Building the Fourth Wall

Introduction

There's been a lot of discussion in game design circles recently about emergent gameplay. The running theory (or holy war, depending) is that games are inherently more fun to play when the game dynamics allow a large number of secondary reactions, many with consequences unplanned by the game designer. This allows the player to have more of a sense that he is directing the progress in the game instead of being kept on rails by the designer, because he is frequently making use of things the designer could not have had in mind.

While that certainly is emergent gameplay, I would like to suggest a wider meaning for the word. British director Jonathan Miller has suggested that, unlike a painting, sculpture, or piece of film, a play is not complete as text, or even following its original performance, but rather is completed by the combination of the text, original, and subsequent performances. The play, in short, emerges from the combination of text and repeated performance. Similarly, I claim that a game is not complete when it "goes Gold," but rather is completed in the experience of the player.

Certain plays have an unusual vibrancy because they allow significant variety and interpretation in subsequent performances. For one example, while there is certainly a "canonical" staging of Hamlet, set in thirteenth century Denmark and focusing on the stereotype of the brooding Dane, a variety of other stagings are possible, from placing it in other times and places to reinterpretations of the core characters. Similarly, the primary goal of the game designer must be to encourage creativity and variety from the player. The game will take its vibrancy from the player's ability to make decisions and experience their repercussion in the game world.

In this essay, I will propose one technique for achieving this level of player/game interaction. At the core of the technique is the idea that two things are necessary to maximize player creativity. First, the player must have a strong emotional involvement in the events of the game world. Without this emotional involvement, the player becomes detached from the game and loses investment in the choices he has to make. Second, the player must be fundamentally comfortable with the rules of the game world. This allows the player to think at a level of secondary, tertiary, etc. effects and reactions, where he is able to go beyond paths left by the game designer and find his own paths and solutions in the game world. I will call this ideal immersion.

While I have not seen these ideas set down clearly before, a number of games have attempted to create immersive environments, with varying degrees of success. I will start by addressing some of the game design issues that have prevented current games from reaching true immersion.

Crimes Against Immersion

I have stolen my title for this section from an essay written by Roger Giner-Sorolla entitled Crimes Against Mimesis. Writing about text-based adventure gaming, Giner-Sorolla noted the tendency for games to break the illusion that the player was reading a story and make the puzzle aspects of the games unnecessarily obvious. While some of Giner-Sorolla's arguments are not applicable outside of his gaming style of choice, and his attempt to remove both chance and "challenges of coordination" from games would defeat any attempt to be truly immersive, his approach still gives valuable insight into modern games.

(Of course, in the end, it was a lack of immersion that killed adventure gaming. The puzzles in adventure games became gradually farther and farther from people's expectations until articles like Old Man Murray's discussion of adventure gaming appeared. However, adventure gaming left a legacy of story-driven gaming that is still crucial today.)

The OED defines mimesis as

Imitation; spec. the representation or imitation of the real world in (a work of) art, literature, etc.

Giner-Sorolla takes this to refer to the impression that the player is reading a work of fiction, and his "crimes" are things that return him to the reality that he is solving a closer to linear sequence of puzzles. Similarly, crimes against immersion are anything inside the game that exposes elements outside the game fiction.

Heads-up displays

The heads-up display is an accepted and probably integral part of modern gaming. While there have been some successful attempts to move information from the heads-up display into the game world (the assault rifle in Halo comes to mind - the number of shots remaining was displayed on the rifle itself, much more usefully than in the HUD), I'm fairly convinced that HUDs are here to stay. Halo 2's experiment with removing the health portion of the HUD seemed to mainly confuse the gamer. While it's true that I don't walk around with a little bar in my left eye telling me how healthy I am, it's also true that I have definite feedback of stubbed toes and broken bones and, pending the publication of Microsoft's DirectPain interface, I'm not sure there's an obvious replacement. The same goes for one's mana/bioenergy/psy/blood/whatever pool.

