This article has since been revised and updated

[size="5"]Introduction The design of videogames has never been a subject of formal study. All too often today games are designed simply by looking at whatever is popular at the moment and copying it. The recent interest in classic videogames has a basis in something more than mere nostalgia. There is a feeling that something has been lost in game design, even while graphics and sound are reaching new heights. This document is an attempt to explore certain concepts of videogame design. It is not a formula for designing hit videogames. I firmly believe designing games by formula leads to stagnation. But I just as firmly believe that a good videogame must be built on a solid foundation. So, I have divided the subject into four elements and written an essay on each one. I've gathered them together here for the benefit of anyone interested in game design. [size="5"]The Element of Presence One of the magical aspects of videogames is the ability to simulate worlds in the machine. These worlds may be tiny or huge, simple or complex. This is an element I call "presence" which is almost like virtual reality. A good videogame creates a feeling of being there, not just watching a video or a slideshow. Now I will list the components that go into making a successful simulation, beginning with the most basic and important, then working up toward more sophisticated frills that can enhance the experience. Take this as a recipe for creating that element of presence in the game: [size="3"][bquote]1. Movement and Collision [bquote][size="2"]Among the first things infants learn are simple rules of movement and collision. Things bump things. These rules are closely tied to our sense of reality. It's why smooth motion and accurate collision detection are so important. The earliest videogames, such as Space War and Pong, were built around movement and collision. Too many adventure and "interactive multi-media" games still are not. This is one of the biggest failings of full motion video (FMV) games.[/bquote][size="3"]2. Internally Consistent Rules [bquote][size="2"]Objects in the game should interact with each other according to certain rules. These rules should be simple and consistent enough for the player to figure them out and use them. He should be able to say, "If object A does this to object B, then it should do that to object C." These rules don't necessarily have to be the same as rules in the real world. The rules may be fantastic or abstract, but they must make sense within the context of the game. These are like the laws of physics in your game world.[/bquote][size="3"]3. Freedom of Movement [bquote][size="2"]The player should generally be able to move where he wishes, when he wishes, for his own purposes. Games that force the player along a fixed path at a predetermined speed lose some sense of reality. FMV games or others that force the player to follow pre-written scripts also come up short. Instead, provide the player an environment and allow him to explore it. It's okay for the environment to be small, or for there to be obstacles for the player to overcome before reaching some parts of it. However, free movement within the bounds and rules of the environment is critical. Asteroids is a good early example: a game which allowed the little spaceship to fly all over the screen, as opposed to games like Space Invaders with mere side-to-side movement.[/bquote][size="3"]4. Realtime Action and Background Activity [bquote][size="2"]The real world does not stop and wait for us to make decisions. Some games, especially adventure and strategy games, freeze in their tracks much like a cheap choose-your-own-adventure book. This is considered harmful. Realtime action does not necessarily mean you keep your player racing against the clock (unless it's that kind of game), but the game should not freeze while he decides his move. Things should constantly be happening in the background, regardless of what the player does. This background activity, even if only partly seen by the player, adds a tremendous sense of vitality and presence to the game. VCS Adventure stands out as an early example of this technique. Regardless of the player's actions, the dragons and the black bat continued roaming about and doing their own things.[/bquote][size="3"]5. Three Dimensional Space [bquote][size="2"]Every game example I've listed so far has been two dimensional in nature. However, people have a natural ability to think in three dimensions. Adding three dimensional space to a game makes it more stimulating and can enhance the feeling of simulating a world in the machine. Games that build realtime 3D polygon environments are perfect for creating this feeling of really being there. Examples would include Frontier, Virtua Racing, Cybermorph, Iron Soldier. Note that these are all fairly new games. Advancing technology is only beginning to unleash the potential of 3D environment gaming. It is probably not a good idea to force a 3D environment onto inadequate hardware, as we saw with some versions of Frontier. Note well also that FMV or pre-rendered scenes do not offer a true 3D environment. Instead they present only a thin illusion of such, and may not hold a player's attention for long.