Transcription

You just go over the mountains over here and you're in high desert. You get past Ellensburg and it's high desert, so I spent a lot of time out there getting reference. Most of the reference came from Eastern Washington. I did a drive out to the Columbia Gorge and took a lot of photos and also a lot of just reference imagery of just like, hey, this is what desert looks like and big cliffs and stuff like that. That drive was two days. It was eight hours down, eight hours back, and I'm not even sure I spent the night. I did. I must have done. Spent the night down there. So while it was definitely made for the Southwest, the local reference was out in the Columbia Gorge. Good morning and welcome to the Black Mesa Transit System. This automated train is provided for the security and convenience of the Black Mesa Transit System. My name is Mike Harrington. I was a co-founder at Valve. I wasn't really used to big companies when I started at Microsoft. It wasn't that big. It was like 1,500 people, like a big high school. You know, nine years later, it was significantly bigger and it's like, oh, this place is too big. That was still like 1996. So they had a long ways to go. I told my managers like a year before, like, I'm leaving. I'm leaving in a year and I'm going to start a game company. And the first person I talked to was a good friend of mine, Michael Abrash. But at that time, he was trying to do something within Microsoft. And by the time it made sense, John Carmack had convinced him to go work at Id. So how could he say no to that? And then I was having lunch with Gabe and I said, Gabe, you know, I'm leaving. I'm just going to leave. And I'm going to start a game company. He goes, I want to leave. I'm going to start a game company. All right. And that was it. You know, on the surface, we should have failed. And realistically, both Mike and I thought we would get about a year into it, realize we'd made horrible mistakes and go back to our friends at Microsoft and ask for our jobs back. But we did think that we knew a fair bit about software development, that there were expertise that goes into it. I think we also had some pretty clear ideas of how to design a company. Right. So as we were building Half-Life, we were also designing Valve at the same time. We had no plan except for by then, like I mentioned, Michael had gone to Id and he said, oh, you're starting a game company. You have to use our engine. So Gabe and I flew down to Mesquite to meet Id. We spent a good deal of time with, you know, John Romero and he told us like, this is what you need to do to start a game company. You need to go out and hire some level designers and, you know, do that kind of thing. So we kind of understood the engineering part. And then, you know, because we were Michael's friends, you know, we walked away with the source code to Quake that day. We had no contract yet. We had kind of an idea of what it could be and they just gave us a CD and we had kind of the crown jewels of Id and that, you know, game on. I think most of us had no game development experience. We were either amateurs working in our homes at night for fun or we, some people were from the software world, but they had never worked on a game. I think we had three or four people in the entire company who had actually shipped a game when we started Half-Life. We have to figure out what is Valve's first game. So we have to come up with an IP. So everything is done from scratch. Gabe came into the office and he said, I read this story, I read this Stephen King story called The Mist. You know, Stephen King, it's got monsters, you know, people in a grocery store, fog comes in, monsters in the fog, you know, use your imagination. We didn't want to write that as a game. That wasn't really the point, but it was just the tension and, you know, how, how it felt. Well, it was interesting at first we had two games. Greg was working on Prospero and then Half-Life was originally called Quiver. My job was then build a team and figure out another game for us to build that'll be the second game Valve ships. What features should it have? How is it going to be different from a normal first person shooter? Every time the engine got a feature for Prospero, Half-Life stole it. And then Half-Life just started getting visually, you know, more impressive and cooler. And eventually it just, yeah, because it was going to ship sooner, it just stole the resources from Prospero and Prospero just faded away. So I had this character had a, like a floating orb and it had wanted to have some effects on it. It's kind of magical energy, something. And so, um, I worked with them on that and came up with the first version of the beam effects and, uh, you know, they were using it, they're, they're happy with it. And then people on Half-Life are coming over and like, what is that? Where did, how did you do that? So very quickly that, uh, got into Half-Life, the disaster sequence has kind of beam effects all over it. And we ended up using it for the, the Oregons. Like there are a bunch of things that we were doing that we thought would help us tackle the problem of what video games were and what video games were, were a rapidly changing field where at the time, 90% of video games failed to turn a profit. And a few percentage of them were, were very, very profitable, right? So it was a, it was a field that really valued talent. You know, like I said, Gabe and I brought some people and then, you know, from there on out, we had to hire people and, you know, Romero said he got a higher level designer. So, you know, we found some people who had industry experience, um, worked on Rise of the Triad first as a pixel artist and then went with Duke Nukem and then everything just went nuts. Well, I was sitting in my office one day and my phone rang just out of the blue and I pick it up and it's this recruiter. Eventually he hooked me up with Gabe and they flew me up here as a big surprise. But then we started looking, you know, the nice thing that we had is we had, um, a whole bunch of people in the world who had been working on mods and so we could start looking at those folks. And that's where we hired a lot of our designers. Steve and Guthrie, Steve Bond and John Guthrie were my favorite hires by far. Other than Doug, Gabe sent him an email and they're like, who is a scammer? You know, never heard of these guys. Just kids, like barely 20, I swear. And then they show up and they're just the most creative, brilliant, oh, fucking clever. Steve had the magic ability to make stuff fun. The fact that those guys had been hired to go work on some game for some company, I think that was the start for a lot of people that something cool was going to happen because those guys were already doing stuff that was cool. And then Mark showed up and like Mark actually knows how stories work, you know, beginning, middle, end and all this stuff. And he had all sorts of great ideas and he was, he also would just, you know, bring good ideas out of everybody else as well. Like the person who wrote more code on Half-Life was a chemistry major who decided to be an IP lawyer in Atlanta, right? And he ended up writing more code than anybody else. The creature designer for most of the AI for the creatures in the game was a manager at a Waffle House. It's not like there are a bunch of people that you could say, oh, this person, you know, has a PhD in video game design. It was that combination