Half-Life 2 - Valve speaks to Eurogamer!

People are pretty enthused right now - around March 15th we hit our alpha date and we took a few days off of production for everyone to play the game from start to finish and it was fun for the different teams to get a chance to see... because we have the different cabal groups, different design groups. Y'know, some of the guys are designing the city levels, some of the guys are designing the outer levels and whatnot and it gave everybody a chance to play through each other's work and actually see the game from start to finish. And it really injected a good amount of energy into the folks after what has been a pretty rough year for us, and to sort of say 'ok, there it all is, y'know, let's clean it all up and ship this thing'.

Once it's all put together it's somewhat tangible and you can say 'here's the good parts, I had a lot of fun here; and we engage in a lot of dialogue and you start pushing from 'what's the game going to be like?' to 'what's the path to completion with this much together?'

Well, we're hoping to be Beta before the show, before E3, and then after that it'll be at final release candidate stage, so we're targeting to be done for the first half of this summer.

We don't have a release date today, but we hope to have one to announce at E3. Like I say, we hope to be in Beta by E3 and if everything comes together we hope to map out a release date that's a little bit later on in the summer.

Sheesh, well to rewind, we came out of last summer aggressively still trying to hit that end of summer last year date [September 30th], we realised that we weren't going to make that, and we announced that we were going to be a little further out, probably in the winter time or whatever, but two or three weeks after that our source code was on the Internet and several dark weeks ensued. To put it simply it was a morale hit that you only can experience, you can't really describe, and to the team's credit folks rallied. One by one people put their heads back together and said 'we're either going to be beat by this thing or we're going to beat it', and in a way it caused the team to be reenergised because not only did we have something to prove in the sequel but now we had something to prove again in that we could get beyond this, that we could deliver something in the face of this.

No, no, I mean we weren't done. The September 30th date was aggressive, and like I said, before the Source code theft we said we weren't going to hit that date and then obviously the source code theft caused further delay so we had to go back and do certain things to look at the code and make sure we weren't entirely compromised in terms of network security, and that it wouldn't just be a haven for cheaters, et cetera. So there was a fair amount of work that had to go back into the code.

In terms of the content, though, I'm sure some people thought 'should we change the characters, should we change the story, should we the levels?' and at the end of the day we thought 'no, we're onto something good here'. We've been on this path for several years, let's keep going. Gamers understand that this isn't the final product. Luckily for us we weren't that close to being done so it wasn't like it was ready to be revealed. If anyone took a look at what was released at that point it clearly wasn't consumer ready yet.

Yeah! Well we were pushing for it, y'know, pushing for it, and pushing for it. Hindsight's 20/20. Should we have said that we weren't going to hit it earlier? Yeah, probably.

This year at E3 we're going to have new demos to show. We'll be showing at ATi's booth as well as in VU's booth and it'll be two separate demos, both of them will be in the theatre, same setting, sort of presentation style that we did last year. Last year we spent a fair amount of time going through the Source engine in demoing the rendering, the physics, the character stuff and talking a lot about that, and then we showed some glimpses of the game. This year, since we're closer to completion we're just going to show the game, and what we're really going to do is focus in on a lot of the gameplay pieces and take people a little further into the story, but we're really trying to keep a lid on the story so that it's fresh and the gamers can discover it for themselves.

No. Same venue and format as we did last year where it'll be in a theatre, it'll be in-game footage but it'll be run for folks to view while we do [the presentation]. It's a certain philosophy that Valve has, when we demo to audiences like that it's just easier to show them [to] more people. The set demo pieces rather than having somebody playing the game and somebody may or may not be pretty good at it or struggling and a bunch of people are watching it while somebody's colliding with the wall or something. It's just not the best way to show the game in a big group like that.

Ah jeez, y'know, particular parts of the game... It's hard when you work on a game to have a favourite. As you go through, and as pieces come online that are currently your favourite because you can see how this piece is close to done and what have you. I think, given what's happened in the past year, the thing I'm most proud of is the team's resolve to come back and do a great job to put it together. I think right now, a lot of the folks are really having a lot of fun with the intro to the game, because we're doing some similar things that we did in first game where you're on a train ride coming into City 17 where the game opens up and we're sort of indulging ourselves in the dark humour that was found in the first game, and inserting those little pieces and putting in little things that will hopefully sort of scare people in places where you think you're okay. Like somebody jumps out and goes 'boo!' - that kind of stuff.

