Designer's Notebook: Interstate'76 and the Principles of Harmony

I write this column to discuss the principles of good game design, and ordinarily, I wouldn't devote a whole one to a single game. But every now and then, a game comes along which so beautifully illustrates a design principle that it deserves special attention. Activision's INTERSTATE '76 is one such game.

In his 1997 Game Developers' Conference lecture, "Listen: The Potential of Shared Hallucinations," legendary game designer Brian Moriarty (who had nothing to do with INTERSTATE '76, by the way) described the principle of harmony. Harmony is the feeling that all parts of the game belong to a single, coherent whole. Brian explained the idea so well that with his permission, I'm going to quote the entire section of his lecture. Read his words carefully, because this is rich, heady stuff:

Harmony isn't something you can fake. You don't need anyone to tell you if it's there or not. Nobody can sell it to you, it's not an intellectual exercise. It's a sensual, intuitive experience. It's something you feel. How do you achieve that feeling that everything works together? Where do you get this harmony stuff?

Well, I'm here to tell you that it doesn't come from design committees. It doesn't come from focus groups or market surveys. It doesn't come from cool technology or expensive marketing. And it never happens by accident or by luck. Games with harmony emerge from a fundamental note of clear intention. From design decisions based on an ineffable sense of proportion and rightness. Its presence produces an emotional resonance with its audience. A sense of inner unity that has nothing to do with what or how you did something, it has something to do with why. Myst and Gemstone both have harmony. They have it, because their makers had a vision of the experience they were trying to achieve and the confidence to attain it. They laid down a solid, ambient groove that players and their respective markets can relate to emotionally. They resisted the urge to over-build. They didn't pile on a lot of gratuitous features just so they could boast about them. And they resisted the temptation to employ inappropriate emotional effects. Effects like shock violence, bad language, inside humor. You know, the suspension of disbelief is fragile. It's hard to achieve it, and hard to maintain. One bit of unnecessary gore, one hip colloquialism, one reference to anything outside the imaginary world you've created is enough to destroy that world. These cheap effects are the most common indicators of a lack of vision or confidence. People who put this stuff into their games are not working hard enough. Harmony is a principle too often overlooked in the creation of today's computer games. There are any number of reasons why a game might lack harmony: a dispersed development team, divided leadership, a poorly-conceived vision.

INTERSTATE '76 avoids these pitfalls, and it manages to do so even though it's set in a game universe unlike any other seen before. For those of you unfamiliar with the game, it takes place in an "alternative" 1976 in which the gas crisis has deepened to the point of chaos. Criminals with heavily armed muscle cars are wreaking havoc in the American southwest, and it's up to you and a few other auto-vigilantes to stop them. The game consists of a series of missions in which you, as the hero Groove Champion, with the help of your friends Taurus and Skeeter, must track down the evil auto-villain who killed your sister.

INTERSTATE '76's degree of graphical and thematic harmony is extraordinary. The cars, the clothes, the music, the language, and every other aspect of the game closely resembles the 1970s TV shows which inspired it. The maps are crude pencil sketches on the backs of paper bags. Even the game's necessary bookkeeping screens - which are so often clunky, fantasy-killing text in other products - are presented within the game world. The options menu appears on a truck stop waitress' order pad, and the save/load menu on a credit card receipt. No effort is spared to make you feel like a guy driving around the desert in a hot rod.

The other thing that's extraordinary about INTERSTATE '76 is that it's not set in the usual fantasy or science fiction world. INTERSTATE '76 is at the same time both derivative and original: derivative because its look and sound all come from 1970s television, and original because no other game has tapped that rich (if hilariously tacky) vein of culture. Here we have a game whose creators dared to eschew the traditional orcs and aliens, and create something completely new for the jaded gamer... and yet still completely familiar.

Rather than speculate about how they managed to achieve this feat, I decided to go to the source and ask. Sean Vesce was the game's director, and although he's no longer at Activision, he agreed to talk to me about it.

Q: INTERSTATE '76 seems like every teenaged boy's dream game: driving like a maniac and shooting at things. But setting it in the 1970's slants it to an older crowd. How did the idea for the '70's theme come about?

A: Zack Norman (Mech Warrior 2, INTERSTATE '76, Interstate '82) and I were at lunch shortly after shipping Mech Warrior 2. We were talking about what kind of game we wanted to do next. Zack was in the market for a car at the time and was reading a local Auto Trader magazine. He stopped on a page with several classic muscle cars, and I asked, "is that the car?" He replied, "No, that's the game!" The rest developed from there. We came up with several key design goals over the course of lunch, the most important being "to make a game with soul." This was one of the main driving forces throughout the design, and I think this is what helped us make such a distinctive game.

Q: Did you ever specifically discuss the target audience with marketing, sales or the executives? Or did you just decide to make the game and let the chips fall where they may?

