Fundamentals of Game Design: Game Worlds

This chapter is from the book 

The Dimensions of a Game World

Many different properties define a game's world. Some, such as the size of the world, are quantitative and can be given numerical values. Others, such as the world's mood, are qualitative and can only be described with words. Certain properties are related to one another, and these groups of related properties are the dimensions of the game world. To fully define your world and its setting, you need to consider each of these dimensions and answer certain questions about them.

The Physical Dimension

Video game worlds are almost always implemented as some sort of simulated physical space. The player moves his avatar in and around this space or manipulates other pieces or characters in it. The physical properties of this space determine a great deal about the gameplay.

Even text adventures include a physical dimension. The player moves from one abstract location, usually called a room even if it's described as outdoors, to another. Back when more people played text adventures, the boxes the games came in used to carry proud boasts about the number of rooms in the game. Gamers could take this as a very rough measure of the size of the world they could explore in the game and, therefore, the amount of gameplay that the game offered.

The physical dimension of a game is itself characterized by several different properties: spatial dimensionality, scale, and boundaries.

Spatial Dimensionality

One of the first questions to ask yourself is how many spatial dimensions your physical space will have. It is essential to understand that the dimensionality of the game's physical space is not the same as how the game displays that space (the camera model) or how it implements the space in the software. How to implement the space and how to display it are separate but related questions. The former has to do with technical design, and the latter has to do with user interface design. Ultimately, all spaces must be displayed on the two-dimensional surface of the monitor screen.

These are the typical dimensionalities found in video games:

• 2D. A few years ago, the vast majority of games had only two dimensions. This was especially noticeable in 2D side-scrolling games such as Super Mario Bros. (see Figure 4.1). Mario could run left and right and jump up and down, but he could not move toward the player (out of the screen) or away from him (into the screen). Two-dimensional worlds have one huge advantage when you're thinking about how to display them: The two dimensions of the world directly correspond to the two dimensions of the monitor screen, so you don't have to worry about conveying a sense of depth to the player. On the other hand, a number of games with 2D game worlds still use 3D hardware accelerators for display so that objects appear three-dimensional even though the gameplay does not use the third dimension. Two-dimensional worlds may seem rather old-fashioned nowadays, but there are still many uses for them in casual browser-based games and smaller devices such as low-end mobile phones.
  

  Figure 4.1 , the classic 2D side-scrolling game

• 2.5D, typically pronounced "two-and-a-half D." This refers to game worlds that appear to be a three-dimensional space but in reality consist of a series of 2D layers, one above the other. StarCraft, a war game, shows plateaus and lowlands, as well as aircraft that pass over obstacles and ground units. The player can place objects and move them horizontally within a layer with a fine degree of precision, but vertically an object must be in one plane or another; there is no in-between. Flying objects can't move up and down in the air; they're simply in the air layer as Figure 4.2 depicts.
  

  Figure 4.2 , with plateaus and lowlands visible

• 3D. Three true dimensions. Thanks to 3D hardware accelerators and modeling tools, 3D spaces are now easy to implement on hardware that supports them. They give the player a much greater sense of being inside a space (building, cave, spacecraft, or whatever) than 2D spaces ever can. With a 2D world, the player feels as if he is looking at it; with a 3D world, he feels as if he is in it. 3D worlds are great for avatar-based games with exploration challenges, such as the Prince of Persia series (see Figure 4.3). Most large games for personal computers and consoles now use three dimensions, but many small casual games still need only two.
  

  Figure 4.3 a fully 3D environment

• 4D. If you want to include a fourth dimension for some reason (not counting time), implement it as an alternate version of the 3D game world rather than an actual four-dimensional space. In other words, create two (or more) three-dimensional spaces that look similar but offer different experiences as the avatar moves among them. For example, the Legacy of Kain series presents two versions of the same 3D world, the spectral realm and the material realm, with different gameplay modes for each. The landscape is the same in both, but the material realm is lit by white light while the spectral realm is lit by blue light, and the architecture is distorted in the spiritual realm (see Figure 4.4). The actions available to the player are different in each realm. The realms look similar but are functionally different places governed by different laws. In the movie version of The Lord of the Rings, the world that Frodo inhabits while he is wearing the Ring can be thought of as an alternate plane of reality as well, overlapping the real world but appearing and behaving differently.
  

  Figure 4.4 's material (left) and spiritual (right) realms. Notice how the walls are slightly twisted in the spiritual realm and the overlay indicator is different.

When you first think about the dimensionality of your game space, don't immediately assume that you want it to be three-dimensional because 3D seems more real or makes the best use of your machine's hardware. As with everything else you design, the dimensionality of your physical space must serve the entertainment value of the game. Make sure all the dimensions will contribute meaningfully. Many games that work extremely well in two dimensions don't work well in three. Lemmings was a hit 2D game, but Lemmings 3D was nowhere near as successful because it was much more difficult to play. The addition of a third dimension detracted from the player's enjoyment rather than added to it.

