A Mini-Review of the Game
Great games come along infrequently. Right now, I’m in paradise, having just picked up Thief: the Dark Project and Half-life. I couldn’t decide which one to start with, so like several people I know I’m playing both. Below are some my (extremely positive) reactions to Half-life.
Level Design and Locational Coherency
This is one of the game’s most beautiful aspects. The fiction is never broken (in any way that I have seen so far) and the sense of location is very coherent from the in-game intro onward. As the player, you feel like the Black Mesa complex could really be a functional location in the same way that System Shock’s Citadel Station felt real. Also, Half-life is not a game in which the player is following a string of pearls. Once you leave an area, it does not cease to exist-often you have to return to the area, finding the lone NPC who you could only look at through the window before. This is one of the best ways to create such coherency-the world is not a single-file trail through the woods, but a looping, contiguous labyrinth, with places that change as you revisit them. Starting on one side of a chasm in Half-life, then eventually looping back around to the other side (via a circuitous route), allows the player to look back across the chasm at the spot at which he once stood. And the feeling of progress and ‘locational coherency’ is powerful.
Also, like Duke Nukem and System Shock, Half-life creates a believable world by playing upon player expectations and real-world familiarity. If the player is told to move into an area where the power is out, he is likely to find a generator in the maintenance room and signs along the way. In the lounge area, of course, there are lockers, bathrooms and a microwave-equipped kitchen.
Feedback informs the player as to the results of his actions. This is one of the simplest, most fundamentally critical elements of computer gaming. Yet few games handle it as well as Half-life. Sound effect cues, voice lines and a simple graphical interface help tremendously here, but one of the more sublte ways in which the game provides the user with simple feedback is through behavioral consistency. The individual elements that make up the game are predictable. The player can learn what a particular monster, terrain feature or weapon is capable of doing; the behaviors of these things stay the same.
Some people-developers and gamers alike-are often tempted to think “predictability equals boredom” in games. Nothing could be further from the truth. Think about Tetris-the player, if given time, can always figure out how a particular piece will fit into the blocks below; there are no amorphous shapes or wild, random block behavior. Similarly, one of Half-life’s most brilliant features is the consistency of the behavior exhibited by (almost) all the game’s elements. This allows the player to a) easily learn how to deal with these elements and b) to work out cool combinations of interaction. The things that prevent this predictability from being boring are numerous-the combined interactions of the elements, the fun of dealing with them and learning to master them, the slight variations created by the player’s input and other ‘randomizing’ simulation elements. Plus, there are lots of these simple, predictable elements, not just a few. The key is that the game as a whole is not predictable-just the individual pieces.
For instance, in Half-life you have laser trip-bombs-if you break the beam, the bomb explodes. So the designers were able to use this device to create various game-play scenarios. There are places where the player must jump over the beam, duck the beam, where the beam is partially hidden, or where the player must trigger the bomb from a distance. In addition, there are two more features in Half-life that I’ll use in my example: crates and headcrabs. Crates hold supplies and can also be moved around for various reasons-for jumping puzzles, to unblock openings and for cover during firefights. Like laser trip-bombs and crates, head-crabs have known, predictable behaviors-they crawl toward you slowly for a while then they jump through the air at you. Each of these alone is cool and makes for some good game-play. But what Half-life manages to do is create a set of game-play elements that actually interact with one another, in addition to interacting with the player. So to cap off the example, imagine a room filled with laser trip bombs-their blue beams criss-cross the room, creating a 3-D labyrinth. Some of the beams are connected to crates, which the player would normally be tempted to break apart or push out of the way. And lastly, there are several headcrabs in the room, crawling about in search of the player. When they see him, they jump, possibly setting off all the laser trip-bombs. There is a powerful adrenaline rush that comes when the player actually sees the headcrabs and realizes the explosive threat they suddenly pose. Their behavior is still consistent-the player can predict what they will do. But now the combinations of elements in the room provide a new twist. This is an example of game-play elements that interact with one another. It’s this sort of thing that elevates Half-life above boring, toe-to-toe shooters and earns it my respect. As with System Shock, the game-play in Half-life sometimes resembles emergent behavior.