A: For the most part we were intent on making a game for ourselves - 18- to 30-year-old males who owned a high-end PC and were looking for something unique to play. We also wanted to capture people who played and enjoyed Mech Warrior 2, who understood how to play a sim, but we wanted to give them a more compelling experience by weaving a great story through it.

At the time we were not experienced enough to understand the changes that were happening around us in the PC market and how that would affect the audience and their taste for games. It's really clear in hindsight that there was a whole new wave of users coming into the market. Spurred by very affordable powerful PCs and a lust for hunting deer, this new majority was looking for something less heady, more mainstream, and much easier to play. This has had a profound effect on the state of game design and production.

Q: Game companies tend to be pretty conservative these days, sticking to well-established futuristic or fantasy worlds. How did Activision's management and marketing departments react to the idea of a game set in the '70s?

A: Our first pitch meeting was disastrous. The vice president of production hated the idea. After an hour of pitching, we ended up following him all the way to the elevator as he was trying to leave for the evening. We didn't stop until the doors shut. In the end persistence paid off. Once management adopted the concept, the entire company got behind the product.

Q: What do you think changed their minds? And did you do anything specific to convince them, like create conceptual drawings, high-concept documents, or a prototype?

A: A couple of events happened which helped us sell the concept. First, we got Scott Lahman, the Director of Creative Affairs on board. He really helped us position the concept in such a way that management could understand its potential for success. Second, we drafted a preliminary design document which included tons of conceptual drawings, excerpts from the script that Zack had begun to write, examples of game play, and preliminary schedules and budgets. This helped tremendously by enabling us to process and organize our ideas and develop a solid plan for its execution. Third, we hit on the idea to re-use the Mech Warrior 2 engine to cut down on production time. Together, these forces began to turn the heads of those who were on the fence and in charge of "green-lighting" projects. We were given a small budget to hire some folks to begin work on a prototype of the game. When we completed and demonstrated the prototype to about forty or so managers, sales, marketing, executives, we hit the jackpot. They loved it.

Q: INTERSTATE '76 is visually more consistent than most games, because the mission briefing movies use highly polygonal characters who look as if they could be in the game engine. Still, this must have been expensive to design and render. Did you ever consider using live actors and real video shoots?

A: An original goal of the project was to experiment with blurring the line between the interactive and non-interactive portions of the game. To achieve this we needed to maintain a consistent graphical style throughout. Originally the design called for all cinematics to be generated real-time using the game's engine (remember, this was a big deal back then). When schedule restraints and technical limitations prevented us from doing this, we decided to do the next best thing: using low-poly models and in-game assets (car models, textures etc.) to create pre-rendered cinematics to be streamed off the CD.

We ruled out live actors and real video shoots for several reasons. First, it went directly against our goal of achieving stylistic consistency throughout the game. Second, live actors and real video shoots are extremely expensive, time-consuming, and require skills that we as game developers did not directly possess. Third, and most important, our team was strongly opposed to full-motion video and live actors in video games. It was bad enough to be working in the midst of the great "Hollywood / Silicon Valley" convergence, we didn't feel compelled to encourage it.

Q: Everything in INTERSTATE '76 seems to be part of a coherent whole: the music, the art, the voices, the opening movie... right down to little details like the fact that the weapons display is done with a Dymo labelmaker. Why do you think more games aren't like this, and what do you think was different about the way INTERSTATE '76 was developed that made it possible?

A: The tight cohesion that you see in INTERSTATE '76 was achieved by the total commitment of the team in paying attention to the details. It was our mantra. We set off to create an amazing but believable universe and to do this, even the smallest detail had to be designed properly to fit within the whole. It was two parts passion, one part insanity, and three parts fear (of making a lame game).

Q: How big was the team, in terms of numbers of people doing each kind of work? Were they all in the same office?

The team grew to include about 23 internal core team members. Everyone was split into one of four teams including art led by Rick Glenn, design led by Zack Norman, programming led by Dan Stanfil, and production led by Scott Krager. The breakdown was roughly six artists, six designer/mission implementers, seven programmers, three production people, and myself. We were lucky enough to get an entire hall of a new section of office space that Activision had rented (the internal studio was expanding at an insane rate around the time we went into production). We each sat two to an office and had a central meeting space that was off to one end of the hall. Because the entire team was located in the same place there was a general sense being one coherent whole. This was a very important component in establishing the collaborative atmosphere that was key to creating a game like INTERSTATE '76. (On Mech 2, we weren't so lucky. We had the programmers, designers, and artist each sitting in different areas of the building, leading to lots of headaches and miscommunication.)

Q: Did you establish any formal procedures or guidelines (like a "bible") to help keep the design in harmony? Were design decisions made hierarchically, or collaboratively?

