One of the most important things to do when talking to game designers is to get information on upcoming titles. People who read these ramblings want hard-core facts, and the ques-tions must be structured to affect that outcome. But after countless articulations, it becomes a little dull asking the standard set, including, “How cool are the weapons?” or “What 3D cards are supported, and at what resolutions?” These things are important, but there is an-other gauge that is more subtle, and perhaps as significant — the excitement of the designer as he or she discusses their game.
It is rare to find someone who discusses their game with the remarkable enthusiasm of pro-ject leader Seamus Blackley. He does not sound like someone making a digital entertain-ment product; rather, he comes across as a gamer making a game about which he is “stoked.” During our telephone call, his voice modulated from quiet and introspective to loud and ecstatic, but his focus at all times was on the things he feels makes Trespasser one of the coolest of the upcoming batch.
His excitement is all the more notable for its endurance. In 1995, he and co-designer Austin Grossman left Looking Glass Technologies, and setup housekeeping at Dreamworks Interac-tive, the upstart gaming arm of Stephen Spielberg’s film studio. The duo had been paired on the renowned PC title System Shock, and Spielberg approached them about creating a li-censed game based on his special effects-laden dinosaur flicks. Eager to evolve 3D games, the pair set forth with a group of engineers, programmers, artists, and designers to bring the dinosaurs back to life on their own self-contained, digital island.
The prolific team started with little more than the blessing of Spielberg, and the ambition to create unprecedented realism; no compiler, no graphics libraries, and no age-old DNA were on hand to ease the effort of the initial steps. As the lead designer states, the process of cre-ating a foundation-shaking game from scratch can be “a nightmare.” Now, in 1998, the fruits — or perhaps underfed brutes — of their labor will be unleashed on our computers. Pick one scene from either film, augment the terror through making it a first-hand predicament, and that is the game. Still, as thrilled as we are that the release is at hand, no one is as excited as Blackley, whose singular goal for the project from the start has been for it to be fun. Is it possible to doubt such ardent enthusiasm?
The phone rings, and Blackley picks it up and responds with a terse, “Hello?” We exchange pleasantries, and I am asked to wait a moment before commencing the interview. Furious tapping, not unlike the sound of someone hammering down alien aggressors in an arcade, can be heard in the background, then a triumphant voice issues forth over the phone:
Blackley: All right! I win! I catch six bugs. I am in the eternal battle between the testers and the programmers, and feel like I am making some headway.
Is that a difficult role?
Well, it is interesting to be the producer, write the physics engine, and be on the program-ming staff. It gets a little hairy.
What has the journey of creating this game been like?
We thought of the initial concept when Spielberg came to us and said he wanted to do a Lost World game, and that he didn’t want to do “a licensed game that sucks.” We said, “How about we do a digital sequel to the film?” He was really into that, because he is a big gamer. That was in early 1996, and we had just been hired on at the end of 1995 after leaving Look-ing Glass. We had nothing, not even a compiler, so the game was born from a standing start. No programmers, no test database, no graphics libraries, nothing. We had none of the stuff we needed get it done, so it has been a hell of a journey.
How arduous has it been taking this concept from nothing, and coding it into exis-tence?
Immensely difficult, but at the same time very rewarding. There is nothing like the feeling you get when things start to work. When you see a tester sitting down and really enjoying them-selves while playing it, that is the payoff. It is always a struggle to do something new. For instance, our enemies don’t have guns. The Raptors and T-Rexes have to run up to you and eat you. Aside from the fact that no one has done physically modeled characters and game-play before, no one has had enemies in a 3D game that didn’t shoot back. No one has done puzzles like we have done. All of these different unknowns have been terrifying on one hand, but as things start coming together, it’s a tremendous relief, and we are very proud the ideas are working. On top of that, as a gamer, I am incredibly stoked, because amazingly cool things are happening in this game that I have never seen before.
Has Trespasser been modified from its initial concept?