The Wave Concept
Like early arcade games, Half-life presents the player with fundamental problems to solve in the 3-D environment that come in waves. Initially the player is introduced to a simple problem, like an auto-turret guarding a passage. In the section of the map after this first problem, the problem is repeated at a higher complexity level-suddenly some creatures are added, a second turret, or the turret is positioned in a hard to hit spot. This, in my opinion, is one of the game’s most brilliant features. The arcade-inspired difficulty curve is extremely well balanced.
In other places this fundamental game-play thinking is also evident. For instance, take the ‘headcrab slide’ area, if you look up in the direction from which the headcrabs are raining down you can sidestep left and right to avoid most of them. I’m not sure if everyone will notice, but this is essentially the dynamic from Space Invaders (or any number of other games).
Other areas work in similarly classic “game problem” fashion. The recon scout maze is a fabulous example of what I’m talking about. The game is littered with other examples.
The truly brilliant part is how well these ‘puzzles’ are worked into the fictional and environmental context of the game. The notion that these game-play dynamics fundamentally resemble elements from classic games will probably never occur to most players, because they will be so completely engrossed in the game. They only become apparent when you step back and look at what’s going on from an objective distance.
Very few games have done this well in the recent past. For me, one of the most notable would have to be Ultima Underworld, by Looking Glass Technologies. (Remember the Pac-Man maze in the mines? A maze-like section of the map was filled with ingots that the player had to pick up in trails, all while avoiding patrolling ghosts…) Half-life not only does this, it does so in fairly subtle ways.
Ammo and Health Conservation
In a game, if the player has unlimited ammo and weapon access, he will rarely stop to do anything more than shoot. On the other hand, if the player occasionally runs out of shots and he has other game-play options, he is encouraged to think before pulling the trigger. Should I take this shot with my Big Gun™, or should I use the falling crates to crush the beast? If the player has unlimited ammo, he never has to ask this question. Half-life gives the player just enough ammo to cause him to pause and consider which gun to use, whether he should save ammo and avoid conflict, use the terrain in some clever way, or whether he should attempt to crowbar his way out of the area.
The health economy could have been a little tighter, but overall I love the way Half-life did it. The periodic health and energy stations work much like System Shock’s, allowing the designers to encourage a conservative, more thoughtful shooter game-play style by spreading them out at key points through the maps. I spent quite a lot of time at 90+% health and 50+% energy, though, so like I said, this could have been tighter. Doing so might have added a stronger sense of peril.
The three best things about them is that they are interesting, not too long or common and they do not stop the player from moving around in the game world. They allow the game to tell its story and to communicate to the player what is going on without being obtrusive.
On the downside, scripting events rather than simulating them can lead to broken occurences in a game. For instance, if you approach a certain guard in a scripted game and you miss his trigger, he will probably stand there stupidly. If the same guard’s behavior had been simmed, he would have reacted to the player or the player’s interaction in with the world. In Thief: the Dark Project, everything in the game is simulated at a complex level…monsters are not ‘triggered’ but rather they respond to the sounds and sights of the player in the world. If a fire elemental draws close to a torch in Thief, the torch is ignited because the elemental has ‘fire’ properties and the torch has ‘combustable’ properties.
Half-life’s scripting is so good and so extensive, for a shooter, that the limitations of scripting versus simulating never really become a problem.
Then why simulate if the same level of game-play can be accomplished through scripting? There are two reasons. First, if you simulate a behavior, it works globally. So once “torches are extinguished by water” behavior is established, it works for all torches and all water. On the other hand, to script a spot where a torch can be extinguished by water, someone has to work on that torch specifically and every other torch in the game as well. Simulating precludes a lot of grunt work. (In fact, years of grunt work.) Secondly, simulating allows for actual emergent behavior; the game often does unexpected things that no one ever intended it to do. In scripting, this can virtually never happen.