A: While we made every attempt to foster collaboration, we established a creative hierarchy to ensure focus of vision and a coherent design. According to our hierarchy, the final responsibility for creative decisions fell to the lead designer and then to the director. While this guaranteed that in a dispute, creative decisions would still be made, it became difficult for some at times. Tough decisions had to be made to discard some really cool ideas, which during the heat of production can sometimes be a bitter pill to swallow.

Q: Licensing is a major issue for many games today. Although all the cars that appear in the game have fictitious names, it's pretty clear from their appearance that they're modeled on real cars: the "ABX Leprechaun" is obviously an AMC Gremlin, and so on. Did you ask for the manufacturers' permissions to use the real thing and get turned down, or did you even try?

A: At its core, INTERSTATE '76 was a tribute to the spirit of American muscle cars of the '70s. It was important for us, however, to create our own makes and models rather than directly replicating and licensing "real" cars (we were after all creating an "alternate" 1976!). By doing this, it allowed us to do anything we wanted to the cars like denting them, blowing them up, mounting heat-seeking missile systems to their hoods, and a bunch of other things that we would not have been able to do had we licensed "real" 1970s muscle cars.

Q: Taurus may be the only black character with a central role in any computer game today. Any thoughts on this? Tell us more about Taurus's character, and the decision-making about the characters in general.

A: Developing a strong cast of characters was an important production goal. Our team was not satisfied with earlier attempts made by game developers to tell story through their games, and we wanted to prove that it could be done. Zack had a background in writing and was keenly aware that at the heart of any great story are great characters. When we began early development of the story, he had already come up with the beginnings of Taurus, Groove, Jade, and Skeeter. They were really caricatures of popular '70s television and movie icons. Taurus was a very special character. He was the embodiment of the attitude and soul we were trying to inject into the game. He was Shaft, Superfly and Samuel L. Jackson all rolled into one. Plus he was a poet. One of my favorites was:

Q: I notice that your opponents don't seem to cooperate very much with each other. It seems to me that they could get together and try to herd you into a box canyon, for example. What can you tell me about the other drivers' AI?

A: The AI was a difficult problem to solve. Since we created a fully functioning auto combat simulation, the AI drivers had to deal with a lot of stuff. At its highest level, the AI had overall mission strategies like "destroy the Gas 'n Stuff," or "defend Fort Davis." This was managed by AI state tables created by the designers. From here things got tricky. We set off to create the most realistic auto combat simulation ever made. To do this we created a complex automotive physics system that dealt with gear ratios, engine sizes, tire pressure, surface friction, and a myriad of other real world calculations. To achieve as much realism as possible, the AI drivers had to know how to deal with these things just like the player. We wanted to limit any AI "cheating." Blow out an AI driver's front right tire, and it has to know how to compensate to keep his car driving straight. Bump an AI driver into the gravel, and it has to know how to properly negotiate the change in surface and navigate back onto the road. Add to this the need for the AI drivers to manage and use their weapons and ammo properly, and you begin to see what a nightmare we set ourselves up for.

Q: The game is more aggressive about strong language than most are these days, and much more aggressive than any '70s-era television show. Was there any discussion of this during the design process?

A: The adult language was really a part of the "coherent whole" we talked about earlier. It was important for us to portray these characters as they would really act. Groove and Taurus were auto-vigilantes who lobbed heavy munitions from the roofs of their cars. These guys didn't have time for subtle pleasantries. Surprisingly, this was not a hotly debated issue. Management understood and encouraged our attempts to make a game with attitude. In the end, only a small portion of the content was edited.

Q: In retrospect, is there anything you would have done differently, or would like to have had more or less of in the game?

A: If could turn back time and do it again, the first thing I would do is make the game more action-oriented with easier controls, clearer mission objectives, and greater and more frequent rewards. I would greatly simplify the game's shell, most importantly the vehicle repair screen that players accessed between missions. I talked to a lot of people who were completely baffled the first time they went through. It was far too complex and difficult to use without the manual. I would split some of the missions into multiple missions (especially the first one), and offer a way to save at any point in the game. On the multiplayer side, I would ensure greater security against cheating and hacking, as this really ruined the multiplayer experience for a lot of people. The last thing I would do is demand more time to balance the overall difficulty and learning curve for the game. We never really had the chance to do this properly and I think the overall game was diminished by this.

Q: What are you working on nowadays?

A: After shipping INTERSTATE '76, I decided to move back to Northern California, where I grew up. I took a three-month vacation and realized I still had the bug to do more giant robot stuff (left over from Mech 2). I met up with a very special team of programmers and artists working out of Accolade in San Jose. They had been working on a next-generation hardware-accelerated 3D engine and were looking for a killer concept. We came up with Slave Zero, a giant robot action shooter due out in spring of 1999. It's a wild game where players assume the role of a 60-foot giant "slave" robot battling it out in a mega-city of the future. It's kind of a cross between Mech Warrior 2 and Virtua-On. Be sure to check out http://www.slavezero.com for the latest developments.