and his small teams at Epic. Those id games weren't sidescrollers. They were 3D. They required bigger teams, Sweeney figured. He needed to pool his best people. Epic needed to make a "Doom-killer."
After a few false starts, in late 1994, the Unreal project began. It was going to be a fantasy game. Sweeney would so some of the programming. They shifted course when they heard about Quake and tried to make it sci-fi.
"It really took shape as a beautiful game," Sweeney said of the first Unreal. "That came from two things: number one was James Schmalz's artwork... and the Unreal Engine 1's technology base." That Unreal Engine was made entirely by Sweeney for 2 ½ years. It was possibly the most important thing Sweeney would ever program and the first draft of one of the most important pieces of evolving code in modern video gaming.
"I looked at Doom and what had been released of Quake and tried to build a significantly higher-quality graphical renderer, paying a lot more attention to lighting, texture quality and other parameters like that. We aimed to use texture maps that were two to four times the resolution of Doom's textures. And that, combined with James Schmalz's artwork, really brought things to the next level."
Initially, Sweeney figured that the Unreal Engine would be a good bit of tech on which Epic could build several games. He'd heard that id re-wrote their tech each time from scratch and that struck him as wasteful. "My idea was that we build this and use it for many games of our own." Then they released some screenshots of Unreal. Other developers called up and asked to use Epic's graphics engine. They wanted an engine? Sure, Sweeney recalled. Epic signed them to licensing deals, their first clients of many.
Unreal looked so good that the elite gaming magazine of the time, Next Generation boastfully put a screenshot of the game on their cover. That was a feat unheard of in an era of gaming graphics that were best left unseen until they were paid for—or at least until the magazine containing them was. The cover, Sweeney recalled, "was an incredible ego trip. That had been our goal all along, to create a game that was just that beautiful."
The engine-licensing was an epiphany. Sweeney and his pals at Epic realized that they could make money by granting other developers the right to use their tech. He loved it, because deals like that made Epic the money it needed to focus on making games. Sweeney's involvement in the first Unreal Engine diminished after the first Unreal game shipped. Other programmers, including Epic's bright AI wiz Steve Polge moved to the forefront.
Polge was the first programmer that Sweeney worked with who he discovered was as good as him. He's still at Epic and a big Sweeney fan since he joined the company in 1997. "We worked side by side on folding banquet tables in a cramped office," he said. "We worked crazy hours, and it's the most rewarding work I've ever done. Tim drove a Ford Probe back then, with the back often filled with old Burger King wrappers. Tim has never been interested in showing off, and still works on a banquet table in his small office among the programmers."
Epic didn't just license its engine. It also got its engine and mod tools out to its gamers. People who played games and people who made games were all customers of Sweeney's efforts. Epic was creating game-making tools with the Unreal Engine that helped make those two pools—players and creators—more and more just one big group.
"That was the DNA," Rein said. "Make the game and put out some tools so the community could keep making it and expand it in ways you hadn't thought of... we couldn't do that so much with the sidescrolling games that followed ZZT, but when we came to 3D, the idea was to develop tools elegant enough that artists could use them. You didn't need to be a programmer to put art in the game like you did in the sidescroller days. You could really give them to artists and they could figure out what to do, and they'd be relatively easy to use."
Sweeney was hands-off on Unreal Engine 2 and believes he did "approximately zero coding" for Unreal Engine 3, the graphics tech that has become ubiquitous in the Xbox 360 and PlayStation 3 era of video games that has stretched from 2005 until now. That engine has run Epic's blockbuster Gears of War games and has been licensed to many of the biggest publishers and developers in the industry, motoring the likes of Mass Effect and Batman: Arkham Asylum. Sweeney shifted to more of a planning and research role for the newer engines, as he has in recent years for the forthcoming Unreal Engine 4. He's already backed off on that one, keeping a high-level eye on it while about 40 people at Epic build it as, Epic and Sweeney hope, the graphical standard of gaming's next generation.
