- [General] Welcome back, commander. ♪ Hit it ♪ - [Soldier] Sir? - [Soldier] Clear. - [Soldier] Orders? - [Soldier] Moving out. - [Soldier] Yes. - [Soldier] Yeah? - [Soldier] Sir? - [Soldier] What? - [Soldier] Yes, sir? - [Soldier] Medic. - [Soldier] Move, move, move. - [Soldier] They're everywhere. - [General] Incoming transmission. - Hi, Louis Castle. I'm the co-founder of Westwood Studios. I'm here to talk about Tiberian Sun, which was built by a team of people that had built many RTSs in a row, and it really shows in the way they innovated and pushed the edges. [upbeat electronic music] - [General] New construction options. - Command and Conquer was really the second game that we had done in the real time strategy genre. At Westwood, the first game of real time strategies that looked like a real time strategy game would be Dune 2, and then, the guys Blizzard did a game called WarCraft, and then, we did Command and Conquer. So, Command and Conquer was, effectively, the third RTS game in the world, and the first one in the Command and Conquer series. Command and Conquer was conceived of as a trilogy. We were going to start with Tiberian Dawn, then go to Tiberian Sun, and then, ultimately, Tiberian Twilight. We started off Command and Conquer One, with this idea that, as a kid in your bedroom with a computer and a modem, you were hacking into or becoming a remote commander of a battlefield. And so, we wanted to deliver that as authentically as possible. So, the idea that the screen would be full screen audio/video was kind of a necessity, because it was supposed to be tying into a real feed, with a real person, somewhere. - Are you picking this us? - With Tiberian Sun, we wanted to fast forward the world. So, we wanted to advance that timeline several decades, so we could do things like standing mechs that would walk around, and stuff. Do more science fiction. And then, of course, the world had also gone through a big transformation, where Tiberian had infected most of the world. You had the GDI areas, or safe zones, and you had the infected wastes and mutants, and all sorts of other things we could bring into the game. So, it really added a lot of texture to the story, and added a lot of complexity to the kinds of things we could do with the characters. Command and Conquer was this idea of you're bring constantly pressed between having to make decisions to build up your economy, while you're waging war. Oftentimes, you're doing it while somebody else is doing the same thing on the other side. So, it's this tension of preparing for battle, executing battles, and having to manage both the macro and the micro at the same time. Blizzard games, the StarCraft games, or WarCraft games, Age of Empires, all of those games lend themselves all the way back to the original Dune 2 or Command and Conquer. There's all these games that people play today, like League of Legends, that simply wouldn't have existed if it wasn't for the talented team that thought, "What a crazy idea, let's make something really hard to do, "and make you put under pressured time to do it." And, that's where RTS got its roots. We always were looking at what was happening around the industry, and saying, "Okay, that's an interesting innovation, we like that one. "We don't like that one." One that was sort of a big problem was unit caps. We didn't want to have any unit caps. - [General] Unit ready. - [Louis] We liked the idea that you could build tons of units and go marching across the world and just mow everything down. That was lots of fun. - [Soldier] They're everywhere. - So, we were thinking about the game series as, "Hey, we'll look around the world. "We'll look and see what other people are doing," and every time we saw some great innovation, we said, "Does that make sense for our series, or not?" And, we were constantly editing, what's in, what's out. The way Westwood generally made products was, we started off with the high ideals of what we were trying to accomplish. This game is going to adopt these things, and so, for Tiberian Sun, it was destructible terrain, it was real-time lighting. It was the sort of fantastic, futuristic units. There was about five or six of these things that were just really core tenants of what the game was going to be about. The single player experiences were ways of understanding the game, so by the time you finished the single player experience, each one of the units that we unlock is explored in a way that gives you an idea how it functions, how you should use it, and how you shouldn't use it. But, it also gives us an opportunity, from a story point of view, to put some meaning behind it, and I think that's something that we were particularly good at, at Westwood. People cared about their wolverines, because they had the mission that made you care about the wolverine, as a unit, as a thing. So later, when you're playing multiplayer, and you're throwing 50 of those at somebody, in the back of your head, that's still means something to you. The unit was more than just a bundle of stats. I kind of miss the fact that people tend to don't spend as much time anymore, around the actual universe, and the characters, and the stories, they go, "Oh well, it's just a shooter game, "it doesn't really matter." For the player, I think it's really important to know your relationship with the game. It's not just about moving a mouse a