A little bit. When we first started working on it, we thought we would make this amazing, big environment that could tell a cool story, like when you were a kid and would explore your backyard. Sort of like Shigeru Miyamoto, the incredible genius at Nintendo. His concept of game design is to make it feel like when you were a kid, and were exploring behind your house. We had the idea of creating a game like that. What we discovered is that it was ex-traordinarily fun but at the same time we were missing something if we didn’t go for the dino-saur combat and puzzles.
I imagine most of the reasons for the long development are technical?
Absolutely. And again, because we just started out with nothing. No game we are competing with this year, or supposedly last year, since it seems like everybody has slipped a bunch, has started out from nothing like that. It is incredibly hard, and requires a lot of people, and a huge amount of time and effort. Actually, these days, with all of the graphics hardware, changes with DirectX, and different machine configurations, it can be a real nightmare.
How supportive has Dreamworks been throughout all this? Was there ever a time when they pressured you to put a wrap on the game and release it, or have they been supportive throughout?
Remember, this place is run by Spielberg, and he is interested in quality. The reason I came to work for Dreamworks is because of that. Here is a place where people know how to make money based on creating outstanding products. We wanted this to be a really great game. A lot of game companies say, “Yeah, quality products are the most important thing. It’s very, very important to ship quality products.” But when it comes down to the grind, and the board is on their back, it’s, “No, let’s just ship it!” Yet every single time, the company goes down, because no one purchases a bad game. Spielberg and the other people running Dream-works learned that lesson a long time ago in films. It can be very painful pushing back dead-lines, and it’s a risk, but in the end, the only successful products are those that take risks.
How involved is Spielberg with the development of Trespasser?
He looks at it frequently. It’s kind of funny, because he has no time for anything. He is an extraordinarily busy guy, but he really loves games. He comes by a lot and plays it, and makes exactly the same comments about things that gamers do. I mean, he is a true gamer. So he is terrifically supportive, given the kind of stuff we are doing. What can I say? It is rare enough to have a boss or CEO who is himself a gamer and knows what he is talking about, and is aware of what else is coming out. It is rather incredible to have a guy as powerful as Spielberg be involved with the project.
All right, this is a loaded question. What are the differences between this and other 3D games? After all, this is not a standard 3D shooter.
Back when I did Flight Unlimited at Looking Glass, I had this concept of the standard flight simulator. All flight simulators at that time had the same features. It was like, “What is your weapon targeting like?” Products were not judged by how cool or fun they were, or how good a time you had playing them, but by the implementation of the standard feature set. The same thing is happening with first-person shooters. It’s as though there is this standard rack of questions like, “What does your water do? Do you have multi-texturing? What is your col-ored lighting like? List your weapons and how are they different.” And I guess it is possible to purchase the first-person shooter construction set, but Trespasser doesn’t fit into that mold because it’s not a shooter.
Sure, we have modeled close to 30 licensed weapons, and we have all these animals and dinosaurs and junk to shoot, but the guns in our games are objects just like any other object, and that is the fundamental difference. One of the things I like to do when I show the product is to have a gun in my hand, and rotate it. You can point your hand around, rotate the gun, and look at it. People are like, “Cool! You can rotate the gun around!” And then I walk over to a table, put the gun down on the table, and use my hands to push it around. Everything in Trespasser works like you would expect…. Whether or not you make it through and survive all the puzzles is more a matter of being inventive enough and clever enough to figure out how to get around them than it is a matter of you second guessing the designer, and finding the keycard or lever. If you need to go over the wall, break the park bench apart, or shoot it apart, take the planks from the middle, put it against the wall, and run up. Don’t go looking around for the lever to open the gate.
Hearing you describe these things, I feel inspired to ask one of those standard ques-tions. How simple will the control set be?
