I suggested MicroProse instead, because it seemed to me that computer code was just as elegant as any literary prose, and it made a nice double entendre with the word “pros.” Bill thought it was a little hard to pronounce, but agreed that it was distinctive enough to be remembered. Turns out it wasn’t quite as distinctive as we’d originally thought, because years later we would be sued over the name by a company called MicroPro, makers of the WordStar word processing program. Though we were arguably better known, their company slightly predated ours, and it was looking like we were going to have no choice but to change the name—eventually. Bill, being just as tenacious as he was enthusiastic, managed to affably drag on negotiations for years, until the plaintiffs suddenly changed their own name to WordStar International instead, and the whole issue was dropped. I’m not sure anyone but Bill could have done it, but that was just one of his many talents. Somehow he could stonewall a person in a way that made them feel glad for the opportunity.*
In the beginning, the sales calls Bill made were based mostly on convenience. If he had a business trip out of town, he’d walk from the train to the nearest computer store to try to sell a few copies. On weekends, he’d load a box of disks into the trunk of his car and drive as far as he could get down I-95, coming back just in time for Monday morning meetings at General Instrument.
Then late one evening, my phone rang.
“Sid, I think we might be onto something here.”
“Bill? Where are you?”
“New Jersey. We just sold fifty copies of Hellcat Ace.”
“Hey, that’s great!”
“Yeah,” he said. “So start copying,” was the unspoken implication.
Every individual sale back then translated to about sixty seconds of boredom in front of my matched pair of floppy drives, making copies of the game one by one. I could try to read a book, but getting work done was impossible—multitasking wouldn’t be a feature in home computers until about a decade after I needed it. Outsourcing, on the other hand, was just making its debut, and it didn’t take long before I hired one of the younger members of our users’ group to make copies for me at twenty-five cents a disk. He and I were close because he was too young to drive and I was giving him rides, so it’s possible his first job ever was making the modern-day equivalent of thirty-nine dollars an hour at a videogame company. Not a bad gig.
In the meantime, I finished my helicopter game, Chopper Rescue, as well as another game I’d been working on called Floyd of the Jungle. At Bill’s suggestion, I added an opening screen to all three games that advertised the rest of the MicroProse “catalog,” and copied the new version over whatever disks we already had in stock. Several more tweaks went in over time as Bill received feedback from store managers, so if any of our originals still survive, they’re probably all different from one another.
Floyd of the Jungle box art.
© 1982 MICROPROSE, WWW.MICROPROSE.COM
Even though it wasn’t an aircraft game, Floyd of the Jungle was usually the hook Bill used to get the attention of store employees. It offered multiplayer competition against their coworkers, which few could resist, and had elements of what would eventually be known as the platformer genre. Similar to bestselling arcade titles like Space Panic (1980) and Donkey Kong (1981), this style seemed to resonate with players on a deeper, more intuitive level: somehow, everyone knows that being at the top of the screen is better than being at the bottom, and if there’s a damsel in distress, you have to rescue her. Flying the Grumman F6F took a little practice, but Floyd made sense immediately, and didn’t require the players to be especially good—they only had to be on par with each other to have fun. With two or three employees crowded around the screen, shoppers would soon become curious, and once Bill handed his joystick over to a customer, the store was almost guaranteed to buy.
Other games had multiplayer, of course, including Hellcat Ace and Chopper Rescue, but only for two people. Floyd was special because it allowed all four joysticks to be used at once, something very few small-market games could boast in 1982. The one major release with four-player capability was Asteroids, which had been developed by Atari to showcase their own machine’s capabilities. Technology manufacturers often had to take a “build it, and they will come” approach, and they never really knew if a feature would prove popular with developers until they’d already invested a few million in providing it. Sadly, four-player functionality would remain on the fringe for a long time to come, appearing in only a handful of arcade titles like Gauntlet (1985) and Teenage Mutant Ninja Turtles (1989), and no mainstream computer games until Doom exploded onto the scene in 1993.
Chopper Rescue’s multiplayer was unique in its own way, though, because instead of controlling two identical characters, the joysticks were linked to different aspects of the same helicopter. Just like in many real-life military vehicles, one player navigated while the other fired weapons, which required plenty of cooperation and communication. A lot has been lost from those early years of technology, but I’m willing to claim this was the first videogame to give simultaneous players different tasks, at least until someone says otherwise.
Chopper Rescue was also when I figured out how to scroll in all directions. Most of my early games were inspired by new programming tricks—either learned, or developed from scratch—that I then found a way to build a game around. The big advancement in Hellcat Ace, for example, was a way to tilt the horizon more efficiently and accurately than other games. Changing the angle of a line may not seem like much in today’s terms, but it’s a lot harder when you’re doing it on a computer whose entire memory could hold roughly three chapters of this book in plain text format.