The player's weapon is a somewhat more interesting issue. In the original Deus Ex, NPCs would notice if you wandered around with a drawn weapon, and wandering into a bar brandishing an assault rifle would draw the same reaction you would expect it to. For Deus Ex: Invisible War, the designers claimed that players regarded their weapons as "part of the HUD," and that it simply frustrated players to have to holster their weapons in non-combat areas of the game. Personally, I can't imagine having a nice conversation about the weather with someone who was pointing a shotgun at my chest, and I can't imagine many of you can either.

Stealth is one of my personal favorite middle grounds. For some time (I would blame this on Thief, but I don't know quite enough of my history of early PC gaming to be sure they were really first), stealth games have come with a shadow meter that told you how far in the shadows you were. It's a useful bit, and, especially in first person games, quite necessary. A more recent "innovation" is to try to cram more situational awareness into the stealth meter. An upcoming game called Stolen, according to a Gamespy preview, will feature a stealth meter that

isn't simply a gauge of how much light is falling on you, it actually measures how likely you are to be seen by an enemy. If the gauge is green, you're good, but it gradually creeps up as an enemy gets closer or as you move into the light.

Similarly, Vampire: the Masquerade - Bloodlines provided both the conventional "how bright is it" meter and an indicator that

shows the range at which enemies can see you — 0 means no one in range can see you; 100 indicates that there are enemies close enough to see the vampire in stealth mode. The colors on the range indicator show the "awareness" of enemies. . .

(from the game's manual.) In addition Bloodlines cramming enough information into a simple indicator to require significant attention to decipher, I'm not convinced that gameplay has benefited. It seems like it should be easy enough to tell from the behavior of guards whether they're alerted: if they're wandering around, shining their flashlights in dark corners, calling each other on the radio, etc., it's a decent bet that they've been alerted by something. On the other hand, if they're sitting around a table eating doughnuts, you can be fairly sure they don't know you're there (or are otherwise unalarmed by your presence). I claim that adding more information doesn't really improve the game at all. If you can't tell how alerted the guards are from their behavior, then the solution isn't to cheat and put an indicator on the player's HUD, but to go back and give your guards decent behavior. If the player can't see any guards, then there's certainly no way he should be able to tell (besides memory) how alerted the guards are, and the indicator will give him information that there's no reasonable way he should know, which completely breaks the player's immersion in the game as a fictional experience. As for 0-100 indicators, that's pure dice-rollery and has no place in a computer game.

Leaking Dice Rolls

To my way of thinking, this describes many of the biggest problems faced by at least current CRPGs. It's inevitable that a CRPG will depend on a certain number of numerical scores, and possibly some random events. Whereas a human storyteller (game master, dungeon master, head geek, etc.) may have no problem describing what happens when Hank, who's not so good with a rifle, takes a shot at the fleeing bad guy 40 Ft away, it's difficult for a computer to handle that without at least some quantification of "not so good." On the other hand, it seems completely unnecessary to expose the player to the game's numerical bookkeeping. If I shoot a rifle at someone, I don't have to roll dice or check defenses. I aim the gun at my opponent, as best I can, and then I pull the trigger. There's no need to clutter the experience up simply because it's simulated with a computer.

There are a number of ways that dice rolls (by which I refer to both chance events and most numerical bookkeeping) can leak into the player's experience. Obviously, things like floating damage numbers and 0-100 stealth scores are examples. In Bloodlines, whether you were spotted while in "stealth mode" (violation number one: when standing, you can always be seen, no matter how deep the shadows) was determined by comparing your sneaking feat (which was increased if you were in darkness) to your potential spotter's investigation feat. If your sneaking feat was higher, you weren't spotted. It didn't matter if you were crouched under a blinding light in the middle of a small passage, and the guard who was failing to spot you had physically run into you and was now walking in place because the passage wasn't wide enough for him to find his way around you. As long as your sneaking feat was high enough, you were as good as invisible. While Bloodlines was more obvious about than many games, it didn't invent that particular misbehavior. In at least Deus Ex you could be far enough in shadow, especially if you were in a corner, that guards couldn't spot you. Kick you, yes. See you, no.