We have spent a tremendous amount of time making the interface easy to use — it weighed heavily upon us. We call the first level in our game the Quake re-education camp! [Laughs] For instance, there is a door, and Quake players just stand there hitting the space bar, when all they have to do is open the door. So we’ve put a tremendous amount of time and effort into the interface in order to make it really, really intuitive to use. Now, much like a fighting game, your success is going to be directly related to the skill you acquire using that interface. There are five buttons, and you have a hand. You can pick things up, manipulate them, and put things together. You are not going to get it all at first, but we are not going to require you to do it all at first.
Is this game attempting to evolve the 3D genre?
That’s a safe assumption. I mean, Quake is a great game, and all the 3D shooters are very alluring, but Austin and I feel that interactivity with a game world is really, amazingly fun. We have testers who come here and spend two hours putting planks together, floating them out into the water, jumping on them, throwing stuff in, knocking things down, and playing with the world . It’s unbelievably, incredibly fun to do things like that. On top of that, you haven’t seen anything until you are holding a shotgun and shooting at a Raptor, and he jumps on top of you, knocks the shotgun out of your hands, stands on your chest, and starts eating your face! That is just outrageous! You just don’t see that in other games.
We have this warehouse sequence where you are trying to get up on this scaffolding, and the animals are thrashing around, crates are flying everywhere and breaking apart, there is this giant, clamoring noise, and the animals are screaming at you and trying to shake the scaffolding apart. At the same time, you are trying to knock oil barrels down to kill them, and the barrels are bouncing off their backs, and the animals are hitting the ground and scram-bling to get back up. That is just how it goes in Trespasser, and that comes for free. It comes as a result of all the work at the beginning getting the physics right.
So we should feel things we have never felt before while playing a 3D game?
Exactly! I want it to scare the hell out of you!
But will it remain consistently fun?
I have to underscore that my goal at the beginning of the project was to make a world that was consistently real enough that it would make for a great game. I am not trying to simulate the world; nothing goes into the game that is not fun. What would be the point? I have been doing physics for computer games since 1993, and have produced a lot of really dull stuff. I have produced a lot of neat stuff as well, but the point is, if it’s not fun, and you are not enjoy-ing it, and it is not working as a game — independent of whether or not it is a perfect simula-tion — then I don’t want it. So the razor of Trespasser is, “Is this fun?”
How fundamental are the physics to this being a new experience?
The simulation engine, which includes the renderer, the physics, the sound, the triggering system, and the dinosaur AI, all stand on top of one another and produce a consistent world. If the world is consistent, and you believe in it, then, when you are attacked by a pack of Raptors, it’s going to scare you because it’s a real threat. You see the T-Rex move stuff that is immovable to you, and that’s a really big deal. If the T-Rex is chasing you, and he knocks over a mobile home to get to you, that’s incredible. You are running for your life.
How integral are the physics to the animation?
The animation is all physics-based, so the animals can handle any situation on any terrain. There is a skeleton underneath, and there are muscles stretched between the bones, and the muscles fire according to a locomotion AI. That makes them walk, run, jump, turn, bal-ance, and do all kinds of things. Also, the joints define various skin frames that are used to wrap the bump map’s surface over the dinosaur. The AI system gives high-level commands to the skeleton, and that’s how the game works. The AI system will say, “Bellow!” or, “Bite here!” or, “I need to put you over here with your head looking in this direction,” and the phys-ics says, “All right!” and does its best. If it tells a dinosaur to put its head inside a tree, the animal will run up to the tree and break its neck against the trunk.
What are some of the things we can expect to see the dinosaurs doing, or what kind of behavior is being fused into the dinosaurs?
That’s a very good question. [Pauses] The point is not to beat you at chess, or to have the best path finding so players will say, “Look! The creature found his way around that crate.” After all, dinosaurs are kind of stupid, although Raptors are reasonably smart. The goal of our AI engine is to make believable animals that look as though they are hanging out in their environment, and you are trespassing. The animals are just going about their lives! They have lots of behaviors like eating, drinking, stalking, and things they are going to be doing to you and the other dinosaurs. You are only going to get in trouble if you can’t find your way around them.
So most of the puzzles involve finding clever ways around danger?