of both finding those people when it wasn't necessarily easy to find them at that stage in the video game industry, and then also convincing them that it was a better place to work than anybody else. One of the things that I definitely learned was how to hire engineers and, you know, we needed some engineers. For us anyway, we had a licensed IDS engine, you know, we're using the quick engine. And one of the things me and some of the other engineers, probably Ken we're talking about was if it would be possible to use the save system to kind of like join our levels together in a really seamless way. So you get to the end of the level, save stuff, and then load some of it on the next level and have it just feel like, you know, the world's all connected. And so we had like every once in a while, a hallway or the train would get to a point and you would have a little stop for the level to transition. And so he wrote that code leverage on top of just the save load code to do the transitions. And I think in Half-Life 1, we got most of the transitions were, you know, a second or two. I remember when we went from 8-bit to 16-bit, like, hey, we don't have to have all our textures fit in a certain, you know, 256 colors. That was a huge liberation. The two tech investments we made for Half-Life was to get 16-bit color and skeletal animation for all the monsters. But at the time, it was completely revolutionary. Your Quake characters were, it's like a text file of vertex positions over time. I wanted so much more animation and I wanted to, the animators to have the freedom to put in 100, 200, 500 frame of animation, if not more. And then you can do all sorts of other things because you have a skeleton, so it's like, oh, I can hook on a gun to a character. I can start to make changes of characters and replay the same animation or reuse most of the same models. So I can swap out heads, we can swap out this, we can swap out whatever. Like all the technology that went into the game to make it as fun as it is was not available one generation before. So licensing the Quake engine, which was, you know, very mature and had level designers who'd grown up just goofing around with it. I wasn't heavily into scripting since back into Doom. Like my old Doom levels have a lot of weird, odd events in them. I think I asked Ken, is it possible that I could, you know, have access to the character's animation, like a trigger? Like what I mentioned, I could make the guys go into an animation, I could make them run and then I would configure everything. It was all hack. This is total hacky stuff. And that is mostly what I built, was like things like characters reacting and jumping out or made to look like, or I'd push, I'd use a push brush to throw a headcrab out to make it look like it launched at something. Or if a loud sound happened, the sound would disrupt it, it would cower. Gabe's dictate was, as you walked forward, something needs to happen, like every three to five seconds. If the player stands still, the universe can be quiet. But if they move forward, if they're active, the universe needs to play back something, needs to do something. Even if it is a scripted sequence or a sound or some vignette or something. So I came up with this interaction method where you could do little AIs where the characters would talk and interact with you a little bit. And then we wanted to do spectacle. And that was scripted sequences where you could tell a character, you need to run here and then jump into this animation. And then have a bunch of you all synchronize your animation because it's somebody flying through the air, they're smashing something or they're doing something like that. And then test it and figure out like, oh, this is real easy to break. That was the roughest thing about scripted sequences, where you'd have like a headcrab jumping on a scientist or something. And it's like, yeah, it works totally fine 90% of the time, but some people just sprint right at the scientist and sometimes we have to put them behind glass. Sometimes getting the level designers making the player want to do the thing that gives them the most fun is a very subtle art. The tentacle sequence is probably the one I remember the most. You're playing it, right? Yeah. That was the first time I think they melded together. Because you had your trigger. So closely. And we wanted it to be triggered. So by the time you start to see it, you see the scientists flying through the air and hitting the wall. Then you could look in and you could see the scientist being carried away. The tentacle itself is separate. It's on its own. And you bring the two together, mesh them together and boom. It felt so seamless though. And that for me is like moments like that, and that might've been the first one, is when you start to see the scripted events in play start to blend a little bit more than just ambience. For God's sake, open the silo door. They're coming for us. It's our only way out. Oh my God, we're doomed. I remember Brett in particular had done some of the lab scenes where scientists would walk from one area to another and just look at a monitor. And they just, it gave them a little bit of life. A ton of the work that went into that area was because we knew it was going to be the beginning of the game and we really needed to make it a vibrant and interesting, fun place to be and explore. So originally they were designed to sort of be behind glass. They were just designed to be spectacle. And then all of the level designers were like, no, no, I'm going to put that right in front of you. And I'm like, oh God, how am I supposed to make this work? It's not supposed to do it. Get away from there, Freeman. I'm expecting an important message. The run speed for the player was so insane. It was like, what, 25 miles an hour or something? It was sprinting. You could get so far so fast. And so you'd either have to make the scripted sequence like so quick that you could barely see what's happening. Most people weren't looking in the right place anyways. You'd have to get their attention with a particle or sound or lighting or something. And then you wanted them to see what was happening. But when we playtested, and it's like a scientist was being pulled into a vent or something, people wouldn't just stop and like, oh. Sometimes they would be like, oh wow, look at that. They'd just run straight ahead and just start shooting. It's hard when you think back to it to recognize that those kind of interactive moments were so new. People just hadn't seen them. It was worth it, because no other games were doing that. No other games had that interactivity. No other games had that spectacle playing out in front of you. And it just added that whole feel of Half-Life. Gabe had started to develop this theory around interactivity. And the environment needed to let you do what you wanted and then react to it. So if you did something, if you shot your gun, there'd need to be a bullet hole in the wall. If you did a thing, other characters needed to react. The world is acknowledging you exist. The world is acknowledging all of your actions. And that's a huge part of reinforcement. And you'd have these conversations where you'd be sitting in a design review and somebody would say, that's not realistic. And you're like, OK. What does that have to do? And it seemed to me why that's interesting, because in the real world, I have to write up lists of stuff I have to go to the