Essay #5: Games - nuwen.net

The world's a lot less scary when you're carrying a rocket launcher around with you

Abstract.

Three first-person shooters with a plot are examined: Half-Life, Halo, and Deus Ex. Their flaws and successes are noted, as well as the impact of such flaws or successes on the immersiveness of the game and how well it can communicate its plot. Deus Ex is found to be the most daring game of the three; though it suffers from flaws, they are caused by fundamental problems arising from the nature of interactive storytelling.

Essay.

Electronic games (variously called video games or computer games) are a new form of entertainment, like books, plays, and movies. Powered by technology, computer games have the potential to be the most interactive form of entertainment ever developed by man. And as it so happens, interactivity is fun.

A major obstacle to the widespread recognition of computer games as popular entertainment has been the same thing which powers the games - technology. With primitive technology, primitive games are the only thing which can be made. This resembles the situation of silent films, but is much more serious; you can't well convey a plot when all you have is CGA graphics and a few dozen kilobytes of memory. At this point (the close of 2001 at the time of writing), computer games are just finally becoming "powerful enough".

"Man is a game-playing animal, and a computer is another way to play games" - Scott Adams
Why are computer games fun, anyways? It turns out that they are fun for different reasons than why books or movies are enjoyable. Games don't often try to imitate books; there is essentially no market for text-based single-player games anymore. (With the proper programming, I think there could be - but that's another matter entirely.) There was a brief fad of "interactive movies" (around the time when "multimedia" was an important word); games which were heavily based on full-motion video and allowed the "player" to make decisions at various points. These failed. Miserably. No, games are best when they do something unique; only computer games make it possible to play a real-time strategy game (a supercharged board game, if you will), or a first-person shooter, or a role-playing game (supercharged pencil-and-paper games), or a flight simulator, and so forth. Many genres of games are fun solely because of their interactivity, such as RTSes or flight sims. You can be a commander of human forces fighting against the terrible orcs, building up a base and slaughtering the enemy, or you can test your hand at building bridges which try to support themselves against the forces of gravity and other insults. Books allow you to be present while great and unusual situations develop (I have stood by while Hari Seldon developed psychohistory from a vague idea into a powerful theory which could save humanity), and motion pictures allow you to actually watch events take place (I have watched while Lain accepted her place as goddess of the Wired), but computer games allow you to actually play a part yourself. In a computer game, you can actually be a theoretical physicist. Well, one who has to save the Earth from alien invasion. Or you can be an eight-foot-tall SPARTAN II cyborg, who wears kinetic-assist MJOLNIR armor, can jump six feet in the air, and can run a one-minute mile (who is also trying to, uh, save the Earth from alien invasion). You can be a nanoaugmented stupendous badass, who is trying to save the world not from alien invasion but from its own (shadow) governments. This is only possible with computer games.

There are a tremendous variety of computer games, so I would like to look at just one genre; actually, one subgenre - first-person perspective games with a plot. First-person games make you become the main character in a way that third-person perspective games do not. Because of the primitive nature of computer technology, it happens that the easiest interactions between people to simulate are rather violent; this is why such games are generally First-Person Shooters (FPSes). Many FPSes don't bother with plot; the (violent) interactivity is quite enough to make a game fun. Sometimes, very fun indeed. However, it is generally agreed that all things being equal, we'd like to hear a good story while blasting alien scum nine ways from Sunday. A good plot makes a book enjoyable, and it has the same effect on a game.

Returning to why computer games are fun in the first place, what would be the ideal computer game? It would simulate in full detail (graphically, physically, and otherwise) a world - real or not - in which the player could do anything. Not anything within reason - anything. It would completely simulate social interaction, and no suspension of disbelief would be needed because everything would be completely believable at all times. We are far - incredibly far - from being able to create this ideal game. For one, it would require massively powerful rendering equipment for the graphics alone. True artificial intelligence would be required as well, the kind that passes Turing Tests, if the characters in the game were to be indistinguishable from actual humans. However, we are fortunate in that a game does not have to be ideal to be fun.