Yes, though we do have keycards. It’s great! You pick it up, and it’s this little object, and you stuff it down in your belt. You can see it down there, and to use it, you grab it with your hand, hold it up, and rub it against a card reader. Then you put the card back in your belt, grab the door handle, and open the door. In order to get to one keycard, you have to look at the pack-ing manifest of some cargo that is in the warehouse by the docks. In the warehouse, there are some crates, which block the passage to the keycard. Of course, while climbing over the crates, you knock a bunch down. There is a Raptor tribe that has its nest near there, and when you make all that noise, they creep into the warehouse to investigate. So they are look-ing around, and are really curious, making calls to one another and stuff, and there is a leader in front of them. It is amazing to watch them interact.
But now they are in the warehouse, and guess what? We have it set up so they are not leav-ing. If you want to get out of the warehouse, you are going to have to get past them. If you are careful, they will just be curious, and in fact I think those guys are set up so that, unless you attack them, they will just follow you around. Now, they might be curious to the point where they will come up and nuzzle you, and push you around and stuff, but you have to keep your cool, because if you annoy them, they will eat you instantaneously.
So survival is contingent upon us using our minds to understand and respond to the behavior of the dinosaurs?
Absolutely! We have seven different species, and a bunch of different individuals per spe-cies, all of which have different personalities. For instance, there will be Raptor tribes that are going to be incredibly aggressive. We are setting up one of the final fights now, where you come across three Raptors eating a downed Stegosaurus. They look around, and as soon as they see you, they go for you, just to kill you, and it’s a really difficult encounter. Of course, that’s a situation where your brain just doesn’t matter.
Talk about your approach with the main character, Anne.
We wanted to have a character that was like a real person, so we thought we would have a woman. Actually, we decided on that right before Tomb Raider was released. We were like, “Damn! A female lead character!” because that was extraordinary back then. But she is from New York, is a regular person, and decides to go on vacation in Costa Rica. For whatever reason, she takes an aerial tour, but there is a problem with the plane, and it goes down in the ocean. She washes up on this island, Site B from the film, and has nothing except her clothes. The idea is to get off the island without getting killed.
What approach will yield the greatest success?
Learning about the environment, and figuring out what does and does not work. All of the gameplay evolves out of the actual configuration of the island. For instance, you need to get to the generator that starts the power; you need get to the town to find the satellite dish; you need to get to the radio transmitter; you need to get to the helipad, stuff like that. In order to do that, you have to deal with the terrain, and the creatures that inhabit it. The island is an amazing, decayed empire.
In terms of structure, is the game one, large, contiguous island or is it broken down into smaller, more manageable chunks for the gamer?
The game is broken down into manageable chunks for us! We have nine levels, and each level has close to 15,000 objects in it, and is 2×2 square kilometers in size. That’s just an extraordinary amount of data. Our clipping plane is close to 2 kilometers, so you’ll see 35 to 50 million polygons at all times. To do that requires an editor that is par excellence, so we use 3D-Max plugged into an incredibly stoked machine, and we have an export process that boils all of this down to something you can actually run with a good frame rate on a Pentium 166 with 32 megs. So the editor is an amazing, prodigious piece of technology, and we have divided the island into these workable areas. At first, we had this idea that it would be one, big space, but our data structures couldn’t handle that, though our original memory manager could. Still, why beat yourself?
At E3, there were reports of the game running quite slow on hardware. How formidable a task has this been accomplishing these things with acceleration?
It’s been an incredible struggle to get it to run on hardware at all. The main technologies that are going to get this game running faster are non-standard rendering technologies. There is a reason you don’t see outdoor games, and that is, it is incredibly difficult to draw anything outdoors that looks good because there is so much stuff to draw. So we use temporal ren-dering techniques, and we have a whole bunch of specialized primitives. For a long time, we thought we would just ship a software version, because the amount of geometry we have to render has surpassed even the capabilities of today’s accelerators. If you attempt to draw 20,000 trees in a scene with a two kilometer back plane, what are you going to get? You are going to get a sick Voodoo 2 that is unhappy with you. Also, imagine the amount of overdraw you get in a forest with a couple of bump-mapped, smooth-skinned dinosaurs running around in it. On top of that, add water, sky, and other prodigious processes.