Another common frustration is the over-use of random events. When properly placed, chance is a very powerful tool, and can go a long way towards allowing the player to define his character. Having obtained some money, does he horde it in the hopes of eventually being able to buy the triple-titanium body armor, or does he go off to the casinos and attempt to win enough for the triple-titanium body armor right now? When given two possible thefts, one of which is almost guaranteed not to alert the city police but which only has a low payday and one which might alert the city police but has a much higher payout, which does the player take? These give the human player a chance to shape the story protagonist and to take actions as he imagines the story protagonist would. Further, it gives the game a chance to realize how the player is playing, and adapt the gameplay experience appropriately.

However, far too many games make chance into a too-regular part of the gameplay experience. While Star Wars: Knights of the Old Republic was a wonderful game in many ways, I found nothing quite so frustrating as watching my character, theoretically an amazingly powerful Jedi warrior, attack a rank-and-file Sith trooper over and over again, only to inexplicably miss every time. Similarly, everyone who has played any of the Civilization games remembers the first time a trireme sank one of their battleships.

In both of these cases, it's possible to imagine ways in which the described events could have taken place. The Jedi, for example, might have been both cocky and drunk. On the other hand, as the player, I had no control over whether he was cocky and drunk, and no opportunity to steady his hand. There's no reason combat should be a set of random occurrences, and making it random simply reminds the player of the artifice of the game rather than the fictional reality it attempts to portray.

It has been claimed that "real-world" combat is far from an exact science, and particularly if the player character is not supposed to be a trained soldier, there should be some randomness in combat. While I can accept this perspective (and without it character development in CRPGs would become rather difficult), all chance occurrences must be motivated from within the game world and immediately understandable to the player. If, for example, the aiming reticle in the HUD changes size to reflect accuracy, then the player can immediately understand how accurate he is likely to be, and chance misses will be understandable. On the other hand, if the reticle is always the same size, but sometimes shots do not go where the reticle is aimed, the player is given no way to understand the chance occurrences, and must accept that he is playing a game with dice rolls dictating the results of his actions. The same argument applies to any other chance successes: if a player's stealth ability depends on a skill level, this should represented in a concrete way inside the game world. Similarly, if a player's ability to persuade people depends on a comparison of skill levels, they should be represented in the demeanor presented by NPCs and the dialog options available to the player, not as random successes or failures in otherwise identical conversations.

Once one begins to examine the modern CRPG, more and more examples of disjunction between the way the game plays and the game fiction become apparent. For example, many games reward each kill with some number of experience points. While it becomes harder to gain experience levels by killing lower level monsters as the player character gains levels, it's not necessarily obvious that killing an infinite number of goblins should leave the player character good at much besides killing goblins. Deus Ex took the opposite route, giving no experience for killing enemies but including a number of experience bonuses for finding secret routes, hiding places, etc. Unfortunately, this ended up either penalizing a combat-oriented player (who would have no need for secret passages) or resulting in a highly artificial style of play in which the player would defeat the enemies in an area, then go back and wander through the ventilation ducts picking up the experience bonuses.

Even the idea of experience and experience levels is, in some ways, flawed. All the experience in the world writing Haskell, for example, will not result in my getting even slightly better at writing novels. In the modern CRPG, the connection between playing style and experience points is almost completely reversed. The character sheet defines what the player can do, which then shapes (to at least some extent) his game play decisions, allowing him to gain more experience and develop his character sheet to further approximate the style of game play he'd like. While you could claim that experience allocation is simply an aspect of the game play, I know few people who play games because they really enjoy putting dots in columns. The character sheet is a means to enabling the real game play (shooting bad guys, sneaking past bad guys, feeding on bad guys — take your pick), but it inappropriately becomes the defining element of the game.

Unfortunately, relatively few games have made any attempt to address this problem. Deus Ex: Invisible War made all player character customization take place through bio-modification canisters that then became part of the game fiction. While this worked well in DX:IW, some things conventionally represented with experience, such as expertise with weapons, did not fit into the system at all and other things, such as computer hacking, seemed uncomfortable. Further, portable modification canisters are unlikely to fit into the fiction of many games.