grocery store to buy. And I've never thought to myself that realism is fun. I go play games to have fun. And so we had to come up with some notion of what fun was. We knew it was an ad hoc definition, and it was the degrees to which the game recognizes and responded to the player's choices and actions, right? You know, in behavioral science, you would say we're explicitly talking about what were reinforcers and what the reinforcement schedules were, right? At that point in time, that was a useful way of making design decisions. The point I would make is, if I go up to a wall and shoot it, to me, it feels like the wall is ignoring me. I'm getting a narcissistic injury when the world is ignoring me. So to me, I was trying to convey to the user a sense of, yes, you were making choices. Yes, you were progressing, which meant the game had to acknowledge that back view. If you shoot at a wall, there have to be decals. If you kill a bunch of Marines, the Marines have to run away from you, right? You have to have this sense of the game acknowledging and responding to the choices and actions and progressions that you've made. Otherwise, it loses any sort of impact. Ted and I, we were on the same page in terms of what the game should look like, which would be just sort of alien in a very naturalistic kind of way. Like these are creatures from another dimension, another planet, that look like they evolved in a way. Chuck had a very, he came from Duke Nukem. He made a bunch of monsters that went in a sort of a different direction, and for a while, the game had a hard time coalescing around a single vision. All the ones that looked like genitals were Ted's, and all the ones with shiny bits that looked like alien soldiers were Chuck's. That was a thing when I first came in, Ted's style and mine, trying to figure that out, trying to find the harmony between the two to make it work. Ted had, he really liked the Wayne Barlow aesthetic, and I had a little bit different, but it worked great. Ted made these really weird, he did the headcrab. That's pretty much the symbol of it all, and I remember taking his headcrab and putting it on a scientist and making the zombie. We pretty much had, we talked, and it's like, okay, you do that, and Ted did the soldiers, I did the scientists, I did a lot of the aliens, and with the hornet gun, I did the hornet gun for that one as well, too. You know, the alien slave. Well, I did some animations for the bull squid, or bull chicken, I have to call him that. But that's a Ted creature there. The hound eye was Ted's as well. Yeah, a lot of the fleshy, if it's got a lot of flesh on it, maybe too much, that's a Ted-ism. You can always tell Ted's designs, because they're much more aquarium than, aquarium and or, you know, like, unspeakables, you know? Oh, Big Mama, I forgot about Big Mama. Oh, God. You know it's a testicle, right? It's like, eh. So disturbed. I concepted a guy in sort of a NASA space suit, a scientist guy, he was actually modeled after a couple of programmers I know, big, beefy, bearded Linux programmers, you know? And so that's where the HEV suit sort of emerged out of. There was a lot of desire for not having a space marine or a soldier be the guy. And then Chuck Jones went off and did a pretty different version of him that became known as Ivan the Space Biker. Oh, God. I remember telling myself, it's like, anytime you make a screenshot, make sure it's good, because you don't know where this is going to appear. We didn't know that. But Ivan is the prime example of that one. So you know, you're looking at it, it's, well, go look at Quake. It's like, OK, I'm looking at that, and he's got this incredibly boxy head. And you're always hearing, don't put too many polys in it. It's like, OK, so Ivan got the square head. And it was kind of an early, just a rough prototype of a Gordon Freeman baby. And I remember, we didn't know what Gordon looked like. So it's like, I was looking for somebody, and I found Mike Harrington. It's like, Mike, what does Gordon look like? He's like, I don't know. Put yourself in there. Oh, OK, OK. And so I went in and put kind of my likeness on the early Gordon model in the game. And it kind of stayed there. Did you have a ponytail? Well, I was ex-tattoo artist, so yeah, I definitely had that. I don't have anything to work with now. But yeah, at that point, I had a ponytail on. If you play the game, you know that it's Chuck. If you play multiplayer and you see the model, it's Chuck. But yeah, if you look on the back of that, there's this little, like, four triangles. And he's got a little ponytail back there. I did the G-Man. And again, that one was, Gabe was talking about the Cigarette Man from X-Files. So we needed our own equivalent of that. Because after a while, after playing with the game for a long time, people realized, like, hey, we made this character. And just in this one office scene in the beginning, you know, where he's behind the window and he's talking to a scientist or something. And he's like, he looks really creepy. It was late in the day that we thought, oh, OK, well, he's going to be the big bad behind all of this. Like, scatter him around in places that you can't get to him. But he's there mysteriously. There wasn't enough in the beginning of the game. It's like, we need another monster. Headcrabs weren't enough. It got kind of repetitious. I think I grabbed Ted and said, I need them to do this and this. And so he whipped it out in not much time at all. I'm going to get the timing just right. And I have no time to do this. I'm just going to bang it out. And then once it was working, everybody said, oh, just go back and sprinkle them back in the maps. And that was this. And we introduced them a little earlier than we were planning. Yeah, because they ended up being, they weren't real tough. They weren't supposed to be tough. But they arranged. But they arranged, and they were fun, and they had a big show. And it was showing off a bunch of stuff. I remember the assassin came in pretty late. That was it. It was like, OK, I have a week and a half. It was already done. It was already all modeled, already all animated. And the requirement was it couldn't have any new animations and any new anything. So I had to come up. I grabbed a bunch of Steve Bond's AI and banged that thing out. I built a level super crude, but it had all the right spacing and all the right heights and all the right geometry. Again, it's just a presentation where this creature gets to shine, this enemy gets to shine, and is your entire focus for that scene. And it's kind of a memorable experience because it just takes place in a different fashion than other encounters. And you get knocked out. And there's all kinds of, like, the lights go out, and you get dragged through a corridor and things. So again, it's something you remember in the game. And if we find the body, body, what body? We just did a lot of play testing and did a lot of level building and used all of the monsters and used up everything that we had developed and started seeing which ones kind of stuck out a little oddly. You know, like, we had the hound eyes that Ted had concepted, and Kelly had done great sounds for them. And so they seemed like a very oddly plausible sort of alien creature. And Chuck did what we called the panther eye, which was this big black panther looking creature that also just had a big red eye for his face and a big shark fin coming out the back. And