What happened is we struggled to figure how to make it better on hardware systems, be-cause like I said, in order to draw the outside, you have to use primitives and rendering tech-niques that are just not supported by today’s hardware. The current generation of cards is designed around Quake II. Period. That’s what the hardware does; it plays Quake II. We are interested in drawing something newer and more innovative. Figuring out how to hack the hardware to do it is difficult, but once the hardware guys saw what we were up to, they got enormously excited, and now we have to turn them away because we don’t have time to talk them through all this stuff. We have a hardware implementation that, in the words of one of the biggest card manufacturers, is “a showcase of this generation of hardware.” It does things you have never seen before in hardware, which is very cool.
Such as rendering the huge outdoor areas, the high polygon count, and the detail. Aside from some of our rendering primitives, such as the skinning of the dinosaurs, which are all bump-mapped, and some other surface-mapped items, we can draw everything in hardware to the horizon. We can put all our temporal rendering techniques on the hardware and make it speedy.
What changed your mind about including hardware support?
More than anything else, the proliferation of hardware, and the second generation of cards, which have the flexibilities that allow us to do most of our rendering in hardware.
What are some of your favorite aspects of the visuals, from the creature models to the environments?
Just the feeling of environment. At the start of the game, you are walking up a switchback road on a volcano, and can look down and see the whole island, and it is totally breathtaking. You can just stand there and stare at it, and look down and see where you came from. In the forests, you feel like you are there, hanging out by a pool of water, and it’s dark, and there’s a Stegosaurus walking around, looking for food. It’s completely amazing. I’ve talked to a lot of people, and they want to know about the action, and the gameplay, and all of that, but you know what I think? I think we will have a lot of closet Trespasser players just hanging out in the forest.
Describe the implementation of the audio effects.
We have a real-time Foley system that works hand-in-hand with the physics system. The physics system moves everything, so everything in the game knows all about itself. When a dinosaur makes a series of sounds, we drive all of the sounds from the physics. Foley and a number of other people created the art of attaching sounds to things such as footsteps. You must have seen, at some point in your life, a couple of Foley artists watching a movie screen with shoes on their hands, hitting gravel. We have a virtual Foley system that processes the different effects in a scene so that it sounds like objects hitting one another at the appropriate velocity. The first time we took a crate and dropped it, and it tumbled down a ravine and made the right noises, our sound guys from Sound Deluxe, who have been doing Foleys for about 30 years, almost cried. It did, in real-time, what takes them an entire day. When you get to the point of a Triceratops running through a construction site knocking oil barrels down onto a decrepit truck, it is just ludicrously cool.
Have there been compromises that perhaps can be attempted in a future generation product?
There have been huge compromises. Whenever you try to do something new, you are figur-ing it out as you go. For example, you are not going to be able to see Anne’s feet. I was like, “Hey, I am running this physics engine that’ll locomote dinosaurs, the player will just have feet, too!” Well, that is just not fun. You look down, and your feet do not do what you’d ex-pect. Our locomotion is perfect; it is world class and no one else has done it in real time, but seeing her feet just didn’t work out.
At the same time, there are things we never thought would work that turned out great. We were going to make it so the doors would open automatically, and the testers were like, “No, man! It’s really cool to have to grab the handle and open the door!” So we have been around the block with these things, and played the game, and played the game, and played the game, and tried things, and tried things, and tried things, and had gamers come in, and their friends come in, and designers from other projects come in, and console people come in, and CEOs of other companies come in, and it has been test, test, test! And no one has ever played a game like this before.
What’s the range of the impressions of all these people?