Finally, it may seem like this is only relevant to games that follow the conventional role playing game model of character development. Many first person games, from first person shooters like Half-life and Doom to the Thief series, increase the player's abilities as the game progresses simply by giving the player more and better tools. However, conventionally, attempting to add more character development seemed to require bringing in the game play overhead of character sheets, experience, etc. On the other hand, any solution that allowed player customization to fit seamlessly into the game fiction could fit into a pure shooter or stealth game, adding to the gameplay without detracting from the main focus or immersive potential of the game.

This is not the crate you are looking for

There used to be a nice convention in adventure games. Room descriptions would look something like the following:

You enter a small, well-appointed kitchen. The wallpapers are a green checkerboard pattern. On the counter is an apple peeler. There is a door out to the patio to the east and a door to the dining room in the south wall.

The correct player choice was the take the apple peeler. It didn't matter whether or not the player saw any particular use for an apple peeler, whether he had any love for the apple peeler or whether, in fact, he was the peculiar kind of misogynist who would rather have his spleen chewed by rats than possess an apple peeler. No, the apple peeler would not have been mentioned were it not supposed to be taken, so taken it was.

In part, this was simply a limitation of the technology of the time. Neither the player nor designer would have the patience to handle complete descriptions of every detail in any room. Designers had to prioritize, and telling the player about things that he would care about is an obvious choice. While some games included red herring objects, either as jokes from the designer to the player or as a serious attempt to add complexity to the game, this could not address the core problem. If the previous description of the kitchen had included descriptions of five objects, the problem would simply have changed to figuring out how many of the five objects would be useful in the future. This still does not come close to capturing the experience of walking into a small, well-appointed kitchen.

It would seem like those days would be behind us. Fully rendered scenes came in about the time of Myst, and with a modern 3d engine, there's nothing to stop you from walking around the kitchen, kicking the cupboards. Unfortunately, very few games are designed this way. For every Deus Ex: Invisible War, in which almost every in game object was movable, throwable, and most were destructable given sufficient effort, you have four games like Bloodlines, in which not only are only certain crates destructable and certain objects throwable but the game displays special icons when you get close to them as if to say "this crate is special. You can bash this crate. The other forty crates you've passed this level weren't bashable, but this one is."

Plenty of new game design problems are introduced by following this approach. There's a purely technological problem: every movable object in a game level increases the amount of work the engine has to do. I'm confident, personally, that a combination of clever coding and ever increasing processing power should get that problem under control soon enough. Perhaps a more interesting problem is that of defining reasonable player reactions. For example, in many games, destructable crates contained ammunition, health packs, or what-have-you. In a game in which all crates are destructable, you run the potential of having players running giddily about the streets of New York, attacking everything square with their trusty crowbar. I think this is addressable by following reasonable conventions. Rarely does ammunition come in unlabeled boxes sitting around the streets, and there's no reason it should in a game either. While the player can still technically run around smashing everything square, he should shortly realize that it's not constructive behavior, and adapt to a more realistic behavior model.

"I am not a puzzle! I am a human being!"

I've taken the title of this section from Giner-Sorolla's essay because I think the problem that he described still exists in almost the same form. He wrote that game designers, by and large:

relegat[ed] NPC's to very simple roles: either roving menaces from a hack-and-slash campaign of Dungeons and Dragons (the dwarf and pirate in Adventure, the thief in Zork) or mere components of a lock-and-key puzzle (the troll and bear in Adventure, the cyclops in Zork).

The same is true today. In the modern CRPG, the majority of non-player characters are either simple enemies (guards, aliens, orcs, etc.), or fairly simple parts of quests. The techniques used to implement these two templates have gotten better. The Covenant (at least in Halo; I haven't played much Halo 2) used surprisingly interesting tactics, and gave a certain amount of life to what was otherwise an straight-forward shooter. The guards in Thief: Deadly Shadows noticed things like missing valuables, rearranged boxes, missing fellow guards, and other previous ignored factors and were appropriately alerted. Similarly, most neutral or allied characters have more detail than was possible in Giner-Sorolla's day. Almost all CRPGs give you some conversational options beyond "ask about ," and most attempt to flesh out at least a couple of characters. The relationship between J.C. Denton (the player character) and his brother Paul from Deus Ex comes particularly to mind.