We're taking the time now to put in those little details now. An analogy is in the first game after the experiment that went wrong. You were sort of going back through Black Mesa, things were sort of blown apart and whatnot. You were kind of safe and you got to that one elevator shaft and you hit the button and the scientist came down: 'Arrgggggh!', screaming down, plunging to their death, and that was sort of like one of those fun... it was kinda scary, it was kinda funny, and it was kinda dark and those little moments, I think, are some of the things that I think people,- when they think of Half Life - that's sort of in their mind. Y'know, there's the cool AI of fighting the military guys in the first game and stuff like that, but I think those moments are part of what makes it stand out.

Oh Jeez...(sighs) I don't know if we have a set idea on that right now. The first one shipped with no expansion packs planned whatsoever, and then we sort of said 'hey, people like this game, maybe we should make an expansion pack' so we hooked up with the Gearbox guys and did Op Force. That did really really well and then we were like 'let's do another one, let's do Blue Shift'. So at this point, y'know, it's hard to say. We built Source, though, with the intention that we were going to build several games on it. So with Half Life games and our other properties as well. So, we're definitely going to continue to work in the Half Life universe, we definitely want to expand some of the stories of Al and some of the other characters that we're introducing in Half Life 2, and how those present themselves, whether they're expansion packs or sequels is certain TBD at this point.

You know I think right now the team really wants to do it themselves - at least initially. We had a really good experience with the Gearbox guys and they did a really good job, but we experimented and interviewed and went down a lot of roads with other people before we made the decision of Gearbox, y'know, and those guys really got it. I think this time around folk are really really invested in the universe this time, and since we do have the engine that's going to take us hopefully through a few iterations we don't have to sort of roll the team off to get working on the new engine, and the big idea for the next one they sort of set the tape a little bit more for ourselves to get. So I'd say at least the first couple of things that you see will most likely be from Valve... Beyond that, who knows?

Everyone went down to Mexico and just got absolutely smashed, and then some sun - which you don't get to see a whole lot of in Seattle - and then we came back and basically in January 1999 is really when Half Life 2 got going. We shipped around mid-November '98, y'know people went out, we all partied together, and everybody went home for the holidays, and then after the New Year people got started, and I think you'll see probably a similar pattern this time. We're not going to buff into the holidays this time - at least not on the current schedule - but we're definitely going to take off and go and have a holiday together and celebrate, and then let everybody else go and spend some time with their families and then after that we'll sit and see what's up next.

Valve's really neat that way in terms of what we do next and who works on that - it's really up to the employees to decide. Gabe created a really great company that has a lot of freedom that way, that we fund our own stuff and we drive our own schedule, and so that's really good. And I think that's why so many people from the original Half-Life 1 team are still working on Half-Life 2. I think only a handful of people have moved on.

The Half-Life stuff is with Vivendi.

The first game featured some AI that we spent a lot of time investing and creating. That was one of the biggest pieces of new stuff we put into the Quake technology when we were making Half Life 1, what we really wanted to do was to have characters that responded to your actions rather than jump out and fight, that kind of stuff, and so what we've really done is just try to expand on that philosophy so that the characters are aware of your actions and they react to the things that you do.

We've also incorporated the physics into the game and one of the things that we wanted to do was to give the AI the ability to manipulate the physics as well, so it's pretty interesting at certain points in the game, y'know, you can grab a barrel and throw it at somebody and then they shoot it back at you and then you may throw it back at them and then they shoot it back at you, and that's all the AI at work And it's really a lot of fun when you see it working, because it like, this is cool - this is almost like playing a human on the other side of a multiplayer experience, or something like that, and they're catching what you're doing whereas in the old days you take a barrel and throw it at somebody and it bounces off their hand and they just keep shooting at you, so we're trying to do things more dynamically like that. It's the same philosophy of the AI collaborating with the player. Whatever the player chooses to do, whatever path they select, the AI reacts to that rather than trying to push you into the canned script of the area it wants you to.