design-wise, he seemed like a, oh, this is a very useful creature. You know, it's like when you design things, you know, you like to have a gradient of difficulty, a gradient of escalation and threat. And so if we had these hound eyes, then, oh, yes, this, like, panther, bigger, scarier looking thing would be more of a threat. It was in the game for a long time, and then sort of just worked its way out because folks couldn't find a way to incorporate a creature like that into the environment and make it seem reasonable. The stuka bat was, it was something that I had did, just a weird, another creature from the air. We had the chub toad that we were going to use as a feeding animal to help. And yeah, they had a mechanic to where they were going to use this to distract monsters so you could possibly get by. And it pretty much reached fruition to where we created the object. But then again, time became the factor, and it had to get cut. I grew up just outside DC, and so there's all these, like, big, really banal office buildings. And that's kind of the direction I went. So it started becoming a facility. And so I started making these kind of linoleum tiles and the drop ceiling, the concrete block wall, the black and white tile floor. She was, like, the texture artist for a long, like, for most of the game. It's just an overwhelming amount of work. Originally, I was hand-painting all of the textures. And you can really see a shift in some of them where they go from hand-painted to photo references. The photo references are much, much better. So I was all over Seattle, Harbor Island, Gasworks Park, getting rusty metal things. What can I get good pictures of that is vaguely industrial and interesting to look at? And then how can we use this? And so, you know, sides of box cars. And I don't know how many hallways in Half-Life are, you know, the side of a train car kind of thing. The Kirkland Costco, a lot of time in the Kirkland Costco, getting just packages of, I don't know why I needed, snacks or something. The first year was all about opportunity. And, you know, it wasn't quite as focused. We were trying, we were finding our footing. We were getting into the three months before supposedly shipping in 97. And it's like, this isn't jelling. This is really not good. This is like, you know, a quick knock-off, cash-grab stupid. And let's not do that. There was a lot of disconnect between what all of the different groups were doing, what engineering was doing, and what level design was doing, and what animation was doing. We had a bunch of monsters that had no plans to get in the game because nobody else was assigned to work on them. And we had a bunch of levels that, like, what's supposed to go in here? Oh, I don't know. Everybody's going to do a bunch of stuff. I'm like, no, that doesn't work that way. So here we were, all these people doing all these cool things. And everything's just this random collection of, oh, here's a cool moment and a cool this. And, you know, and then it's gone. And, you know, oh, now on to the next thing. Because the way we were building the game was every level designer was kind of his own silo, his own universe. And I think that just grew out of our kind of mod developer roots. But it wasn't cohesive. And it certainly didn't have a strategy for, like, maximizing any of that stuff. You must have been in the meeting when we reviewed all the levels. It was like the first time everybody piled into a room. And Gabe had his crazy Hellraiser chair with a beautiful Mac monitor. And for two days, he played everyone's levels. But I remember one moment, he turned around. He was like this with his hair. He's going, we're going to fail. We're going to fail. We're going to fail. And I was just like, it blew me away. I was like, oh, shit, this isn't good. You know, we were doing our best trying to make this game and the story. And, you know, we had a schedule from Sierra. And it was a tight schedule. And we told them, like, we're not going to ship this. And we realized that you're not going to pay us to continue developing this. But we're going to do it anyway. Late is just for a little while. Suck is forever, right? We could try to force this thing out the door. But that's not the company we want to be. That's not the people we want to be. That's not the relationship we want to have with our customers. And so we kind of, Kelly and I kind of sat down and sort of tried to take an inventory of all the cool stuff we had built. And why weren't we just, you know, building more on top of that? And, you know, those discussions kind of led to us building this kind of design cabal process to try to give some cohesion to the game and kind of changing the way we were approaching it. And so a cabal was a small group of people, a multidisciplinary group. You know, there was, like, an artist, you know, level designers, engineers, you know, everything in that small team that would write up a spec for, you know, a level. There was the arc that, you know, Mark had written. So we knew how that fit into the story. And we got a small group of people together. And we started with the beginning. And we went through, literally, what map do we need to build to tell the story we're going to have to tell? And then what does that connect to? And what's the next piece and the next piece? The strongest and most influential thing that came out of them was, to me, the sketches. And they were really the visual meeting notes, right? We came up with a formula for how that worked. We had a certain percentage of the time where you were fighting, a certain percentage of the time where you're exploring, and a certain percentage of the time where you're solving puzzles. And so we could apply that uniformly through the game. And that worked pretty well. Testing. Testing. Everything seems to be in order. Good morning, and welcome to the Black Mesa Transit System. This automated train is provided for the security and convenience of the Black Mesa Research Facility personnel. The whole opening of the game was specifically a reaction to how all the other first-person shooters were, whether they had cut scenes or not. Like, sometimes, like, they would just open without a cut scene. But, you know, you're standing in a room with a gun. There's a thing attacking you. So we could look at examples of all that. And I think there was expectation that that's what Half-Life had to do, because every other game was doing it. I can tell you the number of people who started up the game, were staring at it, and then accidentally bumped their mouse and realized, this is live. And as simple as that sounds, it was mind-blowing for so many people. Because it was too much going on. This isn't real time. This is a recording, you know? That was sort of the way we, you know, stumbled through a lot of the new narrative tricks. It was like, well, what can we do that, you could do this in a movie or a story, but nobody's doing this in first-person shooters. You're this anonymous scientist, and you're riding on this train into work, and rolled the credits at the beginning. It was kind of like a cinematic experience, almost. Like, I don't think games had really gone down that road. It's so funny to look at it now, because it's so primitive. But at the time, it felt like you were really in this Black Mesa world. Pre-disaster survived from the first year. So a lot of that geometry was from that first version of the game, but in a very minimal