Scale

Scale refers to both the absolute size of the physical space represented, as measured in units meaningful in the game world (meters, miles, or light-years, for instance) and the relative sizes of objects in the game. If a game is purely abstract and doesn't correspond to anything in the real world, the sizes of objects in its game world don't really matter. You can adjust them to suit the game's needs any way you like. But if you are designing a game that represents (if only partially) the real world, you'll have to address the question of how big everything should be to both look real and play well. Some distortion is often necessary for the sake of gameplay, especially in war games; the trick is to distort the scale without harming the player's suspension of disbelief too much.

In a sports game, a driving game, a flight simulator, or any other kind of game in which the player expects a high degree of verisimilitude, you have little choice but to scale things to their actual sizes. In old 2D sports games, it was not uncommon for the athletes to be depicted as 12 feet tall to make them more visible, but nowadays players don't tolerate a game taking such liberties with reality. Serious simulations need to represent the physical world accurately.

Similarly, you should scale most of the objects in first-person games accurately. Fortunately, almost all first-person games are set indoors or within limited areas, seldom larger than a few hundred feet in any dimension, so this doesn't create implementation problems. Because the player's perspective is that of a person walking through the space, objects need to look right for their surrounding area. You might want to slightly exaggerate the size of critical objects such as keys, weapons, or ammunition to make them more visible, but most things, such as doors and furniture, should be scaled normally.

If you're designing a game with an aerial or isometric perspective, you might need to distort the scale of things somewhat. The real world is so much larger and more detailed than a game world that it's impossible to represent objects in their true scale in such a perspective. For example, in modern mechanized warfare, ground battles can easily take place over a 20-mile front, with weapons that can fire that far or farther. If you were to map an area this size onto a computer screen, an individual soldier or even a tank would be smaller than a single pixel, completely invisible. Although the display will normally be zoomed in on one small area of the whole map, the scale of objects will have to be somewhat exaggerated so that the objects are clearly identifiable on the screen.

Games frequently distort the relative heights of people and the buildings or hills in their environment. The buildings are often only a little taller than the people who walk past them. (See Figure 4.5 for an example.) To be able to see the roofs of all the buildings or the tops of all the hills, the camera must be positioned above the highest point in the world. But if the camera is positioned too high, the people are hardly visible at all. To solve this problem, the game simply does not include tall buildings or hills and exaggerates the height of the people. Because the vertical dimension is seldom critical to the gameplay in products such as war games and role-playing games, it doesn't matter if heights are not accurate, as long as they're not so inaccurate as to interfere with suspension of disbelief.

Designers often make another scale distortion between indoor and outdoor locations. When a character walks through a town, simply going from one place to another, the player wants the character to get there reasonably quickly. The scale of the town should be small enough that the character takes only a few minutes to get from one end to another unless the point of the game is to explore a richly detailed urban environment. When the character steps inside a building, however, and needs to negotiate doors and furniture, you should expand the scale to show these additional details. If you use the same animation for a character walking indoors and outdoors, this will give the impression that the character walks much faster outdoors than indoors. However, this seldom bothers players—they'd much rather have the game proceed quickly than have their avatar take hours to get anywhere, even if that would be more accurate.

This brings up one final distortion, which is also affected by the game's notion of time (see the section "The Temporal Dimension" later in this chapter), and that is the relative speeds of moving objects. In the real world, a supersonic jet fighter can fly more than a hundred times faster than an infantry soldier can walk on the ground. If you're designing a game that includes both infantry soldiers and jet fighters, you're going to have a problem. If the scale of the battlefield is suitable for jets, it will take infantry weeks to walk across; if it's suitable for infantry, a jet could pass over it in the blink of an eye. One solution is to do what the real military does and implement transport vehicles for ground troops. Another is simply to accept a certain amount of distortion and create jets that fly only four or five times as fast as people walk (StarCraft uses this trick). As long as the jet is the fastest thing in the game, it doesn't really matter how much faster it is; the strike-and-retreat tactic that jets are good at will still work. Setting these values is all part of balancing the game, as Chapter 9, "Gameplay," discusses in more detail.

Boundaries

In board games, the edge of the board is the edge of the game world. Because computers don't have infinite memories, the physical dimension of a computer game world must have an "edge" as well. However, computer games are usually more immersive than board games, and they often try to disguise or explain away the fact that the world is limited to help maintain the player's immersion.

In some cases, the boundaries of a game world arise naturally, and we don't have to disguise or explain them. Sports games take place only in a stadium or an arena, and no one expects or wants them to include the larger world. In most driving games, the car is restricted to a track or a road, and this, too, is reasonable enough.

Setting a game underground or indoors helps to create natural boundaries for the game world. Everyone expects indoor regions to be of a limited size, with walls defining the edges. The problem occurs when games move outdoors, where players expect large, open spaces without sharply defined edges. A common solution in this case is to set the game on an island surrounded by water or by some other kind of impassable terrain: mountains, swamps, or deserts. These establish both a credible and a visually distinctive "edge of the world."