[/bquote][size="3"]6. Virtual Reality [bquote][size="2"]True virtual reality starts by incorporating all five of the previous elements. It builds on that by adding immersiveness: using tracking helmets, sensor gloves and other gadgets to create the illusion that the player is actually inside the electronic environment. Current VR technology is primitive. In a way, it reminds me of Pong and Spacewar, because we see only the first hint of what it can become in due time. To me that's exciting. However, it could be that too much emphasis is being placed on a concept that isn't entirely here yet, either technically or economically. Let's master the five basics before we seriously move on to level 6, okay? [/bquote][/bquote][size="2"]There it is, a simple foundation for creating presense in your games. However, I'm sure some developers will dismiss my advice. Creating worlds in the computer is not easy. It's much easier to throw together some fancy digitized graphics, digitized sound and speech, CD music, and maybe some FMV for a "multi-media interactive experience". The result is a disjointed hodgepodge with large amounts of glitz and small amounts of actual gameplay. It doesn't take consumers long to catch onto the distinction. Virtual reality and multi-media present two very different paradigms, very different views of where information technology should be going. Multi-media, as the name implies, is simply the cobbling together of various pre-existing media: text, pictures, sounds, music, film. It can be useful indeed for some purposes, such as presenting reference material. However, multi-media is itself only another communications medium, one more way of repackaging information. Virtual reality, however, is the paradigm we need for games. It's more than just a way of packaging information. Rather, VR is a tool for people to generate new information. Instead of digging through scraps of words, music and images that someone else created, VR style simulations allow people to interact with the computer to produce new images and ideas that nobody has seen before. VR allows people to play! Play is a natural method of exploration and experimentation, the way that human beings instinctively learn. That is why VR is a superior metaphor for videogames, however useful multi-media might be for other things. Don't fall into the trap of thinking that killer, state-of-the-art hardware is required to carry out the techniques of presence. The first four techniques I listed were all used to great effect in VCS Adventure, a game that ran from 4K ROM on a console with 128 bytes of RAM and minimal graphics and processing power. VCS Adventure became one of the most popular games ever done on the Atari VCS. Needless to say, these same ideas can used far more effectively with even the lower end of today's hardware. [size="5"]The Element of Chaos [size="2"]A disturbing trend in recent years has been toward games where everything is pre-scripted. Each level is the same every time: same powerups in the same places, enemies in the same places following the same patterns. The game becomes a contest of pure memorization. Once all the secrets are found and the levels memorized, it's time to buy another game. That's not to say there's no place for these games, or that it's not fun to explore and learn their secrets. That learning and exploration is the hub of their appeal. But, we are missing an opportunity for something even better. The missing element is chaos. Chaos means that surprising things can happen at almost any time in the game. It means that simple objects can interact in unexpectedly complex ways. The player is absorbed not in learning preset patterns, but in learning how the objects interact. This is not only great fun for the player, but also a more efficient way to program the game. Instead of working to create level after level of pre-scripted secrets, the developer can just create interesting objects and turn them loose in the game environment. He doesn't have to create all the many (possibly hundreds or thousands of) interactions that can occur in the game. These interactions arise naturally from the objects themselves, including interactions the game designers may not have thought about. At this point a couple of examples are in order. First, let's consider Cosmic Avenger. This was a coin-op game of the early 1980s, which was also ported to the ColecoVision. At first glance it didn't seem too different from the other pre-scripted, side-scrolling shooters of the day, such as Scramble and Super Cobra. But, Cosmic Avenger had a small chaos element those games lacked. The enemy forces could accidentally shoot each other. There were also enemy missiles that would home toward the player's ship, and could follow it for quite some distance. This led to great possibilities. The player could lead homing missiles into hitting enemy ships or installations. Enemy objects could collide or shoot each other at unexpected moments. These reactions made the game richer. Now let's also consider VCS Adventure once again. It appears superficially similar to lots of graphic adventures that followed, such as the Zelda series. But, there is a big difference. Games like Zelda were entirely pre-scripted. They presented a series of