However, interaction with neutral or allied NPCs continues to be completely scripted. By comparison, one of the biggest criticisms leveled against Doom 3 was that monster behavior was completely scripted, significantly limiting the gameplay choices available to the player. Nobody mentions that the same limitations are universal in conversation and other interaction with neural and allied characters. This limits the player to doing things that the game designer foretold, and frequently results in leaving the player feeling "railroaded," or forced to make certain choices against how they think their character would have behaved. Two examples come immediately to mind: in Bloodlines, the player at one point has reason to suspect that the leader of the Anarch faction has murdered one of the leading vampires in Los Angeles. However, the game gives the player no option to speak to the other Anarchs, instead forcing the player to return to the Prince and report the treachery -- no matter how much the player may dislike the Prince's faction and agree with the Anarchs. Similarly, one of the early pivotal events in Deus Ex is when J.C. follows his brother and leaves the anti-terrorist organization they both originally worked for and sides with one of the terrorist factions. While the player has plenty of reasons to agree with that decision, it is not left up to the player.

The solution to this problem is not obvious. The AI required to enable more free-form conversation and NPC behavior is not trivial, and it is not clear that much of it has applications outside the gaming industry. While it has become accepted for graphics engines to require long, specialized development, the same is not true for game AI, leaving the work to be done by small, design-centric developers who likely have neither the resources nor expertise to do it well. However, the recent success of the Havok physics engine suggests that the game world may be becoming accepting of more specialized developers and a more component-based method of engine development, which may open the door for raising AI to the level of importance given graphics, sound, and now physics.

"I can see myself from here!"

In a recent discussion on the IGDA discussion boards, Ian Weyna wrote:

When considering first person or third person view, you must consider how you want the player to form an emotional attachment to the character and the world.

For example, in the 3rd person view, the player does not become the character, but rather, becomes a puppetmaster, controlling the characters actions. This leaves the player and character disconnected. Lets use GTA as an example: "Did you at any point feel the emotions of the character?"

In first person view, the character does not become the focal point, but rather the world. The player is actually put into the role of the character, mimicing real life to a certain extent. What occurs, or should occur, is that the player develops feeling for their character. Eg - "Instead of bad guys shooting at Duke Nukem, they're shooting at me!" In this form the player is able to connect that much deeper with the game.

A first person perspective is thus imperative for an immersive game.

There are still a number of limitations with the first person perspective. Primarily, present monitors are not large enough to convey peripheral vision or the self-awareness that would make complicated climbing. jumping puzzles, or many forms of melee combat intuitive. Recently, games like Thief: Deadly Shadows have made small advances by leaving a player model in place, even in first person view. This allows the player to see his arms and other body parts when looking around, and also would give a better impression of the player's environment were he able to hang from bannisters, climb along pipes, etc. Unfortunately, given the breaks in immersion that are unavoidable in a third person perspective, these must be accepted as limitations in the current technology, and games designed to minimize their impact.

Playing God

In the beginning, I said that to achieve immersion, it was necessary for the player to have a strong emotional involvement with the world of the game and the actions of the player character therein, and for the player to have a fundamental comfort with the rules by which the game world operates. The remarks I have made between then and now have focussed completely on things that distance the player from the world of the game; as such, they are no different from the rules of perspective or a knowledge of the lighting instruments available in the modern theater. However, just as following the rules of perspective exactly does not guarantee that the resulting painting will be good, nor does a detailed knowledge of the capabilities and limitations of lighting instruments guarantee lighting that will support and reinforce a play, the rules I have mentioned describe only what a game designer should not do, not what he should.

(It is worth mentioning that there are plenty of excellent paintings which do not follow the rules of perspective, and that a performance outdoors at dusk needs no lighting instruments. I will return to this later.)