In flight simulators, setting the boundaries of the world creates even more problems. Most flight simulators restrict the player to a particular area of the real world. Because there are no walls in the air, there's nothing to stop the plane from flying up to the edge of the game world; when the player arrives there he can clearly see that there's nothing beyond. In some games, the plane just stops there, hovering in midair, and won't go any farther. In Battlefield 1942, the game tells the player that he has left the scene of the action and forcibly returns him to the runway.

A common solution to the edge-of-the-world problem is to allow the flat world to "wrap" at the top, bottom, and sides. Although the world is implemented as a rectangular space in the software, objects that cross one edge appear at the opposite edge—they wrap around the world. If the object remains centered on the screen and the world appears to move beneath it, you can create the impression that the world is spherical. This is used to excellent effect in Bullfrog Productions' game Magic Carpet. Maxis's Spore actually displays the world as a sphere on the screen, not just a wrapping rectangle (see Figure 4.6).

Finally, you can solve the problem of boundaries by requiring the player to move among defined locations. For example, you might let a player fly from planet to planet in the solar system by clicking on the planet she wants to go to. The player cannot go beyond the boundary of the solar system because there are no planets in interstellar space. The user interface for movement creates a natural limit that requires no further explanation.

The Temporal Dimension

The temporal dimension of a game world defines the way that time is treated in that world and the ways in which it differs from time in the real world.

In many turn-based and action games, the world doesn't include a concept of time passing: days and nights or seasons and years. Everything in the world idles or runs in a continuous loop until the player interacts with the game in some way. Occasionally, the player is put under pressure by being given a limited amount of real-world time to accomplish something, but this usually applies to only a single challenge and is not part of a larger notion of time in the game.

In some games, time is implemented as part of the game world but not part of the gameplay. The passage of time creates atmosphere and gives the game visual variety, but it doesn't change the game's challenges and actions. This usually feels rather artificial. If the player can do exactly the same things at night that she can during the day and no one ever seems to sleep, then there's little point in making the distinction. For time to really support the fantasy, it must affect the experience in ways besides the purely visual.

Baldur's Gate, a large role-playing game, is a good example of a game in which time is meaningful. At night, shops close and the characters in the game run an increased risk of being attacked by wandering monsters. It's also darker and hard to see. Taverns are open all day and all night, which is reasonable enough, but the customers don't ever seem to leave and the bartender never goes off shift. In this way, the game's use of time is a little inconsistent, but the discrepancy serves the gameplay well because you can always trade with the bartender and pick up gossip no matter what time it is. The characters do need rest if they've been on the march for a long while, and this makes them vulnerable while they're sleeping. In the underground portions of the game, day and night have less meaning, as you would expect.

Variable Time

In games that do implement time as a significant element of the gameplay, time in the game world usually runs much faster than in reality. Time in games also jumps (as it does in books and movies), skipping periods when nothing interesting is happening. Most war games, for example, don't bother to implement nighttime or require that soldiers get any rest. In reality, soldier fatigue is a critical consideration in warfare, but because sleeping soldiers don't make exciting viewing and certainly aren't very interactive, most games just skip sleep periods. Allowing soldiers to fight continuously without a pause permits the player to play continuously without a pause also.

The Sims, a game about managing a household, handles this problem a different way. The simulated characters require rest and sleep for their health, so The Sims depicts day and night accurately. However, when all the characters go to sleep, the game speeds up considerably, letting hours go by in a few seconds. As soon as anyone wakes up, time slows down again.

The Sims is a rather unusual game in that it's chiefly about time management. The player is under constant pressure to have his characters accomplish all their chores and get time for sleep, relaxation, and personal development as well. The game runs something like 48 times as fast as real life, so it takes about 20 minutes of real time to play through the 16 hours of game-world daytime. However, the characters don't move 48 times as fast. Their actions look pretty normal, about as they would in real time. As a result, it takes them 15 minutes according to the game's clock just to go out and pick up the newspaper. This contributes to the sense of time pressure. Because the characters do everything slowly (in game terms), they often don't get a chance to water their flowers, which consequently die.

Anomalous Time

In The Settlers: Rise of an Empire, a complex economic simulation, a tree can grow from a sapling to full size in about the same length of time that it takes for an iron foundry to smelt four or five bars of iron. This is a good example of anomalous time: time that seems to move at different speeds in different parts of the game. Blue Byte, the developer of The Settlers, tuned the length of time it takes to do each of the many tasks in the game to make sure that the game as a whole would run smoothly. As a result, The Settlers is very well balanced at some cost to realism. However, it doesn't disrupt the fantasy because The Settlers doesn't actually give the player a clock in the game world. There's no way to compare game time to real time, so in effect, the game world has no obvious time scale (see Figure 4.7).

Another example of anomalous time appears in Age of Empires, in which tasks that should take less than a day in real time (gathering berries from a bush, for example) seem to take years in game time according to the game clock. Age of Empires does have a time scale, visible on the game clock, but not everything in the world makes sense on that time scale. The players simply have to accept these actions as symbolic rather than real. As designers, we have to make them work in the context of the game world without disrupting the fantasy. As long as the symbolic actions (gathering berries or growing trees) don't have to be coordinated with real-time actions (warfare) but remain essentially independent processes, it doesn't matter if they operate on an anomalous time scale.

