My "Next" Games: Families, Psychology, and Murder

"What's your ideal game? If you could work on any kind of game you wanted to, what would your next game be?"

They're classic interview questions, of course, and in my role as a columnist I also hear them fairly frequently from readers and the press. I don't have an ideal game; that's a little like asking, "What is your ideal movie?" or "What is your ideal piece of music?" The interactive medium is capable of an infinity of games, and no one of them can possibly be ideal. The most elegant action game I've ever seen is Tetris, but that doesn't make it my ideal game or even my ideal action game. The action game I've most enjoyed playing is Tempest, the old Atari coin-op, but I wouldn't say that it was "ideal" either.

The second question, however, is more interesting. If I could work on any kind of game I wanted to and time, money, and resources were no object, I would design a sweeping fantasy epic of ancient India, filled with gods and demons, battles, intrigue, and high adventure. It would be a combination of RPG and wargame, and you could play it from several different perspectives: as the wise old general, the naïve young king, the evil vizier, the king's tyrannical mother, his brave older sister, his bitter younger brother, and so on. It would have many opportunities for great daring, noble sacrifice, evil plotting and grand strategy. There would be courage and compassion, treachery and terror, heartwrenching moral dilemmas and moments of high exhilaration. Its aim would be to touch your heart and to bring the magical world of Indian folklore to life in your mind. It would also cost at least five million dollars to do properly, and take two to four years to build. It's unlikely that any but an Indian company would take the risk to fund it, and I would certainly have to do a lot of research. Educating myself about the subject would be one of the chief joys of designing the game.

However, that's just the game I would build if I had all the money in the world. Even then it would be fairly traditional, and would break new ground chiefly in the area of its setting and subject matter, not in its overall goals. The average gamer would understand what it was about and how to play it pretty quickly.

If I were to design a truly new game, I would try to make it something never before seen at all. You'll notice that in the game I described above, the player would be given the opportunity to play one of several different members of a family or their close confidants. The family is a central structure in all our lives, dominating our emotions and profoundly influencing the way we live. It's a human universal: although the family has many, many different structures around the world, every single human being on the planet was cared for (with greater or lesser amounts of attention, skill and love) as a young child by someone. During that period we each formed deeply-held expectations about familial roles and obligations, both for ourselves and those around us.

I think there's a good reason that the gods of ancient Egypt, Greece, Rome, and many other places were all related to one another. The gods could all have been independent, self-created beings, each living alone and supervising his own areas of responsibility. But polytheistic gods are humans by another name, and to appeal to the human soul they need human relationships. And our ongoing fascination with the British royal family is due as much to the fact that they are a family as to the fact that they are royal. Quiet, untroubled royal families like those of Sweden or the Netherlands don't interest us much; it's those tempestuous British royals with their pomp and ceremony, their petty infighting and their sex scandals that grab our attention. Democracy undoubtedly makes for better, fairer governments than monarchy, but as entertainment, democracy is as dull as ditchwater. In America we have two cable TV channels devoted to the government, but I would bet cash money that if America had a channel devoted to the British royal family it would get more viewers than both C-SPAN channels put together - even though the British royal family has nothing to do with America.

This is also why the Mafia continues to be such a enduring subject of American fiction - witness the success of The Sopranos. It's about families and their relationships. Criminal families are great sources of dramatic tension because they have a lot of conflicts that don't occur in normal families: the temptations of large amounts of ill-gotten money, the constant danger of being caught, the pressure from the police to inform on other members of the family, homicidal infighting over power and territory, and so on. Of course any criminal gang is going to face those problems, but they don't have issues of family loyalty and obligation to complicate matters further.

We seldom see families at all in computer games, much less the kind of difficult familial interactions that I have just been describing. Family relationships have long been explored in literature, but are very little explored in interactive entertainment. If I were going to try to truly break new ground in this medium, this is where I'd start.