What is required is for the game designer to approach his game not only as a craftsman, but also as an artist. The tasks of the designer: to create a scene that is understandable to his audience (the player), to create characters that are meaningful to him, to create a situation that is compelling, and to develop these, during the game, in ways that keep the player involved, motivated, and compelled, are the same tasks set before the creative team of a film or theatrical production. In both film and theater, there are examples of productions which are considered successful, such as the summer action thrillers or Broadway musicals. There are also productions, rarely the same, which are recognized to have artistic merit. It is imperative, both for the production of individual, immersive games and for the improvement of gaming as a whole, for the gaming community to recognize and apply the same standards for artistry as are applied in the theater or film communities. The tasks are the same, and the result should be held to the same standard.

It would be futile to attempt to elaborate what needs to be done to create a game that has artistic merit, just as it would be futile for me to prescribe one technique which would always produce satisfactory lighting for a play. However, by examination of existing games, we can see some commonalities.

Baldur's Gate opens with an animated sequence of a large, dark, armored warrior throwing people over the edge of a tower. The game then describes (in text) that the player is young, has been in the care of a wizard for some time. The player finds out that he and his mentor must leave the fortress he has grown up in, and, after some initial simple quests, they set out. The player's mentor is killed on the road the first night, and he finds himself wandering through the wilderness, attracting followers, fighting trolls, and attempting to piece together who dislikes him and what is going on. Quickly, he finds himself joined by a less mature childhood comrade, a necromancers (alignment chaotic evil), and a number of other wanderers. Planescape: Torment begins with an animated sequence of a corpse (apparently) being wheeled through a bizarre morgue by an equally bizarre looking creature that appears to be some cross between a merman and a zombie. As the corpse is wheeled in, there are flashback sequences showing incidents from his life, starting tender and concluding violently, as eventually he sees his flesh scar and transform into the corpse. Finally, as the zombie walks away, we see the corpse begin to come to life. The game starts at that moment, as the corpse comes back to life as the player character and a nearby hovering skull offers his assistance and reads a message tattooed on the player character's back with (apparent) instructions for what he should do next. There are two important contrasts to note: first, the introduction to Planescape: Torment is focussed on the player and his character. From the beginning, the player is at the center of the story, as he is at the center of the game. Second, the introduction to Planescape: Torment conveys a sense of wonder and curiousity. In contrast, the introduction to Baldur's Gate draws heavily on well-worn parts of the fantasy epic. While these can provide comfort for the player, and place the experience he is about to have alongside well-known others, they also serve to suggest that the game will not be branching out particularly from those ideas.

(In fairness: many people I know enjoyed Baldur's Gate. I cannot comment on the remainder of the game, as it didn't keep my interest.)

Another example: each of the worlds in Star Wars: Knights of the Old Republic draws on a well known geographic area. For example, the planet Dantooine strongly recalls the midwestern plains, and the world Kashyyyk draws heavily on the pacific northwest. This results in the first experience a player has on a world to be a combination of surprise and recognition. However, the game developers do not extend these ideas once a player has arrived on a world. Having recognized that Dantooine is the midwest, there are no tornadoes, no prairie wildfires, nothing to develop or contrast against the original image.

Every area in Deus Ex is set at night. The exterior darkness becomes an accepted part of the game. The first area of Deus Ex: Invisible War is also set at night, suggesting that the same conventions will apply. When the player arrives at the second area and finds it to be bright daylight, it produces a sense of relief and wonder. While this affect is certainly noteworthy, it is worth questioning whether all the previous areas needed to be set at night. While it's true that night is associated with secrecy and isolation, two ideas prominent in Deus Ex, there are other ways to convey those ideas. Setting every area at night is simply a shortcut.

Finally, both StarCraft and Homeworld are real time strategy games set against an epic, empire building background. However, StarCraft follows many of the standard conventions of the genre (and Blizzard games in particular) presenting units who are cartoony, a standardized size that, while it allows the game to not need a zoom function, results in tanks that are as large as the buildings that build them, and a standard assortment of cracks, blasts, booms, war cries, and explosions. Homeworld, in contrast, has appropriately sized units, requiring the player to zoom in to see the detail on fighters or zoom out to see a battle in full. While this increases the interface, it conveys a sense of scale that StarCraft never achieved. Perhaps most effectively, the combat in Homeworld is conveyed through military-like radio comments, but without any sound from the fights themselves. Not only does this recognize that "in space, no one can hear you scream," but it gives the battles a sense of remoteness and scale that fits into the epic sweep of the game and story. As a result, all of Homeworld feels like part of an epic fight, while the battles in StarCraft are likely to feel like small, almost irrelevant pieces of a larger fight conveyed through cut scenes and narration. When the game attempts to have the player play through epic events, the limitations of scale and the cartoonish style of the game undermine the developers attention.