Letting the Player Adjust Time

In sports games and vehicle simulations, game time usually runs at the same speed as real time. An American football game is, by definition, an hour long, but because the clock stops all the time, the actual elapsed time of a football game is closer to three hours. All serious computerized football games simulate this accurately. Verisimilitude is a key requirement of most sports games; if a game does not accurately simulate the real sport, the league might not approve of it, and its competitors are bound to point out the flaw. However, most such games also allow the players to shorten the game by playing 5- or 10-minute quarters instead of 15-minute quarters because most people don't want to devote a full three hours to playing a simulated football game. This is also a useful feature in testing; it takes far too long to test the product if you have to play a full-length game every time.

Flight simulators also usually run in real time, but there are often long periods of flying straight and level during which nothing of interest is going on; the plane is simply traveling from one place to another. To shorten these periods, many games offer a way to speed up time in the game world by two, four, or eight times—in effect, make everything in the game world go faster than real time. When the plane approaches its destination, the player can return the game to normal speed and play in real time.

The Environmental Dimension

The environmental dimension describes the world's appearance and its atmosphere. You've seen that the physical dimension defines the properties of the game's space; the environmental dimension is about what's in that space. The environmental characteristics of the game world form the basis for creating its art and audio. We'll look at two particular properties: the cultural context of the world and the physical surroundings.

Cultural Context

The cultural context of a game refers to its culture in the anthropological sense: the beliefs, attitudes, and values that the people in the game world hold, as well as their political and religious institutions, social organization, and so on—in short, the way those people live. These characteristics are reflected in the manufactured items that appear in the game: clothing, furniture, architecture, landscaping, and every other man-made object in the world. The culture influences not only what appears and what doesn't appear (a game set in a realistic ancient Egypt obviously shouldn't include firearms), but also how everything looks—including the user interface. Cleopatra: Queen of the Nile is an excellent example of a game's culture harmonizing with its user interface; see Figure 4.8. The way objects appear is affected not only by their function in the world, but also by the aesthetic sensibilities of the people who constructed them; for example, a Maori shield looks entirely different from a medieval European shield.

The cultural context also includes the game's backstory. The backstory of a game is the imaginary history, either large-scale (nations, wars, natural disasters) or small-scale (personal events and interactions), that preceded the time when the game takes place. This prior history helps to establish why the culture is the way it is. A warlike people should have a history of warfare; a mercantile people should have a history of trading. In designing the backstory, don't go into too much depth too early, however. As Chapter 3, "Game Concepts," warned, the story must serve the game, not the other way around.

For most game worlds, it's not necessary to define the culture or cultures in great detail. A game set in your own culture can simply use the things that you see around you. The SimCity series, for example, is clearly set in present-day America (European cities are rarely so rectilinear), and it looks like it. But when your game begins to deviate from your own culture, you need to start thinking about how it deviates and what consequences that deviation has.

Physical Surroundings

The physical surroundings define what the game actually looks like. This is a part of game design in which it's most helpful to be an artist or to work closely with one. In the early stages of design, you don't need to make drawings of every single thing that can appear in the game world (although sooner or later someone will to have to do just that). For the time being, it's important to create concept sketches: pencil or pen-and-ink drawings of key visual elements in the game. Depending on what your game is about, this can include buildings, vehicles, clothing, weaponry, furniture, decorations, works of art, jewelry, religious or magical items, logos or emblems, and on and on. See Grim Fandango (Figure 4.9) for a particularly distinctive example. The game's culture influences constructed artifacts in particular. A powerful and highly religious people are likely to have large symbols of their spirituality: stone temples or cathedrals. A warlike nomadic people have animals or vehicles to carry their gear and weapons they can use on the move. (Note that these might be future nomads, driving robo-camels.)

Nor should you neglect the natural world. Games set in urban or indoor environments consisting entirely of manufactured objects feel sterile. Think about birds and animals, plants and trees, earth, rocks, hills, and even the sky. Consider the climate: Is it hot or cold, wet or dry? Is the land fertile or barren, flat or mountainous? These qualities, all parts of a real place, are opportunities to create a visually rich and distinctive environment.

If your world is chiefly indoors, of course, you don't have to think about nature much unless your character passes a window, but there are many other issues to think about instead. Where does the light come from? What are the walls, floors, and ceilings made of, and how are they decorated? Why is this building here? Do the rooms have a specific purpose, and if so, what? How can you tell the purpose of a room from its contents? Does the building have multiple stories? How does the player get from one floor to another?

Physical surroundings include sounds as well as sights: music; ambient environmental sounds; the particular noises made by people, animals, machinery, and vehicles. If you think about the sounds things make at the same time that you think about how they look, this helps you create a coherent world. Suppose you're inventing a six-legged reptilian saddle animal with clawed feet rather than hooves. How does that creature sound as it moves? Its scales might rattle a bit. Its feet are not going to make the characteristic cilp-clop sound of a shod horse. With six legs, it will probably have some rather odd gaits, and those should be reflected in the sound it makes.

