Free Fall Associates

Jon Freeman (designer) and Anne Westfall (programmer) are best known for "Archon," a combination of chess and simple arcade action that broke a lot of new ground (longtime collaborator Paul Reiche had a hand in there too). "Archon" for the Atari 800 was one of the first releases from Electronic Arts.

Jon, with Paul Reiche and Robert Leyland, also designed "Murder on the Zinderneuf," an open-ended mystery game that makes the "play it once" nature of many modern games seem painfully strict. Jon's name has graced many games before and after these two, including "Swords of Twilight" for the Amiga.

Anne is on the board of directors of the Computer Game Developers Conference.

Jon: Anne worked as a programmer for Disco-Tech, a small division of Morton Technology, a civil engineering firm. She developed the first microcomputer-based program designed to help lay out subdivisions. I did word-processing part-time to supplement an uneven income as a free-lance writer. A science-fiction novel and a book on board games eventually led to consulting work for GAMES magazine and several game companies.

Jon: "Starfleet Orion" was at least one of the first published war games–and certainly the first SF or space war game–written on or for a microcomputer: to wit, an 8K Commodore PET. It was a game of strategy and tactics, not an action game; there were nine scenarios of widely varying complexity but no computer opponent. That came with the sequel, "Invasion Orion." The two players had to alternate turns at the computer, which resolved each round of combat after each player had entered commands for the turn.

"Orion" came about because Jim Connelley, the DM of a D&D game I played in, bought a PET to help him with the bookkeeping chores required of a Dungeon Master. After a while, he wanted to write off the purchase and thought the best way to do so would be to create a computer game he could sell. When he got bogged down, he cast about for some help. I knew very little about computers but a lot about games. Learning I had written a book on board games and was contributing articles to GAMES magazine, Jim recruited me to take over game development, while he handled the programming, in BASIC. When–somewhat to our surprise–we ended up with a finished product, we had to start a company to sell the game. In 1978, just in time for Christmas, we formed Automated Simulations–later called Epyx–and I found myself established in a new career in computer games.

Anne: When I went back to school in the 1970s, I did well enough in my programming classes to get hired by a civil engineering firm to write a surveying program on the TRS-80. I was at the West Coast Computer Faire in 1980 to demo my program, and our booth happened to be next to the Automated Simulations booth, where I met Jon Freeman.

Jon: After dating for six months, I persuaded her to move closer and go to work for Automated Simulations. At the end of 1981, we decided to leave office politics and start our own game-development firm, Free Fall Associates.

Anne: When we left Automated at the end of 1981 to start Free Fall Associates, part of the reason was to get into assembly-language programming and to work with the Atari 800, which at that point was the most appealing platform. With a complete set of Atari documentation and a book on programming the 6502 (the microprocessor used in the Atari, the Apple II, and the C64), I learned the machine and taught myself assembler. Low-level programming, at least on the 6502, made perfect sense to me, and the Atari was very logically set up and very easy and fun to work with.

Jon: Since the notion of three lives, a standard feature of arcade games of that period, had always struck me as artificial and rather silly, we began playing around with game designs in which nobody got killed. Partly just to show it could be done, we wanted to create a few maze games that weren't "Pac-Man" clones. The subject matter came from spending too much time doing our taxes, and the details were there partly to illustrate the absurdities of the tax code. The layout was designed to combine the Atari 800's scrolling capabilities with the feel of a board game.

Jon: Unfortunately, while it was mechanically simple enough for anyone to play, you almost had to be an adult to grasp the humor, and not enough adults used Atari 800 computers. It became a hit with some other developers, but the average young gamer just didn't get it. Mainstream adults loved the idea, and we got written up in some major business magazines, but the market wasn't there yet. Now that there's a PC in every office, a Windows version could be a hit.

Jon: There were three principal sources of inspiration. A fantasy chess set I admired at a science-fiction convention pitted Conan, as I recall, and various heroic types against assorted villains and monsters. A large-scale, live-action chess game put on by the Society for Creative Anachronism at the first Renaissance Pleasure Faire in northern California required the differently equipped "pieces" to fight for possession of challenged squares. The brief holographic game scene in Star Wars suggested the graphic appeal of the concept.

How did the design evolve during development? Was it fairly solid before the programming started or were lots of changes made as the game progressed?

