Interview with Warshaw
GCG: How did you get started in video game development?
Warshaw: In college, I worked on some of the first microprocessors. From there I went directly to large scale software development at HP and I was really bored. I was also kind of a zoo case at HP since I was much more vocal and expressive than most of the other programmers there. One day a coworker explained to me how he would occasionally tell his wife "Howard stories," and how she'd always say that sounds just like the wild and crazy people where she worked, a small company in Sunnyvale called Atari. I interviewed there and they didn't want to hire me because they thought I was too straight. But I knew that I needed to work there. I needed to get back to microprocessors, I loved the idea of making games and I also knew I was right for that environment. So I literally begged and pleaded with them to give me a try. Ultimately they relented and the rest (as cliche as it sounds) is history.
GCG: In the early days of video game development, many games were developed solely by one person as opposed to today where hundreds of people may be involved with a single game's development. Can you talk a little bit about the development of Yar's Revenge and the process of taking the game from idea to finished product?
Warshaw: Yar's was originally assigned to me as a conversion of the coin-op game Star Castle. After investigating this for a little while I came to the conclusion that this conversion would suck on the VCS system. So I did something no one else had ever done, I went to my boss and said that I had an idea for an original game that would use the same basic play principles of Star Castle but was designed to fit the VCS hardware so it wouldn't suck. And to their credit, they let me go with it. Think about that. They blew off a license to let me pursue an original concept with the promise of making a better game for the system. That would never happen today.
So I started working on the play and the controls, and honestly I got it to a point where it really sucked. It sucked because the controller scheme I chose was too close to the coin-op version. I did that because there were too many control functions I needed to accomplish with the primitive VCS joystick. When I realized I needed a better control scheme I relaxed my assumptions that the controller had to do everything, and that's when I hit upon a very fundamental game design concept. When you replace a controller action with a play action the game gets better. In Yar's Revenge, I needed a way to call up the Zorlon cannon. So instead of using the controller to activate it, I made the rule that you had to eat a brick or touch the Qotile to activate the Zorlon canon. By converting the controller action to a play action the game was massively improved. The more you let a player earn the things they need rather than just button for them the deeper the game gets. At least that's my opinion.
Also, the game still wasn't Yar's at that point. It was just an abstract game play. It became Yar's when I found out marketing was looking for a name for the game and the ones they were looking at sounded pretty lame to me. So I made up the name "Yar's" and wrote a whole back story to support it. That was the first back story ever written for a video game and Yar's was the first game published with a comic book detailing the back story. In fact all the events that went into naming Yar's Revenge and the implications thereof is a whole other story. It's a tale of marketing intrigue and adventure, which I can share another time if you like.
But it wasn't just the comic book and the back story. There were a lot of firsts in Yar's Revenge that became industry standards. Reset from the joystick, pause mode, full screen explosion, marketing approved Easter eggs, elaborate cues and a death sequence are among the industry firsts established in Yar's. Of course it was easier to do new things back then because so much hadn't been done yet. Back then the focus of development was innovation. We were trying to set standards rather than maintain them.
And once the game was done, then came the testing. Yar's had more testing than any other game, ever. This was mainly because there was a manager there who thought the game was critically flawed. So they kept commissioning tests. And each time Yar's would do great in the test and then they would commission another test. Finally, after Yar's beat Missile Command in a play test they released the game and it went on to become Atari's biggest selling original game. I learned a lot about making games and the games industry during Yar's Revenge and the fact that so many people enjoyed it has always made me very happy. Even more so than E.T.
GCG: What do you think of the team development process? Is it harder to innovate as a single member of a larger team?
Warshaw: Back in the day a video game development was a work of authorship, now it is a collaborative effort. Neither is better or worse, they each have plusses and minuses. All other factors being equal, the team development process can get a lot more done than an individual. Obviously any project can be ruined by a bad team or a lame individual, but if you compare good team games to good individual games, the team games should kick enormous ass. This presumes that in both cases you are implementing a good idea. A major part of team development is the design team. And team games need to be a lot more spec'd out at the start than do individual games. Also, once you get that juggernaut rolling it is much harder to change directions along the way. No game should ever be an exact replica of its original design, but a team game will likely wind up a lot closer to it starting concept than an individual game. This makes sense because a team game needs to be a much larger concept. It should take an individual at least 10 or 15 years to do a modern team game. That's not likely to be a viable dev cycle.