Really the key there is GPU first, RAM second, CPU third. So, a Direct X 9 level video card - we're finding the best on the 9800XT, and is what we're showing the game on and what will be showing the game on at E3 - but any Direct X 9 level video card, so in ATi's world that's a 9600 or better. In terms of RAM, the min sys spec I believe is a 128 RAM, but once you get to 256 or so, you're gonna be good. Anything beyond that is just sort of gravy on top of that. And then on the processor side, the min sys spec is a 1GHz, but to really sort of have full functionality and all the bells and whistles and voodoo whatnot, probably a 1.4 or a 2.0.

Yep - you get a copy of the game.

We've offered content to run as low as DX7, but on the high end we haven't changed anything, and in fact we hope that when DX10 is introduced, as time marches forward with new hardware and new APIs, we've built the engine in such a way that it can move forward with that, so that there will be new features that we'll be able to sort of switch on with updates that we have.

But really the only changes have been to migrate down to the lower end to be able to support a wider audience, which is something we did with the original Half Life, which ran itself on like a 133 or a 166 or something like that, and that was really critical in terms of us being successful commercially, so we sort of looked back and said 'ok, how far back do we have to go to really hit everybody who's buying games right now, and we've run surveys that are on Steampowered - we ran one very recently - and it sort of showed us that the DX7 level was really where you start to see... Once you get below that you there's really not a lot of people playing beyond that, so I think that's like a GeForce 2 or a Rage Pro-level card I believe, or the Intel 740, in terms of the integrated chip.

I think it's both! We're gonna raise some new ones and leave a few things open. The G-Man... It's almost critical for the game for the G-Man to always have a bit of mystery about him.

Well yeah, there's gotta be that intrigue to keep people going right, but we also want to answer some questions. You definitely will find out more about what happened at Black Mesa and why this is going on and what have you, right? So there's a little bit of that reward and also a little bit of dangling of the carrot to keep people going, we want you to keep moving forward and find out more.

That'd be giving away too much [laughs]... I'll leave you guys to spoil that after you get your review copies!

It should take you about the same amount of time that it took you to play the first one depending on how your play style is. We're finding because there's so much physics, puzzles and interaction going on, that people are spending a lot more time just sort of doodling with stuff, you know. Once people get in there and they grab a barrel filled with oil or whatever and they throw it and it explodes and stuff, they're like 'oh wow, this is a little game in there by itself', so you find people spending an hour in a room that we thought was just a few minutes' worth of gameplay, so we're seeing some people really stretch it out if they're stopping and playing with everything. But our goal was to have, y'know, the same amount of time that it took you to play the first one, and putting an hour figure on that is so hard because some people just play it like whiplash style and some people play very methodically, so it's better to use that comparison.

Well we divided up into chapters, is how we think about it, so there's 13 unique chapters or episodes if you will, that we sort of move through. We sort of did that in the first game where we had Unforeseen Consequences, Anomalous Materials and all those other bits, and they were sort of... they weren't as concretely divided up where it was sort of like the chapter ended and then you'd start a new chapter and you're kind of still in the same area sometimes, and sometimes not, like sometimes you're in one of the Xen chapters and it's a completely different. In this one they're a little bit more complete though. That's sort of how we divided it up and that's also how we divided up the design tasks - this group's working on the first chapter, this group's working on that chapter.

We take you to some pretty unique places - we don't take you off of Earth, though. There are some aspects of Xen, or the other world or the supernatural that now occur here on this planet, but the entire game is on this planet this time. But you do move from City 17, which is sort of inspired by our favourite Eastern European cities, then you move outside of the town and you move through some of these supernatural places here on Earth, and then you move along this Cliffside that was shown a little bit in the demo last year at E3 with the Buggy scene with the gunship and what have you.

It's set over three consecutive days. You start at the beginning of day one and you end at the end of day three. That was sort of a design choice, obviously that focused up would be sort of a cool approach to take to stuff, and again it sort of allows us to be a little bit more deliberate with where some of the chapters and the things end. And it's sort of a good pacing thing, or at least that's our theory, that build the tension, build the action and have a little bit of a resolve and let the player take a breath and say 'ok, now... the next day... we begin'. That's our theory on that anyway - we'll see if it's worth anything after we ship.