obstacles which were always the same, located in the same places, sitting dormant until the player arrived to trigger them. Typically a certain screen in the Zelda adventures would have X number of type Y monsters which would appear and attack when the player character entered that screen. By contrast, VCS Adventure had four creatures in the whole game: three dragons and a black bat. These creatures got a lot of mileage because they constantly wandered about the game world doing their thing no matter where the player was. There were some limits to their movement, such as entering or leaving castles. But, there was no way to be sure when a dragon or the bat might pounce on you. They could attack when you were unarmed, and they could attack singly, or they could gang up on you. There was no way to know. Although the dragons provided the combat aspect of the game, the black bat was the really important chaos element. It followed a simple rule. It picked things up and carried them around. After a few minutes it would get tired of whatever it was carrying; then it would look for something new to pick up, and leave the old object. For example, I might be hunting down the dragon with sword in hand. Finally I find the beast, but about that time the bat suddenly swoops in and takes my sword and leaves me holding something useless like a key. It's time to run for my life! Or, for another example, I might be returning to the gold castle with the chalice, which is the final goal of the game. Victory is almost in sight, but suddenly the bat swoops in and grabs the chalice away from me. The chase is on, and it might be another half hour of one thing after another before I get the chalice back where I wanted it. And, perhaps the ultimate example of bat oddities was the Grand Tour. This happened when the player simultaneously killed a dragon and was eaten by it. The player was trapped in the dragon's belly, but sometimes the bat would come and fly off with the dead dragon, thus carrying the player along for a fast tour of the game playing area. In addition to all the above, there was an extra playing mode in Adventure which allowed the game to start with all objects in random positions. Under these conditions there was no way to know if the game could even be won. But, it added a tremendous amount of extra play value after the standard scenario had been mastered. So, this is what I would like to see in more games: more randomness and more simple objects that can interact in complex ways. These chaos ideas can apply to a wide range of games. [size="5"]The Element of Challenge [size="2"]A videogame, like almost any game, is a contest. The player tests his abilities against the machine or against a friend or against both at once. So, it is critically important to create the right level of challenge and difficulty. A game should not be too easy, lest it become boring. Nor should it be too difficult, lest it become frustrating. And, it's also helpful if the game is accessible. When Nolan Bushnell created his first coin-operated videogame, Computer Space, it did very poorly. People didn't understand how to play. His next attempt, Pong, was a giant success because it was more accessible. Anyone could look at the game and immediately see how it was played. Many of the great hit videogames have had this trait of accessibility, such as Space Invaders and Pac Man. More modern examples would include basic scrolling shooters like Raiden, most racing games, and any pinball machine. That is one reason pinball remains popular. Everyone knows how to play. That's not to say complicated games can't be successful. There is a feeling of accomplishment that comes from learning a complex game. This is found among the players of fighting games with their rather abstruse system of control and "combos", and among players of computer role-playing games, and flight simulators. Even so, such games are naturally limited in the width of their appeal. These kinds of games can become very popular within certain circles, but they rarely reach widespread acceptance with the general public. This is the difference between hard-core game players and the average guy. And, it's a big part of why the rather simple game of Tetris was such a huge hit. It was too small and simple to intimidate people who are not really into videogames. Making a game simple to learn doesn't mean making it simple to master. I note again that complex interactions in the game are desirable: they give a game depth and staying power. But, these complex interactions should arise from simple objects and rules in the game. VCS Adventure, which I have already described as an example of chaos, also had very simple controls and was as easy to play as anything. The joystick moved your little hero, and the action button picked up or dropped objects. There were no statistics to track or inventories to manage, unlike more modern adventures. Anyone could pick up the stick and start playing. As for the more general issue of challenge and difficulty, the easiest way to address it is by adding difficulty settings or some form of adjustment the player can make to suit himself. This is especially important when you look at the wide range of ages the videogame market caters to these days. There