Obsession with Cosmetics

This is by far the most common beginner's mistake, and even experienced designers often stumble over it. I use the term cosmetics to describe those elements of the game that are meant primarily to look or sound good rather than to further the gameplay itself. Another term for the same concept is eye candy.

There are five common motivations to equip a game with good graphics and sound:

1. To further the gameplay

2. To permit the player to show off the superior cosmetic capabilities of his new computer

3. To show off the superior technical prowess of the programmer

4. To keep up with the competition

5. To provide the player with images and sounds that are intrinsically pleasing

Reason #1: To further the gameplay by making the player's situation and options as clear as possible. This is the only good reason for pursuing cosmetics.

Remember, this is a game, not a movie, not a musical, not a book. Everything that you put into the game should support the gameplay. If it doesn't support the gameplay, it doesn't belong in the game.

Many game designers have difficulty understanding what is meant by "supporting the gameplay." They confuse relevance with support, and so figure that any image, animation, or sound that is relevant to the topic of the game belongs. If the game is a dungeon crawl, they figure a picture of a castle or a dragon just naturally belongs in there somewhere. This is fuzzy, sloppy thinking. In order to support the gameplay, the cosmetic feature must in some way provide the player with information that is relevant to the choices that the player must make.

An example of a useful graphic is the addition of shadows to a 3D combat game. If the player notices the shadow of an enemy just around the corner, that information is useful and will affect the player's decisions. Hence, the addition of shadows to a 3D engine is beneficial to the game design.

Many game designers erroneously believe that graphic realism necessarily enhances the entertainment experience. Figure 8.1 clears up the misconception.

8.1. More graphic quality isn't always better.

The common misconception holds that more graphic quality is always better. People don't work this way! The value of almost any commodity to the consumer follows diminishing returns: the more you have of it, the less value each increase offers you. There are, of course, plenty of exceptions and special considerations, but in general, more is only decreasingly better.

Notice that I haven't provided any scales on the axes of this graph; our current situation could lie anywhere on this curve. For example, if current graphic capabilities fall well down to the lower left, then increasing graphic quality will still earn us huge benefits in entertainment value. On the other hand, if our situation lies closer to the upper right of the graph, then we are wasting our time pursuing increasing graphic quality. How can we calibrate our graph?

The Lessons of Cinema

We are fortunate to have a sizable database on which to calibrate our graph: the cinema. Over the last century, the movies and television have built up a mammoth database of material to which we can refer.

We can trace the progress of the cinema up the graph. For example, the talkies swept silent movies aside in a matter of two years; it's safe to say that this cosmetic improvement lay on the lower-left portion of the graph, where the benefit of cosmetic enhancements is high. On the other hand, the transition to color movies was considerably slower; the first color films appeared in 1926, and black-and-white remained viable through the 1950s. This slow transition demonstrates that color lay considerably further up the curve than sound.

The next development in cosmetic quality was the use of special effects, which I shall narrowly define (solely for the purposes of this argument) as any technique that generates images other than those created on a conventional set. For the most part, this refers to the use of models, bluescreening, and (much later) digitally rendered images. Such techniques were used at the very beginning of the 20^th century, but the clay models and miniature spaceships used through the 1950s never attained enough verisimilitude to accomplish much. However, by 1970, these techniques were getting good enough; from 2001: A Space Odyssey through the Star Wars movies to a variety of movies in the early 1990s, such techniques matured enough to claim a solid place in the movies. The introduction of digitally rendered images in the 1990s permitted even more spectacular results, and now such techniques are well established. But it is important to remember that these techniques are still confined to a subset a very significant subset! of the movies. Even when we measure impact in terms of box office dollars, the entertainment value of these cosmetic triumphs remains limited. The majority of movies, and the majority of movie dollars, go to movies that do not rely on such splendiferous cosmetics. Therefore, I think that it is safe to say that the movies are now far up on the upper-right side of the curve.