We're super big fans of bad 1950s sci-fi books, so that's where a lot of our fiction comes from and is inspired by, so there's a lot of Lovecraft, and all of that's in there, and if you look at some of the creature and character and vehicle designs you'll sort of see some of that inspiration coming through.

From a gameplay point of view a lot of the inspirations came from a whole different realm of games. We're really big Soul Calibur fans, believe it or not, and there's certain visceralness in Soul Calibur that we thought was cool, and we tried to bring that in. There are a lot of guys that are still really big System Shock fans, and a lot of guys that are really big Ultima Underworld fans. We play all the shooters, we're all super huge Quake players still.

I really haven't spent a lot of time seeing it. The past couple of months have been pretty nose down in Half-Life 2 stuff. A lot of people have been asking me about that, about Far Cry, and those are all on the list of games myself and a lot of the people at Valve are going to play as soon as we're done.

I still think that people aren't spending enough time with story and character development in shooters. I still look at the majority of shooters and say they're shooting galleries. Right, there's been some really good World War II stuff that's come out, Call of Duty being the most recent one, I think Infinity Ward did a great job with that, and I was about to give you an idea when the last time I really spent any good time with, and I think that and Top Spin were the last ones I spend time with. But they did a great job of recreating the mood of that time period and whatnot, but you really didn't care about them.

I don't mean to criticise those games, but obviously they weren't shooting for that, they weren't trying to take you into the story, they were trying to put you in the mood of the battle and set up that gameplay stuff - and it worked for that - but I don't think that people have really said 'let's give the folks some narrative, let's give them characters they care about' etc. And it's sort of funny for us, because after Half Life 1, after any game that comes out that's pretty well-received people always borrow pieces from it in the games that they're making, and the things people borrowed from Half-Life surprised us, because those are the things we wouldn't have ripped off from us, and the things we would have ripped off from us most people haven't. Keeping it all first-person, no cut-scene, having characters that mean something to players. Those are still things that I really haven't seen people steal from us yet - not that I'm encouraging anyone to do so! That leaves us out there to take advantage, to drive that thing by itself.

You know, we've been approached by just about every major studio there is and a lot of those meetings are confidential so I can't tell you who we've met, but we've met with some pretty serious players. And every treatment that I've seen for a Half-Life movie has stunk so badly I didn't even want to be in the room with the script! I mean, it was just awful.

Yeah, the whole deal, y'know people say 'here's what it looks like, here's the treatment' the whole thing. Y'know, fly you down to LA, put you in a limo, introduce you to movie stars that want to star in the movie, the whole thing. It's like; we're not going to do that. We're not going to make another bad movie.

If one comes along we'll do it, if not... We're in this business to make games, right, we're doing pretty well for ourselves making games, there's no need.

Myself! [laughs] It changes, y'know, In the day when Half Life first shipped we thought Harrison Ford would make a pretty good Gordon, but these days Harrison would be better as Gordon's dad, so that's probably not the right guy. For a while there were some people that Edward Norton would be a pretty good Gordon... I dunno. I think it would actually be pretty cool if we got someone who nobody had ever heard of before who's really talented and that's where they got the break. I think it would sort of go back to how we would want to play the movie in general. Y'know, we don't want to do the big Hollywood cheesy scene where Gordon falls in love at the end, y'know, the G-Man pulls off a mask and says 'I'm your father!' or whatever.

I dunno, I think it would be really cool if someone who's really well known, a Sam Rami or somebody like that could get into it and take a multi-million dollar budget and went crazy with it and had great sets and great stunts - I think it could be cool that way. I think it could also be really cool done on a million dollar budget with three cameras and bunch of unknown actors - it's going to be the treatment and the way it's pulled off that will make it.

But again, at this point, we've been round that track so many times, it's left us so disenchanted with the idea that it's now like, pffft, whatever.

Soon! Soon! Let us finish this one first. Right now we're all about finishing this one. We've had a really rough year, and we shipped the Counter-Strike Xbox, and CZ, and now we want to ship this one, and if we're able to achieve all of those within 12 months of when the source code got out, I think that we're gonna be due a little break and then we'll come out and start talking about what's next. But first thing's first!