is no reason to artificially limit your game to a narrow age group when you can widen it's appeal by providing the difficulty adjustments. It is nice if you can provide a little something extra in the game as a bonus for those who play on the harder difficulty levels. By luring players onto the harder difficulty as soon as they are ready for it, you can extend the life of the game. Some companies don't want to extend the life of the game. They feel they can make games to be solved quickly and cast aside, so the consumer can buy another. This policy is shortsighted and ultimately counterproductive. Game players eventually realize they are being ripped off. They become bored with easy, quickly solved games and migrate away from videogames toward other forms of entertainment. This effect may partly underlie the current slump in the industry. Multi-player options are badly overlooked. The thrill that comes from playing against a human opponent is wonderful. Even games that were otherwise rather bland somehow gain new life. I am a big advocate of game machines with built-in support for more than two controllers, and I think the new horizons of modems and networking need to be explored aggressively, especially the voice-modem approach letting the players talk (or yell) to each other during play. Why is multi-player so great? One reason is that it brings a large chaos element into the game. I've already written about the benefits of unpredictability, and there's nothing much more unpredictable than a human being. The other player doesn't follow any preset pattern. The interactions between multiple players and other game objects can be far more complex and exciting than you would normally see in a one-player game. Also, there's a greater sense of competition against a human rather than a machine. But, it's good to provide a handicap option in the multi-player games where the players are competing directly against each other (as opposed to the cooperative type), so that players of different skill can still enjoy themselves. [size="5"]The Element of Atmosphere [size="2"]So far I have described only game mechanics. Some games have succeeded on the strength of game mechanics alone, such as Qix or Tetris. However, it's much easier to engage the player's interest and enrich his experience by creating an atmosphere. The first rule of atmosphere is that it must not sabotage the underlying gameplay in any way. That would be self-defeating. Instead, take the game and build a mood around it. There are two general ways to create atmosphere. It can be done in the game itself, or it can be done with "incidentals" in the title and intermission screens. The advantage of moving some atmosphere elements (especially the more intrusive ones like music and video clips) out into the intermission is not to distract from the focus of the game itself. In-game graphics are one of the great things that create atmosphere. This component is so obviously important that I won't even write about it much. Just look around at games with great graphics and see for yourself! Examine games like Tempest 2000 with its wonderful pixelshatter explosions and melt-o-vision effects! Look at wonderfully texturemapped games like Ridge Racer or Total Eclipse! Check out the intricate animations in Rayman! Look at the computer rendered graphics in Donkey Kong Country! It doesn't matter what techniques you use, if you can get an effect that's impressive and appropriate to the game -- and one that enhances gameplay rather than sabotaging it. Graphics can't substitute for gameplay. When putting graphics into a game, there's a natural conflict evolving between artists and programmers. In the old days they were one and the same. Now it has become possible for design people to turn out true color pictures, digitized graphics, video clips, music and so forth, and then turn them over to the programmer to be pasted into the game. That can be good. However, I feel the greatest "WOW" factor still comes from a programmer who knows how to do video tricks right in the hardware, such as Jeff Minter has done with Tempest 2000 for example. In some ways it depends on the kind of game. For some kinds of games, which are more reality-based, the use of artist-created graphics (whether computer rendered, painted, digitized or whatever) is appropriate. For some other kinds of games, such as more abstract sorts of games, the programmer-created video effects are still unbeatable. However, I do want to talk more about "incidental" graphics, music and other components. These would be title screens, between-level menus, FMV introductions and intermissions, etc. These must be made so the player can get past them quickly and get into the game without delay if he wants to. Nonetheless, these components can be used to tremendous effect for creating atmosphere. The best example of this is Road Rash 3DO. Still pictures are used in the menus and intermissions, and FMV clips with fine music for the title screens, start and end of races, and so forth. All these components combine to add tremendously to the excitement surrounding the game, but they don't interfere with the actual gameplay. It's a great example of how to take a game, which in this case already had solid gameplay and