way, right? Just some, the basic style of some of the hallways, the curved corners in the hallways, things like that. They more became sort of like Lego pieces that John could use and stitch together and extrude. I know Brett did a lot of the raw geometry from pre-disaster, and he did a lot of that. The raw geometry from pre-disaster, and then John did most of the work to ship it. And then the post and the train ride was all John, I think. The sector C, so those are heavily influenced by NASA control rooms. With my favorite green hammer, I really influenced a lot of like the starting. We started with the kind of the consoles. I'm trying to remember, and then there was just lots of hallways, and people were getting lost. And I did the color stripes to help people follow them places. One of the things that was actually super exciting was how quotidian a lot of it really was. You know, the lunchroom with the microwave, you press the button and it would go splat. My God, what are you doing? Yeah, all this work so you can start a microwave and blow up some soup, which was a reference to real life, because that happens regularly, like we regularly accidentally blow up soup. Mostly, wasn't it? There were a few people who were just, you know, challenged with kitchen appliances. And, you know, at one point we had a company fire. We had a lot of things going on in that kitchen. All of that stuff had a real kind of eloquence that in the late 90s, in a world where everything was so over the top, you know, but we still got to go on flights of fancy. Welcome to the HEV Lock 4. I think like a super pivotal moment for me was when we came in this one morning and we saw the test chamber sequence that Kelly and John Guthrie and Wedge was in that office and Dario Driller, they were all in this corner, dark office where so much cool stuff happened. But so Kelly and John stayed up all night and did the test chamber sequence. And it's pretty much like the one we shipped. I mean, that was a lot of what went on in that room that we called the submarine. It was just improvisation. You know, just what can we do to solve this problem? How do we connect what's happening before to what's happening after? This needs to be somehow important, but we just don't know what to do. John and I talked about it. I went home that night and I sort of sketched out, like, you know, maybe we could do something cinematic. We'll take the camera away. We'll turn the lights off. We'll do sounds. We'll just throw everything we can at it. You know, we built that basically that day. You know, it was just a sort of a one day throw together. That's how fast we were moving. They had audio, right? Yeah, yeah. It's like the heartbeat and the breathing, all that stuff was in there. And I played through it and it was like, damn, I have not seen this in any medium in anything before I experienced something like this where this immersive experience of this actually happening like this. And at that point it's like, okay, I see what it is we're doing. And it's this kind of immersive, unbroken experience that happens to you, the player. And then that was this key principle about how, you know, we never take you out of that role. Personally, the difference between before the disaster sequence and after disaster sequence, I felt we really had, I had an understanding that this product was gonna be special or it was gonna be possible, it was gonna be tied together. Whereas before that, there was a lot of really good stuff and a lot of little things all disconnected. But we, well, you guys managed to tie two very different parts of the game together really well. And I thought, okay, now I feel like we're on track for finishing this product. To me, it shifted my brain to have a lot of confidence that this product's coming together rather than just generating a lot of content that exists independently of other content. One of the ways we were funding our development was we sold a preview copy of Half-Life 2, a video card company, I forget which one. And so they had in their contract that we had to deliver it to them by a certain date. And we got like the first three levels done for them. It was really a good milestone for us, just, you know, from working as a company because, hey, there was a deliverable, we got it done. And, you know, we shipped it off to them. And then I swear the next day it leaked. And I was just livid at first, which was stupid, but I was just like, oh, I can't believe you guys let this get out. Like, you know, this wasn't us, you know. And then it turned into something amazing where people started playing it and it was all over. And we started, you know, seeing it, you know, online, what people were thinking about it. And I forget which magazine who said, usually we don't review, you know, beta software, but they reviewed it and said, this is awesome. And that gave us a tremendous amount of confidence. You know, it was an outside validation of what we were trying to do, you know, that it was the right thing. I think just with first-person shooters already kind of, they were kind of like the vowels of weapons you put into first person. You got to have a shotgun and there's obviously a pistol. You know, you need your weapon that's going to make you struggle and drive you to get another weapon. That was a lesson in how to make all of your weapons as orthogonal as possible. So each one did something completely different. All the crazy weapons, those are my fault. So I did like the rocket launcher and the gauss gun and the snarks. I like snarks. Snarks always made me laugh. It's just always funny to watch people run away from the snarks. The squeak grenade, it's called the snark, I wanted something, a little creature that you could bring up and hold and you could throw it out and it would just start running. Because people were camping and everybody hated campers at that point. So you wanted the anti-camp weapon. But it was kind of, you could literally find a little hidey hole and just sit up there and just throw out 30 of them. And they're all over the map. I had a cat and you'd sit there, you'd do your finger blue. Yeah, you'd do your finger towards him and he'd get mad at you and he'd swat. So that was the inspiration for that particular animation. And then the other, that big alien creature with the, we needed some kind of alien weapon. Drew that one, built that one, animated that one, but that was also another one that whenever you would, you could get in a corner and just shoot them in the air because the hornets would track. I think that was the biggest thrill is brainstorming these with your crew, making them, building them, putting them in the game and then playing with them and then they're fun. So there's sort of a trope of having a melee weapon. I think it was Steve Bond and Ken Birdwell and I were saying we wanted a device where the world was reacting to you. It was really goes back to that theory of fun that we were just running around like idiots, smacking the wall. It's an odd thing to know in retrospect, but at the time it felt profoundly satisfying to be able to smack walls. And that was just an example of how that, that fairly abstract notion of what fun was ended up translating into a set of decisions that were really visceral. When you're going around whacking a wall, crowbar is an obvious thing to whack the wall with. If the sound could let you know what's going to happen, that's really makes it powerful. So