Home Consulting Games Writing Personal Links

Learning From Fiction

This article was published in Journal of Computer Game Design, Chris Crawford's now-defunct journal.

In "Computer Game Design -- The Write Stuff" (JCGD, Vol. 2 #1, p. 4-7), David Mullich commits one of the oldest fallacies in the book: he takes his analogy too far.

The simple fact is that games are not stories, any more than they are simulations, movies, or video entertainment. They partake of aspects of all of these things -- but are identical to none.

The differences between fiction and gaming are many; to ignore them can lead to poorly-designed games. Equally, the similarities between the two suggest some techniques which game designers may find helpful.


The first and most important difference between story and game is that stories are linear. That is, they have a beginning, a middle, and an end. A story happens the same way each time you read it.

A game is best described as a triangle. Every time you boot a game up, you start at the some location, the "beginning" apex of the triangle; as the game develops, you move across the triangle, and end up somewhere along the base. The specific path you take can vary widely from one playing to the next: you can move along either leg of the triangle, or anything in between.

A game sets the parameters for each player's experience -- that is, it defines the triangle; but the player's decisions determine the specific outcome in any one session of play.

One of the great virtues of a game is that it permits players to explore a range of possibilities through repeated play. A book on Waterloo can only tell you what happened; a game based on the battle can show you not only what happened, but how it might have happened differently, and thereby give you a better gut feel for the nature of war during the Napoleonic era.

The simple fact is that, by trying to make a game more like a story, at least in the sense of providing an explicit narrative, you make it less like a game. Consider the text or graphic adventure; it overlays onto our "game-triangle" a tree network of connected decision points. Whenever you play an adventure, you follow a linear path detemined by your decisions, and so you experience a linear "story," much like a written fiction. However, this is achieved at a cost. Most games permit a smoothly-varying set of experiences (e.g., you can increase the supplies to your right flank by one, or two, or a thousand). An adventure restricts your choices much more tightly; you may only choose among several discreet options at each decision point. Thus, the triangle defined by most games collapses to a mere network of lines.

My belief is that the text and graphic adventure, and its associated forms (laser-disk videogame, VCR game, choose-your-own-ending paperback, solitaire roleplaying adventure, paragraph-system boardgame, and the play version of "The Mystery of Edwin Drood") are basically not a lot of fun, precisely because they try to reconcile the contradictory requirements of game and story. Their repeat playability, one of the most important aspects of any game, is close to nil. Conversely, the emotional impact of the narrative, one of the most important aspects of any story, is also close to nil. This is because many "paths" of an adventure just don't make for good stories ("oops, you die"). But such paths must exist, because an adventure has to permit players to make logical decisions, even when those decisions make for bad stories.

In this context, it is clear that a slavish adherence to the idea that stories and games are kin can lead to bad games.


However, both games and stories develop. That is, they are not static, like a painting or a puzzle; they both present the audience with a series of events and an eventual resolution.

In this sense, both games and stories have plots. I use the word with some trepidation, because I wish to use it in its specific, writerly sense, rather than in the commonly-understood way.

Pat Murphy (author of the Nebula-award winning novel, The Falling Woman, Tor Books, $3.95 -- highly recommended) says: "I think plot is one of the most difficult concepts to get across. The important aspect of a plot is rising tension." (The emphasis is mine.)

If you remember your high school English classes, you probably remember that your teacher spent a lot of time drawing curves with positive slopes on the board, showing you a climax, or a series of short climaxes leading to a larger one, or something of the kind -- graphing a story's development.

Like stories, games develop; and in good games, as in good stories, tension rises as time goes on.

In fact, one of the most common game failures is anticlimax. The period of maximum tension is not the resolution, but somewhere mid-way through the game. After a while, the computer opponents are on the run, or the player-character is powerful enough to defeat any opposition. In these cases, the game is subsequently rather dull, although I suspect that most players will finish, if only because they've invested a lot of energy getting to this point and want to see the screen which tells them they've won.

How can a game support rising tension? Here are some ideas:

  1. One of the most common methods is simply to increase the power and number of computer opponents as the game goes on. To my mind, this is not particularly elegant, but it works.
  2. Another way is to make sure that the game is not zero-sum, that is, that the player's gain is not necessarily a computer-opponent's loss. This way, both the player and his opponents may gain increasing resources throughout the game, so that the player always feels that he is progressing, but never feels that his opposition is effectively out of the game.
  3. A third way is to face the player with the problem of depleting resources. The simplest and least elegant way to do this is to make the game a race against the clock: the player has a limited time to achieve his objective. This increases tension as the clock ticks on. A variant is to tell the player that he must achieve his objective before an opponent achieves a different objective; this way, the player can balance the utility of thwarting his opposition against striving for his own goals. But the depleting resource need not be time. It can be any game variable -- oil, money, gunpowder, magical energy, hit points. By forcing a player to husband his resources, you force him to make difficult choices -- choices of increasing difficulty (and therefore increasing tension) as the game goes on.
  4. Another way is simply to increase the scope and complexity of the game as it progresses. New technologies become available, or new rules come into play, or new magic items can be obtained. Each new factor that a player must take into account when deciding what to do increases the difficulty of choosing optimum solutions -- and thereby increases the player's level of tension.

Narrative Color

Poul Anderson says that he consciously tries to include some reference to sensory perception in every paragraph he writes -- describing not only what a character sees, but what he hears, smells, tastes; the wind on his skin, the drone of insects, whatever.