The physical surroundings play a big role in setting the tone and mood of the game as it is played, whether it's the lighthearted cheerfulness of Mario or the gritty realities of the Godfather series (see Figure 4.10). The sound, and especially the music, will contribute greatly to this. Think hard about the kind of music you want, and consider what genres will be appropriate. Stanley Kubrick listened to hundreds of records to select the music for 2001: A Space Odyssey, and he astonished the world with his choice of "The Blue Danube" for the shuttle docking sequence. You have a similar opportunity when you design your game.

Detail

Every designer must decide how much detail the game world needs—that is to say, how richly textured the world will be and how accurately modeled its characteristics will be. To some extent, your answer will be determined by the level of realism that you want, but technical limitations and time constraints will necessarily restrict your ambitions. No football game goes to the extent of modeling each fan in the stadium, and few flight simulators model all the physical characteristics of their aircraft. Detail helps to support the fantasy, but it always costs, in development time and in memory or disk space on the player's machine. In an adventure game, it should, in principle, be possible to pick up everything in the world; in practice, this just isn't practical. As a consequence, the player knows that if he can pick up an object, it must be important for some reason; if he can't pick it up, it isn't important. Similarly, in god games, it's common for all the people to look alike; they're often male adults. Bullfrog Productions once designed a god game with both male and female adults, but there wasn't enough time for the artists to model children as well. People simply were born into the world full grown. Lionhead's Black & White, on the other hand, managed to include men, women, and children.

The camera model you choose, and the way that the player moves through the world, may influence your decisions about the level of detail. For example, in a small stadium such as the Wimbledon tennis courts, the athletes may be conscious of specific people in the crowd, so it makes sense to model them in some detail. In motorsports, however, the spectators will flash past in a blur, and there's no point in putting much effort into their appearance.

Here's a good rule of thumb for determining the level of detail your game will contain: Include as much detail as you can to help the game's immersiveness, up to the point at which it begins to harm the gameplay. If the player must struggle to look after everything you've given him, the game probably has too much detail. (This is one of the reasons war games tend to have hundreds rather than hundreds of thousands of units. The player in a war game can't delegate tasks to intelligent subordinates, so the numbers have to be kept down to a size that he can reasonably manage.) A spectacularly detailed game that's no fun to play doesn't sell many copies.

Defining a Style

In describing how your world is going to look, you are defining a visual style for your game that will influence a great many other things as well: the character design, the user interface, perhaps the manual, and even the design of the box and the advertising. You actually have two tasks to take on here: defining the style of things in your world (that is, its intrinsic style), and also defining the style of the artwork that will depict your world. They aren't the same. For example, you can describe a world whose architectural style is inspired by Buddhist temples but draw it to look like a film noir movie. Or you could have medieval towns with half-timbered houses but depict them in a slightly fuzzy, Impressionistic style. You must choose both your content and the way in which you will present that content.

Both decisions will significantly influence the player's experience of the game, jointly creating a distinct atmosphere. In general, the style of depiction tends to superimpose its mood on the style of the object depicted. For example, a Greek temple might be architecturally elegant, but if its style of drawing suggests a Looney Tunes cartoon, players will expect something wacky and outrageous to take place there. The drawing style imposes its own atmosphere over the temple, no matter how majestic it is. For one example, take a look at Naruto: Ultimate Ninja Storm (see Figure 4.11). All the locations in Naruto are rendered in a flat-shaded style reminiscent of the comic book that inspired the game.

Unless you're the lead artist for your game as well as its designer, you probably shouldn't—or won't be allowed to—define the style by yourself. Your art team will have ideas of its own, and you should listen to those suggestions. The marketing department might insist on having a say as well. It's important, however, that you try to keep the style harmonious and consistent throughout your game. Too many games have been published in which different sections had wildly differing art styles because no one held and enforced a single overall vision.

Overused Settings

All too often, games borrow settings from one another or from common settings found in the movies, books, or television. A huge number of games are set in science fiction and fantasy worlds, especially the quasi-medieval, sword-and-sorcery fantasy inspired by J. R. R. Tolkien and Dungeons & Dragons, popular with the young people who used to be the primary—indeed, almost the only—market for computer games. But a more diverse audience plays games nowadays, and they want new worlds to play in. You should look beyond these hoary old staples of gaming. As Chapter 3 mentioned, Interstate '76 is inspired by 1970s TV shows. It includes cars, clothing, music, and language from that era, all highly distinctive and evocative of a particular culture. Interstate '76 has great gameplay, but what really sets it apart from its competitors is that it looks and sounds like nothing else on the market.