The people who have the hardest time are those who play first-person shooters. Trespasser contains a lot of foreign concepts. For instance, you walk up to a gun, and the gun is just like any other object. It’s leaning against something, or it’s in a cabinet, and you have to open the cabinet, reach out, and grab the gun. To shoot the gun, you have to point it in the right direc-tion. It’s very obvious to people who haven’t played a lot of 3D games. In most 3D action games, guns are little more than 3D bitmaps that point to the middle of the screen, so if you want to lean around a corner and shoot, you can’t! But in Trespasser, you can.
What is the final verdict of the 3D gamers?
What happens is, they complain for a little while, then almost universally go back to Unreal and become frustrated because they don’t have the same level of flexibility as in Trespasser. There will be some boxes in the way of a corridor, and they are like, “Why can’t I just pull that out of the way, or take the gun and wedge it out of the way, like I can in Trespasser?”
So Trespasser makes other 3D games seem stiff in comparison?
It does. That doesn’t mean those games aren’t fun; it’s just a different approach. In Tres-passer, a lot of the gameplay comes from situations like, “I need to get across this ravine, and there is a big boulder here, and my shotgun has no shots left in it, so I can use the gun to pry the rock out, and the rock will tumble into the ravine, make this big crashing noise, and I can jump across.”
So is this an “anything goes” game?
That’s the whole point. How much better is it when there are unlimited possibilities for cool stuff happening in the game than if we had to script every single thing?
Given the tie to the film series, would you say Trespasser has a cinematic edge?
We have tried to work that pretty hard. We have Richard Attenborough and Minnie Driver, and a lot of strong, polished cinematics. Plus, we have tried make a game that is explicitly about placing you in this situation, and making you deal with it. If we can get you to believe the environment is real by forcing you to manipulate it, become accustomed to it, and skilled in it, then, when there is an animal after you, it is genuinely scary. That changes everything, because you are thinking in terms of, “I don’t want to get killed!” And that’s huge! That’s gameplay! You are scared, but you win based on your wits.
What do you think gamers will get the most pleasure from?
Our puzzles. Our puzzles are free-form, and that sounds dangerous, because it implies they might be frustrating, but again, we have tested the hell out of them. The best moments will come when you solve a puzzle by thinking of a motion we didn’t see, or wasn’t obvious ac-cording to the visual clues we gave you. For example, we might give you a plank to get over a fence, but instead, you spring yourself up some clever way. You might set a crate up, shoot it with a gun onto the other end of the plank, and that flings you up over the wall. It can hap-pen! You will be so proud of yourself, strutting around and thinking, “I am the king of the world!” That is where I get my joy out of puzzle solving. That, and after I solve a puzzle, I want to be led into a new area. I want a new environment to explore, a new set of chal-lenges, and more stuff to do. Trespasser is all about that kind of payoff.
After the game is released, do you anticipate people will contact you regarding un-foreseen things that happened in the game?
They’d better! That’s the whole point of writing this engine. The engine is all about emergent properties, and people doing something we never thought of. I really hope players find amaz-ingly clever ways to circumvent our puzzles. I want someone to tell me, “I dragged a couple of planks from the beach, took the rudder of the airplane, used it as a lever to dig a big rock out, moved the rock and put one plank up against it, carried the other plank with me on top of that, and laid it out in front of me. Then, if I ran and jumped at the end, I could make it over to the monorail without having to open the gate. I’ll be like, “All right, man! That is amazingly cool, and more fun than the puzzle.”
At this point, what remains to be done?
All that’s left is bugs, and we are just grinding through it gruesomely. The publishers are go-ing in and doing a lot of tests, and we are starting to get that giddy excitement people feel when they have a really great product, and just want to get it out. Our girlfriends hate us, but we’re almost done.
That’s a nice segue into this last question. What do you plan on doing when Tres-passer is complete? It’s been a long haul.
Yeah! Getting really, really, deeply and irresponsibly drunk.
And washing down over two years of hard work!
If we have done our job and people get into the game, then it will be worth it all. I have said this before, but it bears repeating — what a great time to be a gamer. I plan on playing a lot of great games, then asking, “How can we make it even better?”