Jon: We had a pretty good notion of what we wanted to do from the beginning, although the graphics, sounds, and many of the details (each piece's name, placement, strength, weapon speed, etc.) evolved as the programming allowed us to playtest combat. There was one major exception: "Archon" was designed to be played by two people. At the last minute, relatively, EA decided to include a single-player versus a computer opponent option. In retrospect, this was absolutely the right thing to do, but at the time, in the face of an implacable deadline, the uncertain results of so much unforeseen work terrified us. Nonetheless, we somehow managed to develop, from scratch, the entire strategic and tactical AI in a single month, and the game shipped on time.

Jon: Six months. The same madly hectic six months in which Robert Leyland, Paul Reiche, and I developed "Murder on the Zinderneuf"–which missed the same deadline by a couple of weeks.

Jon: The day EA was incorporated, Trip Hawkins called us because he saw our ad in a trade journal. Ironically, the ad existed only because the magazine gave us free space to jump-start its new classified section, and we had then forgotten about it. In fact, due to a production error, our copy of the issue in question was missing the page with our ad! Since we knew nothing of EA, Trip, or the ad he cited, we were a bit skeptical about the whole thing but agreed to meet in person.

He pitched himself and his new company, and we pitched the two products we most wanted to do: "Archon" and "Murder on the Zinderneuf." His new company's direction sounded good to us, and his offer of development money for both our proposals sounded even better. By modern standards, it was a pittance, but no other publisher seemed willing to risk a penny on anything short of a finished product. After some extremely congenial negotiations, we signed EA's first two contracts, and the rest is at least what passes for history in the software biz.

Jon: The bigger surprise was how elaborate the effort was: our pictures on the products were not mere snapshots but elaborately staged productions; the "We See Farther" poster alone required assembling a whole group of developers–one from Arkansas and the rest from all over California–with a rock-album photographer flown up from Hollywood for an all-day photo shoot at a studio up in San Francisco.

Jon: Playing a complete game against each other as a comparative test of skill was actually fairly unusual. We were far more interested in testing the computer's skills, tracking down bugs, verifying fixes or improvements to particular subroutines, or tweaking pieces by replaying the same matchups over and over. One of the ironies of this business is that you rarely get to play your own games; you're always playtesting.

What led from "Archon" to "Adept"? Were there any flaws in the original that you were trying to correct?

Jon: Trip Hawkins wanted a sequel–initially just the same thing with slightly different pieces or powers. I didn't think that would be sufficient. We held off until we could come up with something that had a similar mix of action and strategy, a sufficiently similar feel to please "Archon" fans, but different enough that nobody would feel ripped-off by a simple "me too" product.

We didn't regard the original as flawed, but there were areas we wanted to improve. The computer opponent, for instance. Mostly, we wanted to make the pieces more interesting and the weapons more varied, the same thing we did later with "Archon Ultra."

Jon: Not well. We think it would have done better if it had been out six months sooner, but we had a lot of trouble finding animators, and the publisher persisted in changing the specifications every other week. Problems with the DOS extender and minimal marketing finished it off.

Jon, what made you decide, in "Murder on the Zinderneuf," to take the more difficult "murderer is different each time you play" approach rather than having one set outcome?

Jon: The biggest single reason was its genesis as an attempt to re-create the experience of playing "Clue"–one of my favorite games–by imagining what "Clue" might have been like had it been designed originally for a home computer. A variable mystery was inherent in the notion. Neither Paul Reiche nor I was an enthusiast of adventure games, and maximizing play value seemed to us to imply replayability.

Jon: We created eight distinct scenarios or master plots, each of which had between two and four independent subplots: e.g., a love triangle, blackmail, secret agent vs. spy, even a vampire. When a scenario was played, one subplot was randomly selected to be the real plot–the one involving that session's murder. The others automatically became red herrings. On a different occasion, one of the red herrings might become the real plot, and vice versa.

Additionally, each subplot had three or four roles and typically four sets of alternative casts to play them: for instance, philanderer, jealous spouse, lover, lover's jealous spouse. By reversing sexes and extending a "spouse" role to include a Significant Other, quite a range of characters could occupy the roles at different times. Further, within the same subplot, any role could be victim or murderer: the jealous spouse might kill the lover or the straying spouse–or be killed by either.