"When you look at what's happening with independent development these days, you've got Engines like UDK and Unity enabling individuals or small teams to create pretty much whatever they can imagine. They can then distribute their games via the internet over Steam, the App store, XBLA, PSN, etc. and become successful enough by doing that to support themselves or form their own studios. Some earn enough to work on the next game, others end up having runaway successes. This isn't really that much of an unfamiliar story anymore. Now contrast that against where Epic started.
"Twenty years ago, Tim Sweeney was 21 years old and didn't have access to any of that. The internet didn't exist, he had to write everything himself, and he was mailing physical copies of his games out to people. To start at that, and grow into one of the most successful Engine developers in the world, with many massively successful games and franchises along the way, and still be at the forefront through several fundamental shifts in technology and in how the industry operated... that's crazy! The fact that I can take it for granted that the Unreal Engine exists, or that Steam and other distribution channels make what I'm doing financially viable, is a real testament to how hard people like Tim and the guys at Epic, or the team at Valve, etc. had to work for that to be possible.
"When you also take into account how many of the successful AAA titles of the last couple of years were running on the Engine he created, I think it's pretty clear how much of an influence Tim has had on the industry, on both the mainstream and independent sides alike."
The Unreal Engine has become a modern version of Sweeney's ZZT. Sweeney says the goal of the first UE was to "create a great editor to create a great game." He says so many other developers lost too much time making games with bad tools. The Unreal Engines, however, were supposed to be easy to use.
"If you're going to try to sell an engine to a big game company like EA or Activision, you know they have smart graphics programmers. You know that any feature we build, if they put the right people on it, they can build the same feature. But what they don't do is they don't make the tools investment. And they don't do the sort of long-term thinking we do and realize that we have to make the game development process as efficient as possible.
"We think long-term at Epic. We're like Intel. We look at what we're going to be doing five, 10 years from now and set our direction that way, whereas most companies are like: 'We need to ship our next game, we'll put all of our resources on this and we'll then worry about the next one.'"
Mark Rein describes Sweeney as opinionated and strong-willed. He says Sweeney is usually right and is quick to say he likes something or hates it, "kind of like a Steve Jobs in that way without all the negative qualities."
Sweeney thinks his greatest achievement isn't what he's programmed but who he has assembled. "We build games with a far smaller team than most of our competitors and we beat them in lots of areas." And the people that came on board early have mostly stuck around. "The fact that we've been able to carry this company and its people and culture forward for so long, that's really unusual in the game industry. I'm really proud of that, and I have something to do with that."
He's done so much and, along the way, he's bought some great cars. About those things, says Polge: "He does have a love for fast cars, and I remember being amused once Epic started becoming really successful and Tim was living in a small apartment in an apartment complex, with a Ferrari, a Lamborghini, and a Porsche parked there."
Rein says Sweeney has gotten rid of some of best cars. "He recently got rid of his last Ferrari, because he said it just wasn't scaring him anymore... He replaced it with a vintage Corvette which just wobbles around the road. I guess at any speed it's scary. That's the thing. He likes to be challenged."
He likes to be challenged? He said he hasn't dismantled a lawnmower in 20 years, but it sounds like there is more to tinker with after all. It seems like there are still Adventures out there for a man who taught himself how to work a computer better than anyone he knew. Might the Hall of Famer have, well, Epic's biggest possible game still on his to-do list?
"We've accomplished so many of the goals I had initially, but the one thing we've really never really done, the goal with Unreal 1 was to have a beautiful game that supported single-player and networking. The network part of Unreal 1 was supposed to be massively multiplayer. Basically, everybody in the world was playing in a single, shared environment and you go between different levels—some created by users—through these teleporters and you'd explore the world seamlessly.
"Crazy idea, right? To ship a 3D MMO in 1998? We ended up cutting that feature a very long time ago."
Sweeney had loved early Multi-User Dungeons and Ultima, those old pioneers of multiplayer gaming. He still wants to make something like that, a game that put all of its users in the same place. "That's something that would be fun. We don't have any concrete plans to actually do that at any point in the future. But that's one of the things that would be a hallmark of a cool gaming experience, to me, to go into a world and everybody and all of your friends are there."