"Then there's the giant tentacle creature. Which is probably the most awesome collection of polygons we've seen in a 3D game. Naturally you want to unload several thousand rounds into its face. Except it hasn't got a face. And anyway, that's the worst thing you can do because it hunts by sound, menacingly tippity-tapping away on the metal grates surrounding it.... The tentacle beast... obviously resides in the lovely warmth of a missile silo, just below the main engines of a rocket. Firing that rocket immediately becomes your sole reason for existence" - Mark Donald of PC Gamer UK, reviewing Half-Life
There were several FPSes, such as Marathon and System Shock, developed around the time that Wolfenstein 3D and others were gaining massive popularity; Wolf3D lacked any pretense at a plot, but games like Marathon and System Shock did have something resembling an interesting plot. They remained niche games which never really attracted an audience. The first game which could possibly be called a popular thinking FPS was Valve's Half-Life. Half-Life cast the player as Gordon Freeman, a theoretical physicist fresh out of MIT (nobody's perfect), newly hired at the Black Mesa Research Facility, where interesting things are happening. Too interesting, in fact - Freeman realizes this rather quickly once the headcrabs start appearing. Half-Life was a fun game (modulo some of the final parts of the game), and it is also unusual in attracting a massive following even after the end of its lifespan, thanks to a number of expansions and especially multiplayer modifications created by its fans. But - was Half-Life really that innovative? It wasn't especially interactive, nor did it have anything in the way of character interaction. Gordon Freeman says nothing during the entire game save for breathing heavily in his environmental suit while scary stuff goes down, nor does he read or hear much of anything interesting during the game. Half-Life is also entirely linear; generally, there is one and only one solution to each problem that presents itself, and progress cannot occur until the player figures out what the problem is and solves it. The triumph of Half-Life is that despite all this, the game is fun - it makes a tired old alien invasion plot seem new again by being very immersive and well-paced. Half-Life is an experience, not a plot, which is why I need not bother describing the plot here. (This creates an interesting problem for historians; a book such as Second Foundation, or better yet The Odyssey, is a collection of printed words on paper, and the words can be copied and translated with reasonable fidelity over the centuries. A game is also a collection of bits, but to play it requires an entire computer architecture. Computers of the far future (read: 50 years from now) will have as much trouble playing Half-Life as our computers would have trouble playing the original Spacewar program. How will the experience of Half-Life be preserved for future generations? It is an interesting problem.) Half-Life has other, greater flaws (such as the last levels), but these are the result of a different kind of bad development decision, and aren't really intrinsic to the flavor of the game.

Next, let us consider a very recent game (at the time of writing): Bungie's Halo. Halo attracted some attention early on because it was a very beautiful game; eventually, it caught the eye of Microsoft, which acquired the developer (famous for, among other things, Marathon) and made Halo into a launch title for its new console, the Xbox. Halo is a very interesting game. (And a very fun game, if I need to bother mentioning that.) The first, and possibly most remarkable thing about Halo, is that it gets most of the mechanics right; it is a console FPS whose controls don't suck. It has no level load times to speak of, even better than Half-Life; this goes a long way to increase the immersiveness of the game. (It has chapter load times, but these are easily ignored and even enjoyed.) It has a game flashlight that doesn't suck. And so forth. Another remarkable thing about Halo is the attention to detail paid to the world, which is a relatively consistent simulation. When fighting Covenant Grunts (little, annoying alien enemies), it often happens that their commander (a Covenant Elite, a very badass alien enemy) is killed in battle. Upon seeing this, the Grunts will yell various things, such as "He's everywhere!" and then run away cowardly. Marines will often walk up to dead enemies and pump a few more rounds into them, just for the effect of it. (I once was coaching a newbie how to play, and I had to explain to him that though the Marines were firing, they weren't attacking him, only the dead alien; he had been shooting the Marines in what he believed was self-defense, making no one happy.) And while driving around the landscape in a vehicle, with a computer-controlled Marine as gunner, occasionally after taking down a squad of aliens the Marine will yell, "Now that's what I'm talking about!". This phrase is (independently) actually a favorite of mine and my friends, and amuses us endlessly. Little touches to detail like this add tremendously to the value of the game even before such things as the perfect weapon balancing, beautiful graphics, and interesting plot are considered. Halo is further remarkable as a prequel novel was released before the launch of the game; Halo: The Fall of Reach is an excellent book in its own right and greatly adds to the enjoyment of the game. However, Halo has its problems too. Halo is an entirely linear game, though not as bad as Half-Life; while battling aliens the player is usually a free agent, but generally there is no choice as to the route that can be taken through the game. More maddeningly, there are gaps in the exquisite detail paid to most of the game; these range from minor (no volume controls; narration is often drowned out) to moderate (the Master Chief is said to be skilled in all weapons and all vehicles, but is unable to pick up a fuel rod gun or a Covenant Wraith tank) to major (plot holes (or so it seems, at this point) involving the purpose of Halo). Worst of all, Halo (unlike Half-Life) is attempting to tell a story. I've mentioned that telling a story is a major component of making a game fun. In Halo, however, the story is massively cut. This is apparent after learning about what the game was to be like pre-release, and is very obvious after reading the prequel novel. Halo should have been much more in-depth. Even after looking at the storyboards for game cutscenes, cutting appears to have been extensive. For example: the main character is an eight-foot tall SPARTAN II cyborg wearing kinetic-assist MJOLNIR battle armor. He's a big, imposing fellow who makes quite the impression on the average Marine. One of the first cutscenes for Halo was planned to have various Marines gasping in awe at the Spartan as he walks up to the captain of the ship; later, as he and the Captain shake hands, a shot was planned to contrast the massive difference in their heights. This was not to be, and the final version of the cutscene barely makes it obvious exactly how tall the main character is. One actually never really gets a sense of the Spartan's height; though the marines seem short, while playing the game your legs are nowhere to be seen (a common occurrence in FPSes for some unknown reason). This has various, unintended consequences; the gravity on Halo seems to be really low, when in fact it is really Earth gravity, but you play an eight-foot tall cyborg who can jump six feet in the air. This detracts from the immersiveness of the game, and the shortened cutscenes and plot detract from the interactivity and fun of the game.