Half the dialogue was based on the nature of the characters, permanent relationships, and the general flavor of the scenario; the other half was assigned to specific roles. Since many statements referred to a character who might be male or female, alive or dead, even the simplest lines could get complicated: e.g., "He/she hates/hated her/him." Since we made heavy use of tokens to adjust verb tense and pronoun gender, the resulting encoded dialogue bore only a slight resemblance to normal English; it was hard to read, harder to proofread, and almost impossible to debug thoroughly.

Are you surprised that more games haven't used similar open-ended techniques?

Jon: I used to be. We tried hard to disguise our methodology, because we were initially concerned that many people would copy it. Now I think it's just too tricky for most people to tackle. It requires planning, care, and the ability to juggle an uncomfortable number of variations, and the work can only be shared by people who really understand the system. It helps to have designers with a grasp of programming and programmers who understand game design. Game companies have focused almost exclusively on hard-core gamers, who seem to prefer a forty-hour game to a half-hour game that can be replayed 100 times. "Normal" people have different preferences, but nobody except Microsoft seems to pay any attention to them.

Anne: Moved on to other machines. Long before the Atari died, we were doing games for the Commodore 64, and we were among the earliest Amiga developers. The Atari experience was very useful even later, because the Amiga owed a lot to the design of the Atari 800, and the game machines through the Sega Genesis shared a lot of similarities with it.

Anne: Yes, but not as much as I'd like! For the past three years, most of my time has been devoted to the Computer Game Developers' Conference and the Computer Game Developers' Association.

Anne: Essentially it's gone from a being almost a private club for an in-crowd of a few hundred to a major event for thousands of professionals in the business. Although it's still casual and friendly and a great place to make contacts or see old friends, it grew too big to be run by a group of volunteers and far too time-consuming for what were supposed to be part-time directors. It also became too much of a financial risk for a handful of people without the resources of a large corporation. That's why, at the end of 1995, we sold the event to a company–Miller Freeman–with the resources and the experience to do it justice.

Jon, in the early 1980s, you were outspoken about computer game plagiarism. Have things improved or worsened in regard to the stealing of game designs?

Jon: Out-and-out plagiarism–pixel-for-pixel copying of entire games–has disappeared except in the shareware fringes; there's no money in it. Getting games on the shelf is so expensive nobody can afford to ignore rip-offs, and the profits of plagiarism tend not to survive a lawsuit. On the other hand, me-tooism is epidemic. Although breakthrough designs and games that are different–"Populous," "Lemmings," "Sim City," "Myst," even the "Microsoft Entertainment Packs"–tend to be the ones garnering the most critical acclaim and popular success, publishers almost universally want only games that are just like what everybody else is doing.

Jon: In most cases, the current state of game design is not significantly different than it was ten or fifteen years ago. Graphics are far more elaborate and realistic; programming is far more sophisticated; budgets and memory requirements have skyrocketed; simulations are more detailed; but it would be hard to argue that, in general, pure game design has advanced appreciably in the last ten or fifteen years. The game design elements of such popular games as "Doom" and "Descent" are not significantly different from those of Automated Simulations' dungeon adventures way back in 1979-1981.

Jon: I suppose there is something to be said for neutering sales and marketing people, but I'm not sure that's what you meant!

Seriously, I would like to change the suicidal overreliance on "multimillion-dollar would-be blockbusters" and the rampant me-tooism. The combination eliminates diversity, which limits our audience, and discourages originality, which makes breakthrough hits unlikely. Publishers wrongly defend this as a cautious, conservative strategy; instead, it's a recipe for financial disaster that guarantees ruin for many companies.

Jon: My favorite card game is "Oh Hell," and my favorite board games remain "Clue" and "Diplomacy." My favorite coin-ops were definitely 8-player "Tank"–a really old machine–and "Gravitar." On the home front, electronically, my favorites are probably "M.U.L.E." and "Krazy Shootout" on the Atari 800; "Archon" in all its incarnations and sequels, if I can include any of our own games; "Garrison," a "Gauntlet" clone on the Amiga; "Golden Axe" and "Herzog Zwei" for the Sega Genesis; "Star Control 2" for 3DO; "SimCity," and "X-Com."

Anne: "M.U.L.E.," "Zak McKracken," "Free Cell," "The Horde," "Star Control 2," and of course "Archon."

Jon Freeman&Anne Westfall

Jon Freeman
&
Anne Westfall