In the game that I would design, you would be playing the role of a forensic psychologist. The gameplay would consist of a series of one-on-one interviews with the members of a large and highly dysfunctional family, all of whom are colluding to cover up the murder of one of them by another. Not a criminal family, just one in which someone has finally snapped and killed someone else - goodness knows it happens often enough. For whatever reason, there would not be enough physical evidence to solve the case. The object would be to untangle the facts of the crime, in effect prying apart the family's story (or stories). In order to do this, you would have to understand the relationships among the members, and those would not be simple. Even in healthy families, most people's feelings are multilayered, and change with time and circumstances. An event as catastrophic as a murder is bound to bring a lot of issues to the surface - and cause a lot more to be suppressed in the face of police scrutiny. Only by uncovering their real feelings about one another would you be able to arrive at the truth.

You'd play the game by asking questions and listening to the answers. You could interrupt an interview at any time and switch to another member of the family, if one person's answer raised a question for another one. There would be clues and red herrings, and of course a great many lies. Some lies might conflict with others, which would let you know you were on the right track, if you were paying enough attention to notice it. Certain members of the family would be stronger or more level-headed than others, and some approaches would work better with some than with others. You would have to observe everyone's reactions very carefully, and perhaps take notes.

Ideally, the screen would show a view of whoever you were interrogating at the moment, with a great many subtle facial expressions and body language for each person. If you were watching closely enough, they would give you clues about what effect your questions were having on the person's state of mind. If the project didn't have much money we could implement this with a large number of photographs of real actors. If money were no object, we could create a 3D model of a person with fully implemented facial expressions such as Jeff Lander described in his Gamasutra article "Flex Your Facial Muscles." Unfortunately, 3D rendered people still don't look much like real people (they're too symmetrical and their movements are too mechanical, among other things), so that might harm the effect somewhat.

Each person's replies would be pre-recorded sound bites, or possibly artificially-generated speech if the technology has gotten that far. At the moment, however, we're even farther behind at generating the subtle nuances of tone in real speech than we are at generating subtle facial expressions. Another and far cheaper approach, although somewhat lacking in subtlety, would be to do the whole game in text. It would print out what the person said, what tone he used, and how he looked when he said it. However, because words are clearer and more direct than body language, the clues about each person's inner feelings could be a little too easy to spot. If that proved to be the case, it might be better to stick only to questions and answers, like a chat conversation. You would have to pick up cues from the person's vocabulary, a bit like reading a Shakespeare play.

The questions themselves would be asked by assembling a sentence out of words from a menu, and as time went on and you learned more, the menus would grow so that you could ask about more and more things. As for the answers, obviously it would be wonderful to develop an entire psychological model of a person with full sentence-generation capabilities, but that's the work of a lifetime, if not several. In any case the goal of the project is not to develop new technology, but to create a new kind of experience for the player, regardless of how that's achieved. Although it's technologically uninteresting, I would probably just treat each person as a very large finite state machine. Asking the same question would not always elicit the same answer, however, because each person's "state" would change as the game progresses. The response you get would depend on when you ask the question. If you ask, "Did you kill your brother?" right at the beginning, the suspect is bound to shout "No!" Later on, once you've delved into his psyche a little, you might get a different answer: "No… but I wanted to."

Now, undoubtedly there are a few pragmatical types reading this who are thinking, "Why build this? Nobody would ever buy it." The game as described doesn't have much replayability, either, but I don't think that's important. This isn't supposed to be a commercial product; it's an experimental project to test the boundaries of what our medium can do. We need these kinds of tests, and as we're now starting to see academic programs devoted to game development, I hope more of them will appear. The nice thing about academic research is that it's not intended for sale; it's intended to expand our understanding (although if it has commercial applications, so much the better).

There's no shortage of games that are mysteries of one kind and another, but most of them are solved by physical exploration and physical evidence. I haven't yet seen anything like what I've described. My goal would be, to put it rather romantically, to allow the player to explore new mysteries, the mysteries of the human heart.

______________________________________________________