we're always on the lookout for that in level design and monster design, communicating because the AI is often hidden from the player. So we use the sound to broadcast the internal state of the AI. You know, soldiers, they tell you I'm injured. I'm going to throw a grenade. I'm taking cover. You know, they are broadcasting as constantly as they can what's about to happen. So it was really important to try to keep those states distinct so that as you play the game, you're actually, whether you're aware of it or not, you're starting to be able to tell what's about to happen. We knew he was a musician too, so he could do sounds, but you know, he also decided he was going to write the sound engine. And then he wrote the soundtrack and he'd never written a soundtrack before. And he wrote the whole, you know, soundtrack for the game and won an award. And it was crazy. There just weren't that many of us. You know, I had a little bit of a background playing music. I understood the sound engine really well. So it was really easy for me to just hook sounds in. And I understood enough about the level design side that I could kind of connect those things a little bit. The sounds could come from anywhere. And then I just heavily modified them. Usually those are animal sounds. Reversing, changing the pitch, chopping them up and rearranging them. Head crabs were rats. The little squeaky, yeah, that's a rat slowed way, way down and then reversed. Yeah, the DSP stuff was another cool thing that Kelly built. When you got an event, it sounded like you were an event and that wasn't really happening, I guess, at least in Quake and other games I was playing back then. You know, we had these huge silos, these cavernous spaces. The day they had their own echoey DSP, it made an enormous difference, right? You believed the space. Kelly tells me, oh God, Chuck Jones asked him if the characters could speak. And I was like, yeah, they could. I asked him if the characters could speak and have their mouths move. And Kelly said, you know, that's super easy. I could just look at the audio and see if they're talking and do that. But getting their mouths to move is stupid, he says to me. And I look, Chuck Jones told me the same thing. I told him making the mouths move was super trivially easy, but knowing when they're talking and knowing how to follow the envelope of the speech would be impossible. And Kelly and I just both looked at each other and like, oh, and so I went back to my office and he went back to his office. And I think within an hour, the characters were just talking and their mouths were moving. And it was only because we sat there at lunch bitching about an animator who was 100% right. Why are you leaving me here? Scientist conversations, Kelly and I made that. When they're next to each other, one scientist will occasionally play a question and the other one will give him an answer. And it's not a script, you know? It just might be, it might not always make sense. I am rather looking forward to this analysis, aren't you? I don't think so. But it had a bunch of personality, it was really cool. And I worked on getting those guys to look at you and turn their heads. When we first added the ability for scientists to respond to the player and other things in the environment, what I would call them, what response rules. The first time that went in, it came flying down the elevator and one of them looks at you and goes, oh, hello, Gordon. And then they explode into a pile of jibs at the bottom. Hello, the automatic diagnostic and announcement system welcomes you to the Black Mesa Research Facility. Remember, privacy is everything. I hadn't really started working on the biggest part of the game, which I ended up shipping, which was surface tension. Prior to then, a lot of the work that I had done was the alien research labs, which turned into Questionable Ethics. And all of that was ditched. There were some concepts made it through, but everything was rebuilt from zero. But you know what did make it through? It was power up. Yeah, basically that geometry survived completely intact. It was so closely designed for that creature that because the creature survived, the levels survived along with it. Yeah, there's this sort of central circular thing where you have to rotate the train and you have to kill the gargantua first because he was always defending the way out. I mean, you have to go to another area which looks very much like it was designed for Quake because pretty much it was. There was a flooded pump room and you have to drain the generator and then you can switch the power on and you can take the train out. Yeah, I remember him being totally central to this whole thing. And then, well, how do we block the player, which is, I mean, that's pretty much like design point number one for all level design. Okay, well, how do we prevent the player from just walking straight out through the exit? And you design backwards from there. Oh! Oh! In the first year, we built a lot of world space, a lot of track, as we call it, and the programmers would try to fit the AI into the existing maps. What we did the second year was we built the AI in very constrained environments that showcased the AI. Let's make the ideal space for the AI and then give that to the level designers to incorporate those elements into their map so that it worked better and was more interesting. And that, Gargantua was an example of that. And the grunts especially. Yeah, we had the AI that Steve had been working on and he had designed all these interesting encounters with them just in his test map. And so, he had them up high and down low. And so, basically, we just got to re-leverage that. Okay, let's start with the player comes in low and there's a marine on the same level. And then the next one, they're up high and then they come running down. Because it was really neat to see them retreat. They'd try to run back up the stairs and try to flank you from the other side. Steve Bond, like I mentioned earlier, was this really young guy who had been making mods in Quake C, really the designer for so much of the AI. I mean, all the stuff that's good about the soldiers, the grunts you fight in Half-Life. Yeah, we had a little bit of a language that we could sort of use to say, this is a place they can climb. This is a place they can jump to. This is a good place to hide. These are good places to go in general and are interesting places. So, you know, the AI had a little bit of help. We were trying to build something that utilized, you know, the train technology that Jay was working on. And we thought, well, let's give the player something that they can control, but in a limited way. We couldn't do vehicles like you see in Half-Life 2. So we did this, the most constrained vehicle which was the train. Players didn't always bring the train with them. And so you had a bunch of these poor people walking down these tunnels. Was that why you electrified the rail? And yeah, and the rails, that was part of the feedback. Part of the feedback was, well, you can just bypass all this. So we made the rails electric. It's like, well, maybe that'll teach them to bring the train with them. But the basic idea was, you know, let's explore that and then also reveal some of the hidden deep parts of the base, you know, the abandoned areas where you might get a lot more of the Zen creatures. You know, we had designed some of the ones that didn't make it into the game