Especially if you are going to do science fiction or fantasy, try to make your game's setting distinctively different. At present, real spacecraft built by the United States or Russia look extremely functional, just as the first cars did in the 1880s, and the spacecraft in computer games tend to look that way also. But as cars became more common, they began exhibiting stylistic variation to appeal to different kinds of people, and now there is a whole school of aesthetics for automotive design. As spacecraft become more common, and especially as we start to see personal spacecraft, we should expect them to exhibit stylistic variation as well. This is an area in which you have tremendous freedom to innovate.

The same goes for fantasy. Forget the same old elves, dwarves, wizards, and dragons (Figure 4.12). Look to other cultures for your heroes and villains. Right now about the only non-Western culture portrayed with any frequency in games is Japanese (feudal, present-day, and future) because the Japanese make a lot of games and their style has found some acceptance in the West as well. But there are many more sources of inspiration around the world, most untapped. Around AD 1200, while the rulers of Europe were still holed up in cramped, drafty castles, Islamic culture reached a pinnacle of grace and elegance, building magnificent palaces filled with the riches of the Orient and majestic mosques of inlaid stone. Yet this proud and beautiful civilization seldom appears in computer games because Western game designers haven't bothered to learn about it or don't even know it existed. Set your fantasy in Valhalla, in Russia under Peter the Great, in the arctic tundra, at Angkor Wat, at Easter Island, or at Machu Picchu.

Sources of Inspiration

Art and architecture, history and anthropology, literature and religion, clothing fashions, and product design are all great sources of cultural material. Artistic and architectural movements, in particular, offer tremendous riches: Art Nouveau, Art Deco, Palladian, Brutalism. If you haven't heard of one of these, go look it up now. Browse the web or the art, architecture, and design sections of the bookstore or the public library for pictures of interesting objects, buildings, and clothing. Carry a digital camera around and take pictures of things that attract your eye, then post the pictures around your workspace to inspire yourself and your coworkers. Collect graphic scrap from anywhere that you find it. Try old copies of National Geographic. Visit museums of art, design, and natural history if you can get to them; one of the greatest resources of all is travel, if you can afford it. A good game designer is always on the lookout for new ideas, even when he's ostensibly on vacation.

It's tempting to borrow from our closest visual neighbor, the movies, because the moviemakers have already done the visual design work for us. Blade Runner introduced the decaying urban future; Alien gave us disgustingly biological aliens rather than little green men. The problem with these looks is that they've already been borrowed many, many times. You can use them as a quick-and-dirty backdrop if you don't want to put much effort into developing your world, and players will instantly recognize the world and know what the game is about. But to stand out from the crowd, consider other genres. Film noir, the Marx Brothers, John Wayne westerns, war movies from the World War II era, costume dramas of all periods—from the silliness of One Million Years B.C. to the Regency elegance of Pride and Prejudice, they're all grist for the mill.

Television goes through its own distinct phases, and because it's even more fashion-driven than the movies, it is ripe for parody. The comedies of the 1950s and 1960s and the nighttime soaps of the 1970s and 1980s all had characteristic looks that seem laughable today but that are immediately familiar to most adult Americans. This is not without risk; if you make explicit references to American popular culture, non-Americans and children might not get the references. If your gameplay is good enough, though, it shouldn't matter.

The Emotional Dimension

The emotional dimension of a game world defines not only the emotions of the people in the world but, more important, the emotions that you, as a designer, hope to arouse in the player. Multiplayer games evoke the widest variety of emotions, because the players are socializing with real people and making friends (and, alas, enemies) as they play. Single-player games have to influence players' emotions with storytelling and gameplay. Action and strategy games are usually limited to a narrow emotional dimension, but other games that rely more heavily on story and characters can offer rich emotional content that deeply affects the player.

The idea of manipulating the player's emotions might seem a little strange. For much of their history, games have been seen only as light entertainment, a means to while away a few hours in a fantasy world. But just because that's all they have been doesn't mean that's all they can be. In terms of the richness of their emotional content, games are now just about where the movies were when they moved from the nickelodeon to the screen. Greater emotional variety enables us to reach new players who value it.

Influencing the Player's Feelings

Games are intrinsically good at evoking feelings related to the player's efforts to achieve something. They can create "the thrill of victory and the agony of defeat," as the old ABC Wide World of Sports introduction used to say. Use the elements of risk and reward—a price for failure and a prize for success—to further heighten these emotions. Games can also produce frustration as a by-product of their challenges, but this isn't a good thing; some players tolerate frustration poorly and stop playing if it gets too high. To reduce frustration, build games with player-settable difficulty levels and make sure the easy level is genuinely easy. Excitement and anticipation, too, play large roles in many games. If you can devise a close contest or a series of stimulating challenges, you will generate these kinds of emotions.

Construction and management simulations, whose challenges are usually financial, arouse the player's feelings of ambition, greed, and desire for power or control. They also offer the emotional rewards of creative play. Give the player a way to amass a fortune, then let her spend it to build things of her own design. The SimCity and various Tycoon games (RollerCoaster Tycoon, Railroad Tycoon, and so on), do this well. Artificial life games and god games such as Spore or The Sims let the player control the lives of autonomous people and creatures for better or worse, satisfying a desire to be omnipotent over a world of beings subject to the player's will. (This may not be a very admirable fantasy, but it's one that a lot of people enjoy having fulfilled.)