Meanwhile, Floyd of the Jungle contained several advancements in one. Aside from managing four active players on the screen, it also included a new technique for animation that involved switching back and forth between slightly different versions of the same character. The hottest game on the market that year, Space Invaders, had used this approach to display a total of six alien types, wiggling their legs in a loop of just two positions each. But the code left room for considerably more, and I wanted to test its limits with as many creatures moving in as many different ways as possible. Aliens had been done already, so I settled on a rainforest backdrop instead, and only then did I think back to the countless Saturday mornings I’d spent watching George of the Jungle cartoons. Unlike later in my career, topic was still secondary to technique—I was making games for the computer, not games that could be put on a computer if necessary, and I wanted to utilize every available feature.
Part of the experiment was seeing how much I could cram on the screen without slowing down the game, and part of it was practicing my illustration skills, since most designers in those days had to be a one-stop shop. I maxed out the code with four-stage images of birds, elephants, crocodiles, snakes, lions, monkeys, and Pygmies (a peaceful set of tribes, in reality, but stereotypes of the day presented them as a formidable challenge for explorers, and it didn’t occur to me at the time to question this received wisdom). Then there was the lovely damsel in distress, Janice, and of course Floyd himself, who had a separate resting animation in addition to all of his running, jumping, climbing, and dying moves. My monkey was a little lumpy, like those animal crackers you can never quite identify, but the crocodile and elephant were downright artistic. This was good news for MicroProse, because it would be another three years before we could afford to hire a real artist.
We started running ads in October 1982, and six months later we finally received our first review, in which the Atari-themed Antic magazine declared Floyd of the Jungle to be both “enjoyable,” and “very good.” Reviews back then were pretty light on adjectives in general. The next month, they covered Hellcat Ace, which was “effective” but “could be improved.”
Bill didn’t mind the lukewarm assessment, though. For one thing, the writer repeated one of his favorite promotional lines, “playtested by members of an Air National Gua
rd Wing,” which was just a slick way of referring to Bill and a couple of his friends. But the truth was his plan never really hinged on the review’s content anyway. He just needed it to exist.
As soon as the articles were published, Bill began placing calls to hobby stores that were farther than driving distance away.
“Hello, I’m looking to buy a copy of Hellcat Ace.”
“Hmm, I don’t think we carry that one—”
“What?” he would fume. “What kind of computer store are you? Didn’t you see the review in Antic?” Then he would hang up in a huff, muttering about taking his business elsewhere.
A week later he would call again, pretending to be somebody else. And a third time a week after that. He didn’t even have to call from different numbers, since caller ID was still as imaginary as Dick Tracy’s Apple Watch.
Finally, on the fourth week, he’d use his professional voice. “Good afternoon, I’m a representative from MicroProse Software, and I’d like to show you our latest game, Hellcat Ace.” Spurred by the imaginary demand, they would invite him in.
It seems utterly transparent in today’s marketing-savvy world, but in the era of mom-and-pop computer stores, it worked. Bill may very well have placed a call to every single outlet in the nation at that time, charming them with his energy and enthusiasm. He and I were the perfect combination, because I had no interest in sales, and he had no interest in the creative side. I could sit at home and program all night, he could get out every weekend and sell, and we never got in each other’s way.
* Achievement Unlocked: Books Don’t Come with a Demo Mode—Time to buy this thing.
2
ADAPTATION
Tic-Tac-Toe (1975) * The Star Trek Game (1979) * Hostage Rescue (1980) * Bank Game I (1981) * Bank Game II: The Revenge (1981) * Faux Space Invaders (1981) * Faux Pac-Man (1981) * Formula 1 Racing (1982)
“ADAPTATION” IS SUCH A flattering word. So much nicer than “copyright infringement.” For 63 percent of these titles it really was an honest adaptation, sometimes even at the request of the property owner. The other 37 percent, okay, I was maybe slightly infringing on an existing trademark. But all it got me was a few bucks’ worth of sales and some free skeptical looks. Crime doesn’t pay, kids. (Unless it serves as inspiration and practice for a lifelong career in one of the most rewarding industries on the planet, in which case it pays fine, both monetarily and spiritually.) My well-intentioned plagiarism also earned me a mild reprimand at General Instrument, where the words “game designer” were decidedly absent from my job title. But long before I was getting in trouble at work for making games, I was getting in trouble at college for making games.
When I entered the University of Michigan in 1971, I had never even seen a computer in person before, but the ultra-logical nature of them was intriguing to me. So, on a whim, I signed up for a programming class alongside my physics and math double major, and by the end of the year, I was a computer science major instead. This decision greatly improved my employment prospects, I realize now, but that didn’t factor in much at the time. Mostly, I did it because computers felt empowering. I couldn’t calculate pi to 10,000 digits—or at the very least it would take me a long, long time—but I could write a program that could. The ability to say, in relatively few words, “Do something cool,” and then have that cool thing pop out the other side, was unbelievably exciting. I wouldn’t even say it was magical. It was technological, and that was better than magic.
Our class learned on an IBM 360 mainframe, programming in FORTRAN on eighty-column punch cards. We would prepare our deck of cards, bring them to the room that held the computer, and watch a staff member feed them into the card reader one by one. Then maybe ten minutes later, we’d walk up to a different desk to collect our results. The good old days were yet to come; these were still very much the bad old days.