keyboard around, it's really about getting into the units and what they're there for, and what their goals are. C&C, in general, always had a lot of technical problems, that maybe were not obvious to people. We'll start with the first one, which is big bucket called the CD-ROM. - [Soldier] On my way. I can make it. [cries out] - [General] Unit lost. - So, what really distinguished Command and Conquer from a lot of other games at the time, was the use of the CD-ROM. CD-ROMs had just come out recently, they were single speed, and they were thought of as a big storage device, where you could load a lot of stuff, but the speed of a CD is really, really slow. So, there were only a few games that could figure out how to use this in a way that was effective, and we used it so much that we ended up shipping the game with two CDs. That changed everything about how we thought about Command and Conquer. We went from 1-MB floppy to 700-MB CDs. We had two 700-MB CDs to store stuff on. We actually consumed all of it, virtually, because we decided, with Command and Conquer, we were going to do full-screen video and audio. And, this was quite a challenge at the time. When you see Kane, Kane was not just the star villain of the series, he was actually our director, and our producer of all of our video shoots. So, we had this great talent in Joe. With Tiberian Sun, we really wanted to step up our game. We wanted A-list, double-A, triple-A, so Michael Biehn, and James Earl Jones, I mean, these are top-notch actors, at the heyday of their careers. - Let's kick some ass. - So, our compression technology for all the video that we did for the games, for cutscenes and such, had to be such that it would be able to be playable from a single-speed CD-ROM. There were no third party codecs, even by the time Tiberian Sun, they couldn't do that. So, we had to create all of the video compression technology ourselves. We enlisted the aid of some people from a university. I think it was University of Washington. They worked on a thing called vector quantization. We would take the video and slice it up into little blocks, and the individual blocks would be quantized into a single scalar value, and they would be put into a multidimensional cloud, and reduced down to the one set of blocks that could represent the most number of other blocks reasonably well. And so, for Tiberian Sun, we were able to get a much higher quality visual with our technology than was available for the same kind of data rates as other things. There was certainly better codecs, as far as in total quality, but nothing that could run at the same speed as ours did. So, CD-ROM ended up being both a huge benefit, because we could do these great sequences, and also an incredible, torturous pain as a thing to use, because there was this massive amount of data, for the time, and this little, tiny straw that you had to sip everything through. So, I guess the second big bucket, I mentioned CD was one of them. The other one was just the idea that we had these realistic units. You have this imaginary space that's represented by mathematics, and you need to move an object from one place to another. The simplest way, like if you've ever wanted to always solve a maze, you could always go to the right, until you can't go to the right, and eventually you'll find your way out of the maze. And, if I said, "Take this unit, to go from here to here," and it went around the edges of the map all the time to get there, that would look ridiculous. You also can't just do a beeline, because if you try to do a beeline, something might be in the way, and you'll have to go around it. So, we had to figure out how to do that. Because we had lots of choke points in the game, and because you could just deform the terrain, you couldn't precalculate all the pathfinding, so all of the stuff was done in real time. So, something like that doesn't seem like a big problem, but just the sheer computational need to calculate hundreds of units pathfinding is actually one of the things that brought all the CPUs down to their knees. The computational need to calculate all these possible branches, all these possible locations, becomes too many computations, too many tables, and it brings the processor down. Even to this day, pathfinding is still a problem in every single game. In RTSs, it just happens to be a really hard problem, because you have hundreds of units, especially in C&C, and they all move in different kinds of ways, and the terrain that they're moving through can be very dynamic. And so, it's changing all the time. Most pathfinding solutions that are fast rely on the fact that the things around them are not moving, and so, you can do pretty good pathfinding, node-based stuff. When a lot of stuff is moving, and virtually everything is moving, it becomes really difficult. I think one of the other things that people underestimate is how hard you have to work to make a game not do something stupid. The best part about artificial intelligence, if you want to call it that, is you just want to avoid artificial idiocy. Pathfinding algorithms have this weird edge cases, where something will thrash back and forth, or it will go up to an edge, and it's clear that it should go one way, but it goes all the way around another way. Those kinds of edge cases that do things that are really illogical ruin the game experience for a