Sid Meier's Memoir!

by Sid Meier's Memoir! (retail) (epub)

  Unfortunately, the rapid third-party expansion wasn’t necessarily great for quality. Before the year was out, the Microplay division would publish ten new games across multiple platforms each. Many of us on the development side felt their standards could perhaps have been a little higher, and we rebelled by referring to our core team as MPS Labs. The term “labs” was being thrown around a lot in those days as computers became more associated with scientific progress, and Bell Labs in particular had recently won several Nobel Prizes for their research, so at the time it felt like a hip way to elevate ourselves a little from the Microplay name. We designed a new logo to be displayed at the beginning of our games, and someone went so far as to take a picture of me in a lab coat and tape it to the main development door, with the stern safety warning that “You Are Now Entering MPS Labs.” We had no shortage of self-esteem.

  Not all the third-party games were iffy, though. As in many industries, there were two types of independent outfits: those who were still knocking on publishers’ doors trying to establish themselves, and those who had proven their talent so thoroughly that they could wait for publishers to knock on theirs. Dan Bunten was one of the latter.

  I had met Dan at my first Computer Game Developer’s Conference a year earlier, but I’d known his name much longer than that. He had been making games since 1978, and by 1983 had already created M.U.L.E., which many consider to be one of the best computer games of all time. Loosely inspired by the Robert Heinlein book Time Enough for Love, the box described it as “a game in which up to four players attempt to settle a distant planet with the so-called help of a mule-like machine they all learn to hate.”

  I’d like to say that at least I beat Dan to the punch on the four-player thing, since Floyd of the Jungle had been released one year before M.U.L.E. in 1982. But in fact Dan’s very first game, Wheeler Dealers, had shipped with a custom four-player controller that he had designed himself. He’d wanted to play a particular type of game, and the fact that it didn’t exist yet was no deterrent—hardware included. It was a feeling I could relate to.

  Meanwhile, the game he signed on to make for us, Command HQ, was one of the first to include head-to-head online play over a modem. Dan was absolutely evangelical about multiplayer, and many of his games shipped without a single-player option at all, even when his publishers begged for one. He felt that the most important thing computers could offer us was a connection with each other, and without it, they were essentially worthless.

  Another thing Dan was well ahead of his time on was gender issues. He felt that more designers should be women, and failing that, more designers should be seeking the input of women—and failing that, more designers should at least have a woman’s influence somewhere in their lives. At one early CGDC, he gave a speech in which he urged designers to get married, have kids, and “stop spending all their time alone in front of computers.” In 1992, he underwent a sex change operation, and became Danielle Bunten Berry. I’m proud to say my fellow designers had a very progressive attitude about it, especially for the era. It was a little awkward at first, but only in the way that a room full of nerds would have been awkward around any woman, and she never suffered any outright rejection that I’m aware of, at least not from us. We were a community of people who, on average, had experienced a fair amount of social rejection ourselves, so perhaps we were more sensitive to the hurt it could cause.

  Pronouns are a big deal these days, to the point that it would be almost impossible to talk about my friend without angering someone. But Dani, as she went by after the surgery, always envisioned herself as a different person in the second phase of her life, and never wanted to erase who she used to be. She often specifically referred to her transition as her “pronoun change,” joked about how she only did it to increase the number of female designers in the industry, and once said of her former self, “I’m not as good a programmer as he was. I’m also not as willing to sit for hours in front of a computer. . . . I tend to need to socialize far more often than he did.” I honor both Dan’s and Dani’s memories by speaking about them the way she preferred.

  In any case, Dan was already a strong voice for equality in gaming before his transition, and by 1989, the American industry was just barely starting to listen—the Japanese and European markets being ahead of us on this one by many years. American gamers tend to mark our own Lara Croft as a critical turning point for female heroines, and her contributions are not to be dismissed, but Tomb Raider wasn’t released until 1992. Nintendo had already cast Samus Aran as a woman six years earlier in Metroid, which itself was notable only because it infiltrated the American market. Plenty of earlier Japanese games had lead characters who openly admitted to being women on the title screen, rather than in the final seconds of the game as Metroid had.

  But just as Dani would have quipped in her soft Arkansas twang, slow progress was better than no progress. Text-based games had been accidentally inclusive for years by asking players to enter their own name, and the adventure game genre inched ahead of its peers again when some began offering a choice between a male-ish or female-ish clump of pixels. Electronic Arts released Murder on the Zinderneuf in 1983 with a selection of six male and two female sleuths, and Atari’s well-known Gauntlet series offered one female action-adventurer out of four. It was something, anyway.