to appear there, you know, bats that would dive bomb you and while you're on the train, yeah, Stuka bats. And then we also tried hard to not make it feel like it was the super linear thing. So we tried to give little branches. And so there were some challenges there. Well, there is a really good sketch that Kelly Bailey made, which basically charted surface tension. And it was pretty extensive. Like, it went through the cliff side and it went through areas where tanks were and like a lot of soldier holdouts. Yeah, it even went through the desert, right? So it was pretty extensive. And then we also tried to make it feel like it was a little bit more like a real-world thing. So we tried to make it feel like it was a real-world thing. And then we also tried to make it feel like it was a real-world thing. Yeah, it even went through the desert with the cactuses and the helicopters and then across a dam. It was pretty extensive. You know, when I saw that sketch, I was so motivated to just start building all of that stuff. So I ran ahead with a lot of level design, a lot of geometry. And it seems like a lot of it stuck in some way or another, which is really amazing. I did the, oh yeah, the dam, which was funny. I just did that in a day. Like, it was like about four or five hours. I did a Hoover Dam thing because I liked the Hoover Dam. I saw a concept of the cliff side. I think Harry Teasley built that render of the skybox. And I just thought, holy crap, this is perfect for Vertigo. Like, whatever we do here, we have to play on Vertigo and narrow little edges and places you would like to normally dodge a grenade, but you can't. And that's the moment. That's really the big part of that level is that reveal. The play is kind of, kind of supports it afterward. The helicopter comes in, you're fighting the helicopter, which is awesome. First of all, I had a bunch of textures. And then every time I made new ones, whoever was working on the new levels would be like, oh, fresh textures, I'm going to use those. Somebody making another level would start using them. And I was like, no, this is chaos. We need to restrict this. So I started naming the texture sets by the level that they were made for. I was trying to enforce some sort of like visual cohesion. And so that really ended up working. There was one level I did, it was the introduction of the sort of the bounce pads, the alien bounce pads in the real world. And because, you know, story-wise, like, okay, you're getting closer to hitting the alien world. So I was, well, let's get a little bit of alien world infection into the real world. So I had done it just as a visual thing. John Guthrie made it 10 times better. And that's actually where he found the photo of my daughter because I didn't put my daughter in Gordon's locker. I hid her in like this destroyed office someplace, you know, as a little Easter egg for myself. And John went and put my daughter's photo in Gordon's locker, which I think was a good decision. I still get questions about stuff that I spent five minutes on 25 years ago from people who've now thought about this for 25 years. And they're thinking that there's 25 years where the development went into this. I'm like, on some level, when you start talking about lore, I'm like, well, there's Black Mesa. That was just a matter, we needed a name for something that was evocative. In that entrance, when you get off the train and there's that lobby where there's a big map on it. And when we were making the map, I put a dot in New Mexico. And then that turned into, I think Mark saw that and he was like, oh, where is that? This is where we are. And then he named it Black Mesa. That is a favorite thing of mine, actually. If you come up with a name for something and it evokes something, then you don't have to write the whole story. It's because it's a conspiracy plot. Everybody knows more about it than you do. So you don't have to do, you don't have to answer those questions. You just keep raising questions and making them. And I think that's what I like about it. You just keep raising questions and make the mystery kind of thing. Like dialogue came in later. Also, a lot of the scripted sequences came in later and those helped to kind of sharpen the storytelling aspect of it because, I mean, setting expectations to get up to the surface and then like, oh no, they're not here to help us. The whole thing about diegetic writing, right? Like keeping the exposition in the game instead of outside the game. You know, I give a lot of credit to Mark. After play tests, we'd realize that maybe you need more of a dialogue cue about something. That's when we started to do like scientists who would literally pop up in a corner and go, go that way. There's a few of those that we stuck in right at the end when we'd like, well, we get one more session with Hal. He can say, go down here, you know, go around that corner. Well, so much for the government. Their idea of containment is to kill everyone associated with the project. Judging from your hazard suit, I'd say you were part of what went wrong. I knew Hal from San Francisco as an artist, cartoonist. He was also an animator and an actor in some science fiction movies. So when we got down to the wire, we did a bunch of voice casting and nothing was right. And I just kept thinking of Hal. I finally just called him and said, Hal, I want you to at least talk to people here so they can hear your voice. And I got him on a speaker phone and I had a bunch of people gathered in my office. And I'm like, okay, Hal, everybody's here. And he's like, well, what would you like me to say? And everybody is like, this is him. This is the guy. Why do we all have to wear these ridiculous ties? Another interesting voice thing was the G-Man. So we liked Mike Shapiro's work. So we knew we wanted to use him, but we brought him in to do Barney and also the G-Man. So we did a take on all the G-Man lines with kind of a straight voice. And I was thinking that, okay, we have something safe, but we're not super excited about it. And give us something else. And he does this crazy lizard voice. The border world, Zen, is in our control for the time being, thanks to you. Quite a nasty piece of work you managed over there. I am impressed. And we took it back and everybody just loved the crazy lizard voice. And so that's how it worked out. Did you do the Nihilanth? I did. Ladies and gentlemen, the Nihilanth. Just imagine his head floating in space as a baby. I went into Gabe's office and I demanded Gabe to either fire Ken or I quit. And Gabe just laughed at me. Correct response from Gabe. Yeah, I was like, oh. Until Bill got there, I was the oldest guy there. It was hard because, you know, I had to go home in the evenings, whereas the kids who just, the young guys who just moved out from Florida or wherever, they did not have to go home in the evenings. I'd go home, see my kids when they were in bed. I would come back down the hill to Valve. Nobody else I worked with had to get up to get their kids off to school in the morning, but I did. You're wired to just take full advantage of the situation. Where you're like, you put everything into it. So when we say we worked 18 hour days, that's no exaggeration. Yeah. Not at all. I can edit all day. I can make, I can design all day. I don't have to worry about anything. You know, two guys from Microsoft start a game studio with a bunch of incredibly capable, you know, with an incredible