Part of my scholarship at the university involved a work-study program to offset tuition costs, and after completing my one programming class, I boldly took a job with a professor who wanted some computer work done. It was a bit of a gamble to claim I was qualified for it, but not many students were in those days, and fortunately the work turned out to be pretty simple. Most of it was early explorations of educational software, like multiple-choice tests that could branch into different questions depending on your answers. But the equipment that Dr. Noah Sherman had in his lab was far more advanced than the stuff offered to second-year students like me. I now had access to a real teletype terminal, which allowed programs to be entered directly into the system without any punch cards acting as middleman. I could examine my broken output, correct the code, and verify the improved output on a much shorter cycle than before. Dr. Sherman could sense my enthusiasm, and he encouraged me to try out my own experiments on the machine after my work was done each day. He even left me with a key to the lab while he was away in Italy one summer.
By then, I was immersing myself in every computer-related topic I could find, most especially this new thing they were calling artificial intelligence. Precise instructions for a computer could be complicated enough to plan out, but teaching a computer how to make its own decisions, maybe even to learn from its mistakes, was on another level entirely. Alan Turing had famously called for an imitation of social behavior as the ultimate goal for a thinking computer, but I thought the more interesting prospect was a computer that could outsmart a human. Not just a math workhorse, but one that could predict my behavior, and be clever about what to do with that information. I wanted a computer that could model complex future possibilities, and eliminate undesirable outcomes until it had settled on the ideal course of action. In short, I wanted a computer that could game.
The classic starting point, I thought, would be tic-tac-toe—and history backed me up on this, though I didn’t know it at the time. In 1950, just two years after Turing’s invention of the stored-program computer, a man named Josef Kates had built a twelve-foot-tall behemoth he called “Bertie the Brain,” which stood on display at the Canadian National Exhibition and beat all comers at tic-tac-toe. (Historians often distinguish between this and the first videogame, Tennis for Two, because the latter used a video screen for its display rather than simple lightbulbs.) Other engineers created independent versions of tic-tac-toe during the 1950s, and eventually followed them with operational renditions of checkers, blackjack, and even chess. Most recently, in 1975—the same year I was attempting to teach myself the tenets of gaming AI—a group of students at MIT had built a mechanical tic-tac-toe machine out of Tinkertoys, which was surprisingly similar to the original totalizator with its gears and piano wire. It would have been super helpful to know all this, but without the advent of the internet, I was largely isolated in my educational pursuits. So I plowed ahead on my own, without the benefit of others’ wisdom.
The lab was mine as long as Dr. Sherman explored the Italian hilltops, so I put the hours to good use and worked every day on my self-assigned project. First, I created a simple text input scheme that allowed you to enter one move at a time. I hadn’t figured out how to get the computer to display its next move on the screen yet, so instead I instructed it to send the grid to the nearest printer, which was stored in a separate room and shared by everyone in the building. I’d go over, collect my printout, come back to my desk, and enter my next move. It was slow, but at least I got some exercise. (If only I had known that games forcing you to walk around would be all the rage forty years later.)
After the third or fourth document containing nothing but Xs and Os, the woman running the output desk was on to me.
“Wait a minute!” she said, snatching back the paper she’d just handed over. “What do you think you’re doing? Computers are not for games!”
I had no satisfactory answer to give her, since it seemed clear to me that was exactly what they were for.
“I’m going to have to report you,” she scolded, already looking up the details of my account on her own terminal. She located the name and contact i
nformation of my supervising professor, and for a while I was afraid that my dream of a gaming computer would be cut short before it could even finish one round of tic-tac-toe. Dr. Sherman hadn’t given me specific permission to do what I was doing, and maybe he would agree that it was frivolous. I might even be banished back to the world of punch cards.
Fortunately, he vouched for me, once they located him by phone on the other side of the Atlantic, and graciously told the staff that I had blanket permission to continue for the rest of the summer in whatever capacity I saw fit. I doubt he had any idea what he was setting in motion, but I was grateful.
After graduation, I began working for General Instrument, and was once again given access to technology I could never have afforded on my own. The sixteen-bit Nova minicomputer—“mini” being relative, in this context—was considered a top-of-the-line machine because its processor was contained on a single printed circuit board, with no spaghetti wiring coming off the back. It was housed in a cabinet the height of an eighth-grader and cost more than a new car, and not only did I have one for my own personal use, but so did most of my coworkers. In addition, all of the minicomputers in our office could talk to each other directly, rather than being hobbled en masse to a central mainframe. We had a network.
Like the university teletype, GI’s business machines only supported plain text, no graphics. But I wasn’t the first to face this dilemma. As far back as 1865, even before the invention of the typewriter, Lewis Carroll was giving the publisher of Alice in Wonderland instructions on how to lay out their movable typeface in a way that drew pictures with the story itself. After the typewriter became widespread, so-called “artyping” exploded as a hobby, and newspapers around the country paid cash to reprint complex portraits and landscapes drawn one character at a time. In 1963, the practice went digital, after the publication of an official binary code for text known as the American Standard Code for Information Interchange, or ASCII. Typewriters hung around for another two decades, but the new acronym took hold, and from that point on pictures made with text characters were commonly known as ASCII art.
Sid Meier's Memoir! Page 2