  We had actually discussed the possibility of a female option in Pirates!, but it would have required an alternate set of art for the entire “wooing the Governor’s daughter” portion of the game. There was only so much we could hope to get away with in 1987, even as our own publisher, and a female pirate making the moves on a female aristocrat was definitely not on the list. It would have been fun to animate a fastidious governor’s son getting swept off his feet by a tough and capable piratess, but doubling the romantic content would have meant cutting an equal amount of something else, and we simply couldn’t afford the memory. Someone pointed out that there were female pirates who had lived and dressed as men, so why not offer the choice and then keep the game exactly the same, masculine player and all, but that just seemed like inviting trouble from both sides. So, Pirates! shipped as it was.

  But now we were nearly in the nineties, for Pete’s sake. I was already a couple of months into development on a new title called Covert Action, which we’d pitched to Bill as “like Pirates!, but with spies,” and from the outset I knew that the game was going to be gender-neutral. We fit the extra data into the game by always addressing the character as “Max” Remington, and asking the player at the beginning whether that was to be short for Maxine, or Maximillian. In reality, it was short for nothing: Max Remington III was our lead artist on the project, and he agreed to let us steal his very espionage-suitable name.

  Spies fit the Pirates! framework in the sense that they did a lot of different activities, such as code breaking, clue chasing, and the occasional hostage taking. Even better, it was possible for a spy to identify the bad guy by focusing on the skill he or she happened to enjoy most. Wiretap enough phones, and you’d eventually get the incriminating evidence without sneaking into the building—or you could just strap on a gas grenade and walk through the front door, if that was your preferred method. Either way, the henchmen would still be going home to their families, because all of your weapons were nonlethal, up to and including the rubber bullets in your top-of-the-line spy pistol. Covert Action was the closest I had ever come to making a violent game in the immediate, bodily sense, and I was determined not to cross the line.

  This is not to say that I’m in favor of any form of censorship. Videogames are an art form, and it’s never a good idea to stifle creativity. I can say with personal certainty that gamers are mature and intelligent people, and we have the ability to distinguish between fantasy and reality. But when it comes to the creations that happen to inspire me, I don’t think violence is necessary. The world is often a very negative place, and I’d rather push it in the opposite direction whenever I can. There’s an argument to be made that
by exposing the unpleasant reality of violence, you can inspire others to push against it, too, but this generally requires a removed perspective, rather than the inherent first-person nature of games. It’s hard to claim that our products are immersive, but somehow insist that the experience has no impact. A game with no impact is simply a bad game, and the hand-waving dismissal that anything we make is “just a game” is even worse when it’s coming from a designer’s own mouth. Excessive gore is, at best, a cheap and short-lived path to player engagement. There is a line to walk, an audience to tailor to, and a purpose to consider in everything we do.

  My partner on Covert Action was a designer named Bruce Shelley, who had previously worked at the board gaming company Avalon Hill. It wasn’t unusual for us to hire someone with no background in computer games, since the industry hadn’t been around long enough for anyone to build a worthwhile résumé. Most of our employees began with a particular nondigital expertise, and learned on the job how to integrate it within the structure of computer games. But even with specialized artists, sound engineers, and playtesters, design at that time was still mostly synonymous with programming. For us to bring a designer on board without a coding background meant he had to be pretty darn good.

  Fortunately, Bruce’s talent was obvious to whoever interviewed him. I wasn’t usually involved in hiring—pretty much everyone got a “sure, they seem nice!” appraisal from me, so my input wasn’t very helpful—and I don’t know that I would have been able to pick Bruce out as someone special, in the beginning. He was quiet, and humble. But as he had studiously chipped away at various tasks within our large F-19 Stealth Fighter team, I had begun to notice a core of determination and insight. He was the type to stick with a problem until it was solved. He liked for things to be done right, and no matter what you showed him, he always had an idea for at least one detail that could be improved.

  Like many introverts, we bonded over the things we liked at first, rather than a particular affinity for each other. We talked about television shows, and historical fiction. We played board games in the break room, including several he’d designed himself back at Avalon Hill. After F-19 shipped, Bruce was officially assigned to work on another flight sim, but unofficially, he became my trusted assistant and sounding board, helping me clear away the debris and figure out what exactly this spy game was supposed to be.

  One way that spies did not fit into the Pirates! framework was the inherently linear nature of their story. A pirate could choose to sword fight indefinitely, but a spy can’t just travel the world wiretapping every building he comes across for no good reason. Clues lead from one to the next. You might have a few different ways of gathering those clues, but eventually they were all going to lead to the same evil mastermind. Once the mystery was solved, why would anyone play our game a second time?

  No problem, I thought. We’ll just have the computer write new mysteries!

  It was only a little bit impossible, which is not the same as completely impossible. One early proponent of computational creativity was Christopher Strachey, who had been a college classmate of Alan Turing’s in the 1930s. Their paths diverged for a bit after graduation, but Strachey eventually reconnected with Turing in 1951, after hearing about his new Manchester Mark 1 computer. Strachey later reported to the national conference of the Association for Computing Machinery that his work on Turing’s machine had been a success: the Manchester Mark 1, he declared, “will in fact play a complete game of Draughts at a reasonable speed.” In other words, he’d programmed it to play checkers.