As a consequence, Anderson's fiction has a sense of place and a feeling of sensory depth lacking in many other authors' work. The need to establish and sustain narrative color -- not only in terms of perception, but also in terms of the emotional mood and the nature of the objects with which characters surround themselves -- is highly important in fiction. It is especially important in the genres of science fiction and fantasy, because stories of both types take place in imaginary worlds which are often quite different from our own. The author of science fiction or fantasy cannot rely on the readers' shared experience of the world which allows fiction set in the present day to set a scene with a considerable economy of detail.

A lot of science fiction and fantasy is, as a result, rather colorless, despite the exotic nature of the places it depicts. Plot and character are the primary requirements of fiction, and many authors concentrate on them to the virtual exclusion of narrative color.

So it is with games. Despite spectacular graphics and sound, many games don't look or feel a bit like their putative subjects. I think in particular of many SSI computer wargames, which do not make the slightest attempt to create an atmosphere appropriate to their subject. The Eon group's Lords of Conquest, an otherwise fine game, is also a good example of colorlessness: the warring nations it depicts have no personality, and it isn't even clear whether the technology they're using belongs in the sixteenth century or the twentieth. Lords of Conquest can be a lot of fun to play, but its lack of color robs it of emotional power.

By comparison, Ancient Art of War is practically dripping in color, from the Ming-dynasty box design to the elegant unit symbology. It is also, in my estimation, a lot less interesting to play than Lords of Conquest -- but boy, is it colorful.

And frankly, color makes up for a lot. An illustration may be helpful. In 1981, I bought a boardgame entitled Axis & Allies from its publisher, Nova Games. It had the paper map and cardboard counters typical of hobby games. Its map was crudely-drawn and had one of the most garish color schemes I'd ever seen; the counters were equally hideous. I looked at it, decided it was garbage, and put it on the shelf, where it has remained ever since.

In 1986, I bought Axis & Allies again, from its new publisher, Milton-Bradley; the game was essentially identical, except that the mounted map's graphics were gorgeous, the cover art attractive, and it came with a million little plastic tanks and airplanes. I looked it over, thought "wow!", and have played it at least a dozen times since. Frankly, as a game qua game, it has severe flaws; but the sheer tactile joy of pushing around all the little plastic aircraft carriers and soldiers makes up for a great deal.

However, color must be appropriate to its subject. I am minded of Crawford's Trust & Betrayal, a highly innovative study in interpersonal relations packaged as a kind of sci-fi duel between aliens. Not only does the fantasy overlay have nothing to do with the putative subject, it in fact sabotages what the game is trying to do. In the game, the player must negotiate with a series of computer characters using an iconographic language. But the characters are depicted as alien creatures -- leading this player, at least, to try to deal with them on non-human terms, which is not the point of the game at all. We have here color of a sort, but color which obscures rather then reinforces.

Narrative color is not a particularly difficult thing to incorporate in a game. All you need to do is to choose graphics, sound, text, and packaging which are appropriate to and help sustain the fantasy central to the game. However, it is also precisely the kind of extra-effort polishing which deadline-pressed designers are likely to forgo.

Player Characters

Many games define a player's game-token as a "character." Many, of course, do not; they call it an empire, or a business, or an army. In all cases, however, there is a set of game variables that a player can manipulate. A game gains power when a player identifies with his game-token, and therefore, in this sense, "character-identification" improves a game.

However, the computer game notion of the player-character is in almost direct contradition with the fictional idea of character. In fiction, a character is more than a name and a set of abilities; if that's all he or she is, the character is termed "cardboard," meaning cliche, and the writer is treated with contempt. To be "well-rounded," a character must have a personality -- a set of attitudes and beliefs -- and he must act consistently with them. The creation of well-rounded characters is a central preoccupation of all fiction, and writers struggle constantly to describe and define unusual personalities.

Ascribing a personality to a player-character is, however, a bad mistake for a game to make. To the degree that a game predefines a player-character's personality, it robs a player of free will: if he wants to kill the elf-maiden, and you have defined him as a chivalrous knight, you must stop him from doing so. Arbitrarily frustrating a player's desires is not a recipe for a fun game.

Non-Player Characters

The idea of personality -- character -- is, however, a useful one to apply to non-player characters. By "non-player character," I mean virtually any game-token not directly controlled by the player -- computer opponents, information sources, subordinates, etc.

The simple fact is that humans are social animals. We enjoy interacting with humans a lot more than we enjoy operating machines. Indeed, to the degree that computer games are successful, it is because they provide the kind of interaction that, say, operating my mimeograph doesn't.

By describing game-tokens in character terms, you give a player an immediate handle on how the token is likely to act in a game context. Ancient Art of War is again a good example; it allows players to choose among several computer opponents, giving each a name and ascribing a personality to each. This is a lot more immediate, and a lot easier, than describing how each software routine works in detail.

In general, the use of personality-hooks -- names, graphics, depictions of individuals, dialog -- can greatly add to a game's emotional appeal and narrative color.


The basis of all gaming is decision-making and the exploration of outcomes. The basis of all fiction is plot and character. To equate the two is to make a bad mistake.

However, both forms share certain requirements. Among these are the need for development over time, for color and context, and for personality. Game designers can learn a great deal by considering the methods authors use to address these problems.

Return to Articles page.

Home Consulting Games Writing Personal Links

© Copyright 1988 by Greg Costikyan .