Is it harder to innovate as a single member of a team? Yes. The only way to truly be innovative on a team game is to be part of the design process and get the team to adopt an innovative design from the start, because the rest is mostly game mechanics and will rarely constitute fundamental innovations. So since the design must be set in hard clay (not quite stone) and lots of resources committed to it you are very unlikely to innovate after the design is approved for production. Software and art development are big tasks and once begun it is very difficult to change their direction in any significant way. A project is more likely to be cancelled than fundamentally changed and continued. So your window for innovation is smaller (only during the design phase) and you have to convince others to accept your innovation rather than just go ahead and test it out yourself. It's much harder to innovate in a team environment.
But the biggest killer to innovation today isn't team size, it's financial expectation. Games cost a lot of money to make these days so the cost of failure is huge. That means the amount of risk people accept is smaller. And whenever you talk about innovation you are talking about risk. It takes a new concept to break out and make a runaway hit, but most new concepts fail miserably. That is why you see narrowcasting in games. People keep repeating the same tried and true formulas because that is what sells. Just like Hollywood blockbusters recycle the same 4 or 5 story lines, the video game industry recycles it's 4 or 5 game plays that are the perennial sellers. They are trying to minimize their risk because the cost is so high. And just like you see innovation in low budget independent films, you see innovation in lower budget games for PCs and online. But you won't see that on big console games because the financial pressures overshadow the creative possibilities.
Team development is necessary today and fortunately it yields some truly amazing games. For me personally I prefer a work of authorship because it's just plain more fun to explore and develop.
GCG: There's so much emphasis on what experience is necessary to get into the game industry, but so little thought given to the experience one takes away from the game industry. What have you learned from your time in the game industry?
Warshaw: I've learned many things over the years. Many lessons I gathered at Atari took years to digest. I did the whole documentary series "Once Upon ATARI" to clarify them in my own mind and share them with others. One key thing I've learned is that creative industries start out as hotbeds of wild innovation with exciting dynamic environments and inevitably over time (if they succeed) they mature. This means developments get larger, slower, more compartmentalized and the goals shift from freshness to profitability. The decision making power shifts from the creators to the management and from engineering to marketing. And I've learned that this isn't a bad thing or the end of the world, it is rather the natural order of business. It is very frequently the case that the CEOs who create successful startups get replaced by classic managers as the company grows. Large companies in established industries are a totally different entity than small start-ups in new industries. It takes different types of people to make them successful. I was able to start in the beginning of the game industry and see it transition all the way through. The biggest change over time in terms of what it takes to get into the industry is this: originally it was about breadth, now it is about depth. I happen to be a person who prefers breadth so I miss that. But team development is still frenzied now just like individual development was frenzied then, now it's just happening in a bigger hive with more specialized bees.
GCG: Do you like to play any video games? If so, which ones?
Warshaw: My favorite games of recent years are the Grand Theft Auto series, and not for the storyline either. I think the game play is brilliantly designed. I really believe that GTA3 will stand as the first real crossing point between 2D style gaming and immersive VR. I see design innovations in those game unparalleled for at least a decade before and in many cases unmatched today. That was the first time I saw a 3D environment really used to its potential. I could go on and on about the significance of GTA, but let's just say I totally respect the series as a rare design innovation from which all gamers stand to learn a great deal. And of course they are milking it now, but they've earned the right. I also like the Jack and Daxter series for different reasons. They came up with a truly beautiful and aesthetically pleasing game to play and then they took it in totally new directions, and I think they did it well. I also still like a lot of the classic coin ops. Defender, Robotron and Millipede are some of the all time greatest for my money.
Today, a video game development team is made up of various specialized roles: programmers, artists, designer, producers, sound engineers, etc. Howard makes a ton of great points about the development process which we'll come back to in future articles, but for now, let's get on to some more immediate topics.