To create suspense, surprise, and fear, use the time-honored techniques of horror films: darkness, sudden noises, disgusting imagery, and things that jump out at the player unexpectedly. Don't overdo it, however. A gore-fest becomes tedious after a while, and Alfred Hitchcock demonstrated that the shock is all the greater when it occurs infrequently. For suspense to work well, the player needs to feel vulnerable and unprepared. Don't arm him too heavily; the world's a lot less scary when you're carrying a rocket launcher around. Survival horror is a popular subgenre of action game, as seen in the Silent Hill and Resident Evil series, that uses these approaches.

Another class of emotions is produced by interactions between characters and the player's identification with one of them. Love, grief, shame, jealousy, and outrage are all emotions that can result from such interactions. (See Figure 4.13 for a famous example.) To evoke them, you'll have to use storytelling techniques, creating characters that the player cares about and believes in and credible relationships between them. Once you get the player to identify with someone, threaten that character or place obstacles in his path in a way that holds the player's interest. This is the essence of dramatic tension, whether you're watching Greek tragedy or reading Harry Potter. Something important must be at stake. The problem need not necessarily be physical danger; it can also be a social, emotional, or economic risk. The young women in Jane Austen's novels were not in imminent peril of death or starvation, but it was essential to their family's social standing and financial future for them to make good marriages. The conflict between their personal desires and their family obligations provides the tension in the novels.

A good many games set the danger at hyperbolic levels with extreme claims such as "The fate of the universe rests in your hands!" This kind of hyperbole appeals to young people, who often feel powerless and have fantasies about being powerful. To adults, it just sounds a bit silly. At the end of Casablanca, Rick said, "The problems of three little people don't amount to a hill of beans in this crazy world," but he was wrong. The whole movie, a movie still popular over a half century after its first release, is about the problems of those three little people. For the duration of the film, these problems hold us entranced. It isn't necessary for the fate of the world to be at stake; it is the fates of Rick, Ilsa, and Victor that tug at our hearts.

The Limitations of Fun

• Weaver's Law: The quality of an entertainment is inversely proportional to the awareness of time engaged in it.
• —CHRIS WEAVER, FOUNDER OF BETHESDA SOFTWORKS

Most people think that the purpose of playing games is to have fun, but fun is a rather limiting term. It tends to suggest excitement and pleasure, either a physical pleasure such as riding a roller coaster, a social pleasure such as joking around with friends, or an intellectual pleasure such as playing cards or a board game. The problem with striving for fun is that it tends to limit the emotional range of games. Suspense, excitement, exhilaration, surprise, and various forms of pleasure fall within the definition of fun, but not pity, jealousy, anger, sorrow, guilt, outrage, or despair.

You might think that nobody in their right mind would want to explore these emotions, but other forms of entertainment—books, movies, television—do it all the time. And, in fact, that's the key: Those media don't provide only fun; they provide entertainment. You can entertain people in all sorts of ways. Movies with sad endings aren't fun in the conventional sense, but they're still entertaining. Although we say that we make games, what we in fact make is interactive entertainment. The potential of our medium to explore emotions and the human condition is much greater than the term fun game allows for. A good game is entertainment that involves the player on a number of levels.

All that said, however, bear in mind that most publishers and players want fun. Too many inexperienced designers are actually more interested in showing how clever they are than in making sure the player has a good time; they place their own creative agenda before the player's enjoyment. As a designer, you must master the ability to create fun—light enjoyment—before you move on to more complex emotional issues. Addressing unpleasant or painful emotions successfully is a greater aesthetic challenge and is commercially risky besides.

You Can't Paint Emotion by Numbers

The idea that games should include more emotional content and should inspire more emotions in players has been gaining ground in the game industry for several years. Unfortunately, this has produced a tendency to look for quick and easy ways to do it, mostly by relying on clichés. The young man whose family is killed and who is obsessed by his desire for revenge or the beautiful princess who needs to be rescued both belong more to fairy tales than to modern fiction. That may be all right if your game aspires to nothing more, but it won't do if you're trying to create an experience with any subtlety. Contrast, for example, the simple themes of the early animation films and the more psychologically rich stories in the recent Pixar films.

Beware of books or articles that offer simple formulas for emotional manipulation: "If you want to make the player feel X, just do Y to the protagonist." An imaginative and novel approach to influencing the players' feelings requires the talents of a skilled storyteller. Paint-by-numbers emotional content has all the sensitivity and nuance of paint-by-numbers art.

The Ethical Dimension

The ethical dimension of a game world defines what right and wrong mean within the context of that world. At first glance, this might seem kind of silly—it's only a game, so there's no need to talk about ethics. But most games that have a setting, a fantasy component, also have an ethical system that defines how the player is supposed to behave. As a designer, you are the god of the game's world, and you establish its morality. When you tell a player that he must perform certain actions to win the game, you are defining those actions as good or desirable. Likewise, when you say that the player must avoid certain actions, you are defining them as bad or undesirable. The players who come into the world must adopt your standards or they will lose the game.