Recently I went to the Austin Computer Game Developers’ Conference. A friend, Rob “Xemu” Fermier, gave the most valuable seminar there. He is lead programmer for System Shock 2 and one of the original System Shock guys. And he understands immersion better than most of the people I know. I don’t want to steal from his excellent presentation (which you can catch if you go to the next CGDC), but the three things Xemu sees as critical to immersion are worth mentioning in relation to Half-life. They are: completeness (the player can do what he expects to be able to do), realism (the illusion is never broken, the player is not jarred out of the game) and investment (the player cares about what happens).
In Half-life, there is a fairly strong sense of ‘completeness’ in the game’s world. Occasionally this is broken when it occurs to the player to try something that is not possible. For instance, there are many ‘non-usable’ computer terminals scattered throughout the world. It’s a bit frustrating sometimes to have to pick through the various consoles, hoping that one of them is interactive. Limits on character interaction occasionally cause a break in the world’s completeness too. Normally, in Half-life, though, these things are not a problem because the limitations of the game (which are in part simply the limitations of the shooter genre) are so well hidden; the player almost always has multiple options in a given location. So Half-life’s “completeness” feels thorough.
Also, the level of realism is high-the world in Half-life often acts in ways the player accepts as realistic. The world and its inhabitants look and act in a believable way and the story unfolds in a plausible fashion.
If there is a flaw specifically in the immersiveness of the game, I feel like it comes from the ‘investment’ part. There are times in Half-life when something happens and the player doesn’t care as much about it as he should. I feel like Half-life encourages players to save the game before a new area, then move in repeatedly to watch all the possible variations in the scenario. This is great fun-sometimes the commandos win, sometimes the aliens win, and often the elements interact in some surprising way. But the downside is that death comes to be less meaningful. After a while you don’t care if a scientist gets gobbled up or killed by friendly fire. In fact, sometimes it’s fun to watch them die. So the player cares less, which equals less investment, which equals less immersion. Also, I’m not Gordon. So when people constantly call me by that name, I am somewhat jarred; no longer is it “me” there trying to solve the Black Mesa problems, but it is suddenly some other guy. (There’s a reason the character in System Shock did not have a name…) But I understand the trade off-sacrifice some of the player’s freedom to differentiate and express himself in exchange for stronger story-telling tools (namely in the form of ‘character’). These are minor points, but I really do feel like they detracted from my sense of immersion.
Creature Differentiation and Behavior
Visually, the Half-life menagerie is really amazing. The artists who did the modeling, texturing and animating deserve some serious applause. Likewise, the creature sound effects are first rate. Additionally, though, the design thinking behind the behavior of the creatures within the Half-life ecology and the fundamental roles they play was really intelligent.
This is such a key element to enjoying the game that I don’t want to ruin anyone’s fun by giving too much away. I’ll use an example from outside the game. Remember the T. rex scene in the original Jurassic Park? As a ‘game unit’ the t-rex could be described as a powerful, hostile entity with such weak perception that it can only spot movement. Stand still and you become invisible. This kind of unit differentiation allows the player to think strategically about how to get past the unit, which of course can lead to interesting game-play. I cannot say enough about this. Facing off against game units that exhibit some interesting, consistent behaviors is much more fun that going toe-to-toe with some giant boss monster. Instead of just shooting at the thing for five minutes, the player suddenly has more freedom in interacting and his problem solving abilities are engaged. Way to go, Half-life. Let’s hope more games catch on to those concepts.
Go get this game–it’s an amazing accomplishment. Though the technology and artwork are great too, Half-life is all about game-play. I’m going on record as saying that Half-life is one of the greatest 3-D shooters ever made.