  Shortly after that, the improved Ferranti Mark 1 model was released, and Strachey again pushed it toward artistic, rather than mathematical, purposes. First, he devised a way to alter the pitch of the computer’s usual clicks and grinding, and arranged them into renditions of “God Save the Queen” and “Baa, Baa, Black Sheep.” Then his focus shifted once again in 1952, when he decided that what the Ferranti Mark 1 really needed to do was write love letters.

  Strachey programmed a template that randomly combined a few different sentence structures and word choices within a basic letter format. The results were stilted, but comprehensible. Occasionally the computer even created something approaching poetry, like “You are my covetous burning, my affectionate yearning,” or “My adoration keenly sighs for your infatuation.” Though most of it is not especially romantic by today’s standards, the word list still serves as a fascinating reminder of what passed for terms of endearment in 1950s Britain. “Little liking” and “fellow feeling” are considered synonyms for love, and the list of salutations includes now-bewildering items like “duck” and “moppet” alongside classics like “honey” and “dearest.”

  The industry term for this type of randomized template—or more specifically, the ideal of perfecting randomized templates into actual creativity—is known as procedural generation. Start with Shakespeare’s Hamlet, for example. But instead of Denmark, set it in Africa, and instead of people, make them lions. All of a sudden you’ve got The Lion King, complete with a murderous, usurping uncle and an angsty protagonist who disappears for a while then returns to claim the throne. Or start with Romeo and Juliet, except swap fair Verona for urban New York, turn the feuding families into rival street gangs, and make everyone a little older for decency’s sake. Now it’s West Side Story. The more individual elements you change, the closer you are to a brand-new story that no one recognizes anymore. Give a computer a starting point, tell it what pieces are allowed to change and in what kinds of ways—you can’t swap Denmark with bees, for example, only other locations—and that’s procedural generation.

  Interestingly, the popular children’s game Mad Libs was being invented elsewhere at almost exactly the same time as Strachey’s first attempts at formulaic love letters. Chris Crawford, the founder of CGDC who had once thrust a sword in the air for art, became particularly obsessed with the idea of a computer that could make up new stories, and eventually left the games industry to develop his “Storytron” invention full-time. I was not prepared to go that far, but I did know that without some level of procedurally generated plot, Covert Action would be dead in the water.

  After several months of work, the prototype I had was not dead, exactly, but kind of treading water with its boots on. Bruce and I had come up with about twenty or thirty crime story templates that could feature different bad guys, cities, and shadow organizations each time. It was enough variety to keep the casual player happy, but wasn’t the breakthrough I’d been hoping for. The patterns were recognizable after a while, the templates too predictable. The very nature of a fill-in-the-blank story meant that everything outside the blanks was set in stone, and no randomized piece of information could have any effect on the ones that came after it. Usually the answer to this sort of problem is just more data—more templates, more swappable elements, longer lists to choose from. But even if we’d had the computer memory to spare, which we didn’t, the result still wouldn’t have satisfied me. I wanted a story that was laid out from the beginning, but not apparent until the end, like a Sherlock Holmes mystery.

  To be honest, I’d imagined this whole project as a technical warm-up to a better game along those exact lines, and I’ve never stopped dreaming about it to this day. How cool would it be to have a game that could sneak in that one crucial piece of evidence in the beginning, just waiting for you to deduce its meaning? Not from a preplanned list of clues, like a worn keyhole or a dirty shoeprint, but from a more general understanding of what is normal, and what would therefore not be normal. You’d have to set down rules for the real world, establish all the cause-and-effect we take for granted, and then follow up with a nearly infinite ruleset, laying out the consequences of a break anywhere in the chain, and how those breaks would affect each other . . . anyway. Still dreaming, like I said. The point is, Covert Action wasn’t it.

  Bruce could sense my energy for the project flagging, and I think he felt the same. We both knew it was an okay game, but not a great game, a
nd it probably never would be. At the same time, my wife Gigi had just become pregnant with our first child, and I was experiencing the usual priority shifts and personal reevaluation that all new parents go through. I was contending with The Future, and it was making The Past look sort of universally rusty.

  But it was hard to admit defeat when I still believed wholeheartedly in the idea’s potential, if not its current execution. I had walked away from failed prototypes before, but never after spending this much time and energy on one. Plus, it wasn’t just my own time we were talking about anymore. When I created games alone, I had only myself to apologize to if something fizzled out, but Bruce had been working alongside me from the beginning, and I didn’t want to feel like I had dragged him into anything unfairly.

  I wanted to drop it. I didn’t see how I could drop it.