In some respects, the morality of a game world is part of its culture and history, which are part of the environmental dimension, but because the ethical dimension poses special design problems, it needs a separate discussion. The ethics of most game worlds deviate somewhat from those of the real world—sometimes they're entirely reversed. Games allow, even require, you to do things that you can't do in the real world. The range of actions that the game world permits is typically narrower than in the real world (you can fly your F-15 fighter jet all you want, but you can't get out of the plane), but often the permitted actions are quite extreme: killing people, stealing things, and so on.

Moral Decision-Making

On the whole, most games have simple ethics: clobber the bad guys, protect the good guys. It's not subtle but it's perfectly functional; that's how you play checkers. Not many games explore the ethical dimension in any depth. A few include explicit moral choices, but unfortunately, these tend to be namby-pamby, consistently rewarding good behavior and punishing bad behavior. Such preachy material turns off even children, not to mention adults. But you can build a richer, more involving game by giving the player tough moral choices to make. Ethical ambiguity and difficult decisions are at the heart of many great stories and, indeed, much of life. Should you send a platoon of soldiers to certain death to save a battalion of others? How would you feel if you were in the platoon?

In many role-playing games, you can choose to play as an evil character who steals and kills indiscriminately, but other characters will refuse to cooperate with you and might even attack you on sight. It's easier to get money by robbing others than by working for it, but you may pay a price for that behavior in other ways. Rather than impose a rule that says, "Immoral behavior is forbidden," the game implements a rule that says, "You are free to make your own moral choices, but be prepared to live with the consequences." This is a more adult approach to the issue than simply punishing bad behavior.

You must be sure to explain the ethical dimension of your game clearly in the manual, in introductory material, or in mission briefings. For example, some games that have hostage-rescue scenarios make the death of a hostage a loss condition: If a hostage dies, the player loses. This means that the player has to be extra careful not to kill any hostages, even at the risk of his own avatar's life. In other games, the only loss condition is the avatar's death. In this case, many players shoot with complete abandon, killing hostages and their captors indiscriminately. In real life, of course, the truth is somewhere in between. Police officers who accidentally shoot a hostage are seldom prosecuted unless they've been grossly negligent, but it doesn't do their careers any good. You can emulate this by penalizing the player somehow. To be fair to the player, however, you need to make this clear at the outset.

The ethical dimension of multiplayer games, whether online or local, is an enormous and separate problem. Chapter 21, "Online Games," discusses this issue at length.

A Word about Game Violence

It's not part of this book's mission to debate, much less offer an answer for, the problem of whether violent video games cause violent behavior in children or adults. This is a psychological question that only prolonged and careful study can resolve. Unfortunately, a good many people on both sides of the issue seem to have made up their minds already, and arguments continue to rage in government and the media, supported for the most part by very few facts.

For you, as a designer, however, consider these suggestions. The essence of many games is conflict, and conflict is often represented as violence in varying degrees of realism. Chess is a war game in which pieces are killed—removed from the board—but nobody objects to the violence of chess; it's entirely abstract. American football is a violent contact sport in which real people get injured all the time, but there are no serious efforts to ban football, either. The only way to remove violence from gameplay is to prohibit most of the games in the world because most contain violence in some more-or-less abstract form. The issue is not violence, per se, but how violence is portrayed and the circumstances under which violence is acceptable.

Games get into political trouble when they have a close visual similarity to the real world but an ethical dimension that is strongly divergent from the real world. The game Kingpin encourages the player to beat prostitutes to death with a crowbar, with bloodily realistic graphics. Not surprisingly, it has earned a lot of criticism. On the other hand, Space Invaders involves shooting hundreds of aliens, but it is so visually abstract that nobody minds. In other words, the more a game resembles reality visually, the more its ethical dimension should resemble reality as well, or it's likely to make people upset. If you want to make a game in which you encourage the player to shoot anything that moves, you're most likely to stay out of trouble if those targets are nonhuman and just quietly disappear rather than break apart into bloody chunks. Tie your ethical realism to your visual realism.

Computer games are about bringing fantasies to life, enabling people to do things in make-believe that they couldn't possibly do in the real world. But make-believe is a dangerous game when it's played by people for whom the line between fantasy and reality is not clear. Young children (those under about age eight) don't know much about the real world; they don't know what is possible and what isn't, what is fantasy and what is reality. An important part of raising children is teaching them this difference. But until they've learned it, it's best to make sure that any violence in young children's games is suitably proportionate to their age. The problem with showing violence to children is not the violence, per se, but the notion that there's no price to pay for it. For a detailed and insightful discussion of how children come to terms with violence, read Killing Monsters: Why Children Need Fantasy, Super Heroes, and Make-Believe Violence by Gerard Jones (Jones, 2002). Ultimately, the violence in a game should serve the gameplay. If it doesn't, then it's gratuitous and you should consider doing without it.