player, because you get mad at the unit, you go, "What is this guy, just completely stupid? "He's going the wrong way." The player doesn't know how much work you had to do to make it work, they just know what comes out. So, I always tell people, it's like, "Well, if you spend more time "just making sure it doesn't do something stupid, "it'll actually look pretty smart." I think, with Tiberian Sun, we really complicated some of the simple stuff in RTSs. Talk a little bit about pathfinding, and that idea of finding your unit from one place to another, we make it really hard when the unit can suddenly burrow underground, and one can't. Or, there's going to be a wall that pops up in increments. So, when I go to test my path, the wall is there, and then, it's gone, and then it's there, and then, it's gone. So, you have to know, over time, whether or not you're driving something you can go through. All these problems become really complicated. So, there's all these algorithms out there, people can do it. You can implement it. You can have one unit go across the map and it looks great. When you start putting hundreds of units across the map, and people fighting each other and trying to exploit the system, it becomes really problematic, and it takes a ton of CPU time. - [Soldier] I can fix that. - [General] Bridge repaired. - We start with a pathfinding saying any friendly unit that's moving with my guys that isn't stationary, assume those are not on the map, and then do your pathfinding. And, that actually works surprisingly well, because, if they're moving, they're probably moving because you grabbed a bunch of units and told them to all go one place, so they're probably all doing that, anyways. And then, the second tier to that was, he said, "Okay, once we do that, "if run into something you can't move, "because there's a stationary unit that's a friendly unit, "wiggle that unit a little bit." Make him move in one direction or another a little bit, to see if it unlocks, and there's sort of a series of decisions after that, you sort of wiped out the edge cases, one at a time. Like, if there's a bunch of units, and you try to jam them over a bridge, they would keep shuffling, and doing this kind of weird little dance, and they would eventually get all the units over the bridge. Each one of the layers that they added, starting with not considering friendly units that are moving, when you're doing you algorithm for pathfinding, so it's quicker, and then, ultimately, moving units around to try to get them through openings, and then, finally, if you just have to punch, find another way around the map. Those types of layer, after layer, after layer, of innovations finally got the game to where, even with hundreds of units, and with playing against another player with lots of different things happening, the terrain changing, most of the time, the units did what you thought was fairly reasonably and logical, and so it became an acceptable thing. So, the funny part about it is, it's not trying to write a perfect pathfinding algorithm, it's about trying to write one that isn't stupid. - [Soldier] These critters don't look too friendly. Medic. [cries out] - [General] Unit lost. - At the end of Tiberian Sun and shipping that one, it was a lot of working hard, and there were a lot of folks at Westwood that wanted to try some other things. And, now that we were part of Electronic Arts, we had a lot of freedom to do other kinds of games. We were focused on a first-person shooter called Renegade, which is in the Command and Conquer universe. A really fun game, by the way. And, we were doing PlayStation 2 stuff, and a lot of other things. So, most of the team that had built Command and Conquer were off doing other kinds of games, and stretching out into other genres. I guess what was really endearing about Tiberian Sun, for me, personally, was the team had taken a lot of risks and done a lot of new technology. It was really neat to see so many new innovations, so much new thinking, get into a game, and get out there. Even by that time, there was so many RTS Me-Too's out there, that so many games just copied everybody else's stuff. It was neat to see the team innovate in lots of ways that ultimately made it into other products, as well. For me, it's always been very charming. It was definitely the last C&C made in Westwood Las Vegas, and it was done in a way that was a really a labor of love, do the best job with the franchise that we had built. Tib' Sun really reached beyond what anybody would have thought was rational at the time, and tried to do some really innovative things. From all the RTSs that we built, I've often said, my favorite one, Red Alert 2: Yuri's Revenge. Many of the things that that game did were facilitated by the innovations that were built by Tiberian Sun. All the ideas about the veterancy of the units, the very tech itself, and just some of the thought processes of design, and how we would make these very different kinds of units interact with each other. These things were played out in full, when a game could start with that baseline and build up, instead of having to innovate everything. So, for me, it's always going to have a very special place in my heart. It sold very well, and that's always important. And, the quality of the people I worked with in the game, they just made it very special. - [General] Objective complete.