excellent in-game graphics, and then surround it with incidentals that raise it over the top and make it an extra special experience. This was was tough prior to the advent of CD-ROM storage technology, because such incidentals would have squeezed out actual gameplay elements. Now that we have CD-ROM, storage space should no longer be much of an obstacle. Earlier in this document I railed against the use of full motion video. I think FMV has a place. The biggest mistake is trying to build a game around FMV, but building FMV around a game can work quite well. Road Rash 3DO demonstrates the great use of FMV outside the actual gameplay. Another good example would be FMV clips embedded in the game at certain appropriate points, such as when the player attains a goal in an adventure game. Again, it's important to let the player quickly abort out of the video clip if he's tired of seeing it. A good example of how FMV could be useful is Alien Vs Predator. In the course of the game, the player will sometimes find computer terminals in which crewmen have recorded log entries. These are plain text. FMV clips could have been far more effective in creating an impression of the desperate last days on board the space station. It was impractical in AvP because of the limitation of the Jaguar ROM cartridge, but the advent of CD-ROM gaming allows these kinds of uses. Some games of the pre-CD-ROM era have used various kinds of animation to create these embedded sequences. CD-ROM and FMV allow us to do it better. Sound effects are an important atmosphere component, just like graphics. All too often overlooked, great sound effects can sometimes contribute more to the mood of the game than graphics do. For example, Alien Vs Predator on the Jaguar has fine graphics, but the excellent digitized sound effects go perhaps even farther toward setting the mood of the game. In many cases, background sound effects are superior to music for setting a mood in the game. For that matter, even eerie silence can be useful sometimes. Music can be used to set mood, but only with some caution. Music is a powerful, and in some ways heavy-handed, tool. It is an arrogant way to set a mood. Rather than create an environment with graphics and sound effects and let the player react to it as he wishes, music grabs the player by the nose and leads him around, saying "THIS is what you should feel!" That's fine for introductions and intermissions. It's not so bad either for shooters or other heavy action games because there's really little doubt about how the player should feel. He should feel excited, and you can use the music soundtrack to pump him up. However, for other kinds of games, such as adventure games, a constant background track during the gameplay is probably bad. It is better to let the player explore the game in his own way, and save the music for brief moments when something special happens in the game. Myst used music very effectively in this way. Most of the time it used sound effects to set the mood (and rather effectively, I should note) but used music when the player found some very special places in the game. The element of atmosphere has always existed in videogames. It's possible to pick out fine examples from the early 1980s, such as the beautiful opening screen of Phoenix or the intimidating aura of Defender. Artistic sensibility and attention to detail are more important than technology. But, with the advent of newer and more powerful graphics and sound hardware and CD-ROM storage, it has become easier than ever to created atmosphere in a game. So, I think it's appropriate to end this section with a warning. It becomes very tempting to pour all your efforts into the glitz and glamour of FMV introductions, intermissions, rendered graphics and the like while overlooking gameplay. This is what many game players fear from the new consoles. The elements of presense, chaos and challenge should come first. When they are in place and working well, then it's fine to go back and tart up the game with atmosphere components. [size="5"]In Closing [size="2"]Since I began writing this document, my thinking has evolved greatly. What began as a rather narrow-minded tirade against the evils of interactive movies and multimedia games has turned into a guide that, I hope, may be useful or at least thought-provoking to anyone involved in videogame design. I still have to get in a parting shot, though. I've noticed the Hollywood Effect intruding onto the world of videogames. Games have become far more expensive to create. To make something commercially competitive on CD-ROM requires a lot more resources than the old days of 8K ROM carts on the VCS. Because of the great expense, game companies become less and less willing to take risks. Increasingly they focus on well worn genres and near copies of already successful games. I wish I knew the cure to get more innovation, which I feel is needed to shake off the current malaise in the industry. But, the economics are against my view. I hope there will always be hackers making games on a shoestring in their basements, garages, and attics. Their product may not be as slick and glamorous, but we can count on them to turn out something different from time to time. This document is dedicated to them.