team. You're gonna push yourself harder than ever before. I don't usually crunch real hard. I get it done in, like, I come in, I work, I go home. And a lot of people were living, like, you know, really long hours. You know, the classic, well, I'll come in at one and then goof around for a while and then work late into the morning, you know, late at night. I was employee 17. There was a woman that was the kind of the office manager and eventually Mona Lisa Guthrie came in at the desk. I was the only woman on the team. I was not awesome. I decided to become a baseball fan here after the birth of my first daughter. So she's the one in Gordon's locker, Isabel, but she was, she's very special needs. She's 25 today, you know, cognitively she's about three years old. So after her birth in the middle of Half-Life production, in the middle of Half-Life production, it was really hard. It was really hard. Cause we were getting about two hours of sleep at a stretch cause she had to be fed through a tube in her stomach every two hours. I had just a difficult time relating to the world when she was a baby. I had a hard time just talking to people cause I was just kind of tired and angry all the time. Yeah, my wife, she worked at Valve for a while. She did a bunch of textures for the game. Like I say, she did those soda machines. She did just a bunch of level textures. She textured a big mama, the sort of boss creature, she textured him. But when Isabel arrived, she obviously had a lot of issues. So yeah, I just had to stop working and stay with her at home. And it was the last year of Half-Life and we pretty much crunched the whole year. So I was at work a lot. It was hard. It was definitely a hard year. Games have often had this thing where they're just a little too attached to Star Wars, Blade Runner, Aliens. So it wasn't that hard to try to get people to go, let's think of something really alien. Like what Xen was initially, was you're gonna go inside a huge alien organism and kill it or figure out how to turn it on or off or something like that. I didn't know we couldn't actually do that or what was involved. What does it mean to have alien architecture or a planet that's biologically defining the structures in it? As opposed to what the tools were good at, which is making rectangles. And then slowly it scaled back until it was more and more corridors and that everybody was really good at. People were finishing up stuff and I was still working on Xen because it was like it went through a lot of iteration or at least we're trying to figure out what it was. Different people were taking swings at it either in static concepts. Or a bit of a timeline crunch for it as well. Yeah, it was the end of the game. I wanted to cut stuff out and they were like, no, no, leave it in. I'm like, yeah, but I don't have any, there's nothing, I don't have any support for this. And then after a lot of it was built, it was like, oh yeah, we're gonna change the gravity to give you a jump pack. That came later. And I was like, ah. There's a lot of stuff, especially in Xen, where you're getting close to a first draft of some things, right? Because we just didn't have time and we got to the end of the game. And so we considered not going to Xen, but the art concepts people did were so cool. For the Xen textures, I was deeply inspired by electron microscope imagery. And we'd had so much of that regular structural texture material that I wanted to do something more organic. Insects as well. I think there's a lot of textures that were based on beetles. It was really fun and incredibly challenging levels to make because the editor was not friendly for organic shapes. I built the little behaviors for the tree that hits you if you get too close and the one, the light stock that kind of hides. But Ken did a bunch of AIs there, that the floating controller guys. And the flying guys were great, but you had to set up air nodes for them to travel properly. I prototyped a lot. Some of those prototypes were used in the final battle, I believe. And also for the gonad, which I think Randy did. We, at some point, have to stop polishing. At some point we have to like, it is what it is. And if the player hasn't had fun up until that point, well, then we failed. They're not going to get to that point, so who cares? It just has to finish, and it did. Yeah, and there's always Half-Life 2 to win them back. I love that sensation of being alone in an alien dimension. But the experience that people really liked in the game was scientists pop up and Barney pops up and you have companions and you have this feeling of companionship around you with all the other stuff that's going on. So a lot of people missed that, which I understood. And I think if we'd had thought about the ending from the beginning, we would have figured out some ways to do that a little more. I remember coming back from downtown Cabo very drunk and then running to the beach, but didn't know there was like a six-foot retaining wall there so I just went right over. And then I was hurting so much, I slept the night where I fell. Yeah, and I woke up the next morning with a really stiff neck. Oh, my God. It's a good thing you were pretty young. I remember after we'd gone final on the game, I was home one morning, you know, taking a shower and my wife asked me, you know, like, it's done, is it any good? Is the game any fun? And I go, I don't know, like, I hope so, you know. I don't think it's exaggerating by saying if any one of the crew had disappeared for a month, we wouldn't have shipped. You know, it didn't matter who they were. You know, everybody was mission critical. My inspiration from Valve, what I really took away with me was that the thing that is most important to me is what are we making? Coming in as a texture artist, you don't have a lot of expectation that you're going to get to influence that. And I was able to have quite a bit of influence on the what. It was so fun to work with the people who were making Half-Life. The collection of people were diverse, at least like in their professional backgrounds was kind of all over the place. And so figuring out how to work with all those people, but also just the fact that we had hired really stellar people who were good at collaborating, like the conscientious people, deeply passionate about what they wanted to create. That just was a recipe for a professionally, like really good time. It seems to me that one of the things that good games do is make two games worth of stuff and then throw away the bad game. And fortunately, like Valve was able to do that. A lot of companies couldn't do that, but backstopped by Gabe and Mike, you know, Valve could afford to do that. I always have absolutely nothing interesting to say when people say, would you reflect on your legacy? Like, I really don't look back a whole lot. I'm always excited by the future, right? So to me, it's like, I can look back at the things we did, but to me, they're just sort of like the stepping stones to what we're gonna be able to do in the future. That's just how I'm wired, right? Anyways, game on. сть Let's do this. Two, three, GO. Defense. functioning execution complete Defense. Using enemy. Reading. Impact. Shooting. Cross,ifire. Now, focus on the targets. Right, target in sight. Target in sight. We are going to throttle at war. Lovely. Let's go. OK, go you satele. I see the SACTO. Oh, no... LASER ACONECTOMY LASER ACONECTOMY VEYEL ЛАСЕР АКОМПЛЕМЕНТ