A textbook experiment was played out for game designers in the summer of 2001 with the nearly simultaneous releases of two major digitally rendered movies: Shrek and Final Fantasy. These two movies pursued completely different strategies. The creators of Shrek concentrated their creative energies on the story; the cosmetic quality of the movie was merely adequate. Final Fantasy, on the other hand, perfectly exemplified the values of game designers indeed, the movie was created and directed by the designer of a successful series of games. Cosmetic quality was the entire point and purpose of the movie; the story seemed almost an afterthought. The people voted with their dollars and the electoral result was a landslide for Shrek: It outsold Final Fantasy by ten to one. It's difficult to imagine more decisive evidence for the relative value of story and cosmetic quality, yet even today, some game designers insist that Final Fantasy was a great movie.

One could of course argue that a movie combining the top-notch cosmetics of Final Fantasy with the brilliant storytelling of Shrek would do even better, but this argument deliberately obfuscates the lesson of the two movies. A nuclear bomb equipped with a tape recorder that screams insults as it approaches its target would be even more demoralizing than a plain old everyday nuclear bomb, but that doesn't mean that insulting tape recorders are a significant military weapon.

LESSON 16

The absence of graphic information can be just as stimulating to the player's imagination as its presence.

Graphic Realism and the Human Imagination

Some argue that graphic realism stimulates the player's imagination. The relationship between graphic realism and the human imagination is considerably more complex than many designers realize; it works in both directions: Imagination can just as easily be smothered by excess information as stimulated. This basic truth has been a female secret for thousands of years, and most males are too dumb to catch on. Every woman knows that she can be more seductive by careful concealment of her allures than by direct revelation. The sexiest clothing accentuates a woman's curves without actually showing anything. This is best exemplified by a variety of female dances, such as the dance of the veils, the belly dance, or the cruder striptease. The entertainment value of the dance lies in what is concealed as much as what is revealed. Indeed, completely nude female dancers are generally regarded as less titillating than the semi-clothed variety. Graphic realism stimulates the imagination, but it must leave room for the imagination to run free.

The game designer must therefore exercise great judiciousness in applying cosmetics. One of the simple but little-recognized secrets of the first smash hit 3D graphics game, Doom, was its dark environment. The monsters were poorly illuminated, a trick the designers used to cover up the weaknesses of their rendering technology, but they had serendipitously hit upon a powerful way to stimulate the players' imaginations.

Cosmetic excess becomes injurious to good design when it suggests what is not true. Blood splashes are my favorite example of this. You shoot up the bad guy and gratifyingly realistic blood splatters all over. However, this blood has no reality other than its splashiness. It remains on the floor and walls, but it is not slippery as one would expect, nor does it coat objects nearby, making them slippery. An even worse blunder is the maintenance of images of dead bodies on the ground after the slaughter. Sure, it looks realistic, but you can walk right over the pile with no additional effort. A swarm of bad guys can charge your machine gun, piling up a mountain of bodies in front of your position, but the newcomers plow right through the pile as if it weren't there.

LESSON 17

Realistic graphics must be matched by realistic behavior; if you can't achieve the latter, ditch the former.

Cosmetic realism communicates the intended realism of the experience as a whole. Consider, for example, the classic cartoon scene in which Character A foists a stick of dynamite with burning fuse onto unwitting Character B. The dynamite explodes, and when the smoke clears, Character B is revealed singed but unhurt. "Ooo, that smarts!" exclaims Character B. Now, if this scene were played out in full graphic realism, with a real human being and real explosion, it wouldn't entertain. It wouldn't be funny, it wouldn't be exciting, it wouldn't be dramatic; it would be confusing and stupid. The disjunction between the graphic realism and the fantasy unrealism jars the imagination.

The best single gauge for the ideal degree of cosmetic realism is the indirectness of the conflict in the game. Socialization serves to deflect our primal conflicts into more indirect channels; the most direct (i.e., violent) forms of conflict are forbidden. Thus, direct, violent conflict is socially and psychologically unrealistic. Such themes, therefore, require a more cartoon-like presentation. Extreme graphic realism should be confined to games in which the conflict is pursued in the most indirect manner: legal dramas, corporate politics, and the like.