I claim that the central part of each success I have listed is a refusal to compromise, and at the center of each failure is a compromise. It saves development time to set an entire game at night instead of producing wind storms, rains, cracks of thunder, and whatever else could contribute greatly needed detail. It saves development time to develop one feel for each world instead of developing those core ideas into a progression as the player moves through the world. It is safer to produce a game in line with existing successes than to attempt to truly convey the epic story and different nature of the combat in an epic science fiction story.

These compromises undermine their games; in essense, they are artistic failures. The games I have pulled them from are all good games --- several, I would claim, are great --- but they indicate how even in the best games, an attitude of compromise and an inattentiveness to artistry haunt the gaming industry.

To create a game is to create and people a world; if a designer wants a player to be involved and motivated by the world he has created (and he must, if he wants the player to be immersed in his game), he must create a world that is as detailed, as vibrant, as filled with life as the world the player inhabits. This requires an artistic discipline that is rarely discussed in the gaming world, but is at the heart of every successful game and must be openly embraced if gaming is to advance.

Postscript: Rejecting Immersion

There are two bases for rejecting immersion as the central premise of game design. The first is that immersive games, such as they have existed so far, have not sold well. Deus Ex, for example, did not compare in sales to Doom. System Shock 2 pales in comparison to WarCraft III. This answer, however, ignores the potential for games to be more than profit for large publishers. I claim that games can be art, and that artistic concerns must guide game development; from this vantage point, we can recognize that sales figures are irrelevant to game design decisions.

The second is to point out that my suggestions have paralleled the naturalistic movement in theater and film. I readily admit this: I have proposed a naturalism that, while it is not tied to the particular nature in which you and I live daily, seeks to create things just as natural. Just as surrealism, expressionism, etc. followed naturalism in the theater, it could be claimed that naturalism in gaming must only be a phase, to be followed by better artistic ideas. Personally, I would welcome this change; however, it must come with the same artistic integrity that I have urged in a naturalistic view of gaming.

Patriot Dreams

I've put off writing this entry for a long time, because it is fundamentally cynical — but seeing a car with patterns of yellow and red/white/blue ribbons on its hood pushed me over the edge.

There was an article in last Sunday's Chicago Tribune about the changes to the memorial and tower planned for the site of the World Trade Center. The original plans for the space have been changed and mauled beyond recognition. What was once an off-center, life-affirming tower has turned into a monolith suitable for an Arthur C. Clarke novel. Not only is it not particularly affirming of anything besides a steel-and-glass corporate future, but it has all the architectural interest and originality of a square rock: none. We've all seen square buildings. To make things worse, some pointy-headed fool came to the conclusion that the originally proposed glass ground floor could pose a danger if someone exploded a car bomb next to it, so the lowest 200 feet is now slated to be concrete.

They're calling this fortress "Freedom Tower." If this is a monument to freedom, book me on a one way trip to Moscow, circa 1950.

I was going to put the whole thing aside as an example of bureaucracy in action when I came to the following part near the conclusion of the article:

There have even been battles over the arrangement of the names of the victims (should they be randomly located or should they reflect such things as clusters of co-workers or what Trade Center tower, north or south, the victims worked in?).

And I gave up. Why not have a grand lottery to determine the name placement? People's names can be drawn from one of those machines with all the little balls bouncing around inside them, the survivors and families can look on, hoping their loved ones will be next, and screaming and crying when their names are finally chosen. After a tight bidding war, the proceedings can be carried by ABC, and run against Survivor on Thursday nights. Finally, the proceeds from all of the above can contribute to the machine gun emplacements surrounding Freedom Tower in case somebody should attempt to drive a bomb-containing car too close.