"Creating a really believable world is just insanely hard" - Warren Spector
Why can't anyone get this stuff right? Mistakes made in games are so obvious in retrospect (and, indeed, some are obvious from the beginning). Fortunately, there is one good example of an FPS with a plot that gets the important things right: Ion Storm Austin's Deus Ex. Deus Ex has a number of flaws (which, remarkably, the designers all recognize and have vowed to correct in Deus Ex II); its AI can be fooled too easily, its graphics are not groundbreaking, and it lacks some features which would increase its immersiveness greatly (for example, a physics engine). One reason these problems are so significant is because Deus Ex tries to simulate the real world, or something close to it (a world torn apart by nanotechnology and conspiracies, circa 2050); since even gamers spend most of their lives in the real world, it is very obvious when something is not right. In a game which simulates something other than the real world; say, Halo's simulation of Covenant Elites, the player doesn't know anything about how such an alien "should" believe (other than it should fight well). Whereas if you kill a guard and go into hiding, and then the guard's parter turns the corner, sees the corpse, and does nothing, you know that something is wrong. Or (as in Deus Ex), if the guard goes into a state of high alert, but then a couple of minutes later says "I must have imagined it" and goes back into normal patrol while his buddy is sprawled out on the ground, you still know that something is wrong. The standards for a real world simulation are unbelievably high. Deus Ex manages to pull most of it off; the gameworld is incredibly interactive, detailed, and immersive. Unlike any other game I've seen, in Deus Ex you can read books, datacubes, and E-mail which are lying around the gameworld; actual (and ficticious) clippings from novels can be seen. Deus Ex also gets game conversations right. The game will cut into a third person view while the conversation occurs. Decisions in conversations can be made, but only when they are key and will affect the course of the game. This resolves the old problem of having a conversation branch, which the gamer attempts to exploit by saving and reloading, trying every alternative to pump the most information out of the non-player-character (NPC). In Deus Ex, the main character J.C. Denton always says and asks what is appropriate in any given situation; to find out more, all the player must do is talk to the NPC again. (We lack total artificial intelligence, so the characters must eventually repeat themselves; however, it usually takes a couple of chats with an NPC before this happens and the suspension of disbelief is broken. Besides, if you were to continually question someone, sooner or later they'd probably start repeating themselves telling you to go away, anyways.) Deus Ex is a minute-to-minute nonlinear game; there are always multiple ways to get around obstacles (obstacles, not silly puzzles). The decisions the player makes affects the game not only on a timescale of minutes, but also long-term; Deus Ex is actually not a FPS but a role-playing game from a first-person perspective, so players become specialized over time.

One of the worst problems of Deus Ex is a fundamental problem with interactive stories. A completely linear story, such as Halo's or Half-Life's, allows the developer the most freedom to create an interesting experience. When the story is allowed to branch, then the actual story experienced by the player is limited because the developer still needs to take the time to create every possible branch of the story, but the player will only follow one path while playing the game (and most casual gamers do not replay games). Deus Ex solves this by being minute-to-minute nonlinear but globally linear; everyone who plays through Deus Ex gets the same basic story, but problems are solved according to the player's style. If there are secondary objectives (in an RPG, this would be called "side quests"), they can reward the player and enrich his experience, but they lead right back into the main story. This is a great solution to the problem of interactive storytelling. However, there is a point in Deus Ex where failure is forced; there is only one course of action which must be taken; without doing so, the player cannot progress. This is more or less vital; to allow branching at this point would essentially require the developers to program two games in one. But it still annoys players. Though while making Deus Ex, the developers recognized that forced failure should be minimized, now they realize that it needs to be hidden even moreso; it will be interesting to see how this is tacked in Deus Ex II.

"Ultimately, all I wanted was for players to feel like they were in the real world. I wanted them to be able to apply real world common sense to the problems confronting them, and I thought recreating real world locations would encourage that kind of thinking. There's also just a real power, a real thrill, when you fire up a game and see a place you've been or want to go, and then get to do all the stuff you WANT to do there but know you'll get arrested if you try! If that isn't the stuff of fantasy - far more than exploring some goofy dwarven mine or alien spaceship - I don't know what is!" - Warren Spector


http://nuwen.net/essay5.html
Stephan T. Lavavej
stl@nuwen.net
Updated a long time ago.