LESSON 18

Match the realism of the cosmetics with the realism of the fantasy.

Consider, for example, the examples provided by two movies: Jurassic Park and Shrek. Believe it or not, Shrek is actually the more violent of the two movies. For all the dinosauric violence in Jurassic Park, there is actually very little on-screen violence: Only one person is actually killed on-screen: the weaselly lawyer. All the others who die in the movie the animal handler, the fat programmer, the chain-smoking programming supervisor, and the Aussie hunter suffer their fates outside of the camera's direct view. Even the sacrificial cow and goat are dispatched off-screen.

Shrek needs no such visual delicacy. We see a scene almost exactly like Jurassic Park's lawyer-meal, only this time it's a dragon eating a prince and the scene does not shock; it's funny. In addition, we see a graphic torture scene, a character catapulted high into the air and crashing through a roof surely violent enough to kill, but the character survives unhurt. We see four characters smashed under a huge barrel, a bird explode, and a frog and a snake violently abused. In Jurassic Park, a boy falls 20 feet; in Shrek, the donkey falls 50 feet. Shrek has two fight scenes with a number of characters badly beaten up; Jurassic Park has no violence between people.

The conclusion seems obvious: You can't use realistic violence in mass entertainment. You can suggest it with realistic imagery, turning the camera away at the moment of violence. Or you can reduce the graphic realism and retain the violence. Either way, the basic rule is clear.

This lesson runs counter to current industry wisdom, which extols the most detailed graphic realism in games of extreme violence. This is an understandable phase of an industry still in its youth. There is, to be sure, a permanent segment of the population that revels in psychopathic fantasies, but if the industry is to reach a larger audience, it must cater to the requirements of the larger population.

Other Reasons for Overdone Cosmetics

All the remaining reasons for snazzy cosmetics are wrong-headed, yet they play far too large a role in the thinking of game designers.

Reason #2: To permit the player to show off the superior cosmetic capabilities of his new computer. There was a time when this reason actually sold games. New graphics cards were rare, expensive, and special, and when somebody dumped $500 on one, they were quite willing to pay $50 for some software that would justify the expense. Nowadays, however, most computers come equipped with strong graphics capabilities, and the people who pay extra to get the latest, most powerful graphics capabilities are not numerous enough to justify utilization of those superior capabilities.

Reason #3: To show off the superior technical prowess of the programmer. Again, this was once a major factor. Game designers were programmers, and showing off your talents was one of the perquisites of the job. People appreciated those snazzy graphics tricks, and a number of people developed reputations solely on their ability to produce stunning graphics effects. Nowadays, however, this kind of grandstanding is no longer tolerated. It still plays a role in the thinking of some team members, and subconsciously seeps through the organization, but usually this reason uses Reason #4 as a Trojan Horse.

Reason #4: To keep up with the competition. This argument is nonsense, but it still gets plenty of mileage in some companies. The obvious riposte is the same one that parents use with whining children: If everybody else jumped off a cliff, would you do it, too? If the competition wants to waste thousands of dollars chasing after some stupid graphic, that's their business. If you can't demonstrate that the graphics are justified under Reason #1, then aping the competition is a waste of time and money.

Reason #5: To provide the player with images and sounds that are intrinsically pleasing. This one is a bit tricky, and requires careful distinction from the valid arguments that arise in Reason #1. Perhaps the best way to understand the flaw in this reasoning is with a variation on the old reductio ad absurdem argument. Consider the Mona Lisa. Now, there's a nice painting, and it certainly is pleasing to gaze upon. So, why not interrupt the game action every now and then and show the player a picture of the Mona Lisa? Sure, it sounds ridiculous in this context, but there are plenty of games out there that commit pretty much the same blunder. If the image or sound in question advances the play of the game, then run with it; if it's just a pretty picture, lose it!