Seriously, a big, square, fortress-like contribution to the "tallest building in the world" dick-measuring contest is not a monument to freedom and this kind of bureaucratic wrangling is no tribute to the people who were killed. I firmly believe that the terrorists won't win the war they started on September 11th, but I really wish a certain number of Americans would stop trying to win it for them.

Do It Yourself Education

The problem with my attempts to write personal journal entries is that nobody comments on them...

So, I graduated a little while ago and, in the midst of looking for a job (but not too hard) and trying to help around the house (but not too much) and catching up on some of the reading I've meant to do for quarters (or years), I've found myself thinking about graduation. I heard a lot of speeches over graduation weekend — about what we'd get from college, what would last after we'd forgotten all our classes, whether commencement was really a beginning, or really an ending, or really a transition, or really a continuation, or really a moment. (That last was our valedictorian. He also talked about getting Philly Cheese Steaks from the cafeteria and I didn't get what he meant by that either.) It came to a lot of words, and none of them really got close to what I was feeling, or, as a result, meant that much to me.

It is, though, tied to a couple of incidents from my last couple of quarters. One was a conversation with an 07 friend who was rather concerned about her department not offering any of the courses that tied into her concentration when she could take them. I didn't have much to offer her — besides sympathy, being a functional programmer in a department that wouldn't give a rat's ass for functional programming myself — but it reminded me of something of a common problem at D (and, I imagine, most other small schools). During my freshman year, and I suspect this true for most people, it seems like the entire world is available. Professors are nice and supportive, the course catalog is enormous, and every class is full of excitement. As time goes on, it turns out that the course catalog actually contains very little of what you're really interested in, and all the professors would be happy to support you doing whatever you're really interested in, but it's not really their area.

The story ended well for me. Even though my department fired the only programming languages professor my junior year, I found an emeritus professor whose interests matched closely with mine, and managed to have quite an enjoyable time researching and writing my thesis with him. I hope (and believe, actually) that things will probably go as well for my 07 friend. It certainly wasn't as smooth as it could have been, though, and I doubt it will be for her either. It really helps to be willing to do a lot of the footwork yourself.

The next incident starts with two things. First, I'd gotten a reputation (don't need to talk about what kind) around the theater department as a lighting designer. Second, the hip-hop dance group on campus, named SHEBA, lost their lighting person the week before their big spring show. As a result, I spent one Sunday near the end of Spring term standing in front of the biggest set of speakers I've ever seen, running a lights board and attempting to improvise lights for a dance show I'd never seen before. I think it ended up well — or at least no one in the audience knew enough to notice the places I fucked up — but I was left wondering why the most talented dance group on campus (at least the most talented student-run dance group) was left in the basketball gym with a set of rented lights and "tech rehearsal" that consisted of a one hour frantic run when the hired help finally got around to setting up the lights (about two hours before the show was supposed to start). I've seen much, much less entertaining, interesting productions get full technical productions from the theater department; how did SHEBA fall through the cracks?

There are lots of answers: SHEBA traditionally runs their spring show in the gym; there's more seating; the theater department already has more shows than it can manage; other things are higher on everyone's priority lists than a student organization with no (I think?) faculty oversight. The fact remains, though, that the show I was trying to keep up with and lighting in broad strokes (ooh! ORANGE!) could easily have made good use of technical capability the school had — but which just wasn't available.

I'm angling towards a point, I promise. In some sense, the freshman year impressions are all correct: the professors and staff at Dartmouth are all quite supportive (with some stunning exceptions), the course catalog really is quite large, and I think there are actually are pretty limitless possibilities. Unfortunately, it seems like relatively few of those opportunities come without a fair amount of, uh, creativity up front, and it's almost always possible to look back at something — anything — and find something that could have been done better, some opportunity that was missed. It's strange to look back at four years that I've described, honestly, as the best years of my life and mainly see regrets; I'm sure it would be possible to recognize that I'd done what I could and what I thought was the best thing to do at the time and to be satisfied with that, and there are times I wish that made sense to me.