alf-Life 2 garnered many Game Of The Year awards when released on the PC in 2004, and its impending sequel is just around the corner.  At the 2006 Game Developers Conference, we got the chance to sit down with Valve's Founder/Managing Director Gabe Newell, and Marketing Director Doug Lombardi to chat about Episode 1, Steam, console development, Team Fortress 2,  and a whole lot more.

GI: What was your biggest challenge with using Steam to digitally distribute Half-Life 2?

Gabe Newell: Deploying a large scale service is challenging, right? Customers need it to work perfectly 24/7, 365 days a year. We have this curse, which is we have millions and millions of customers and having them all simultaneously try to purchase Half-Life 2 or get an update, there are a lot of technical and provision challenges with making sure every single one of those customers has a good experience. That was our biggest challenge.

GI: Are you guys going to continue on with Steam and be done with retail or are you going to stick with retail products?

Newell: We totally expect to continue at retail because our customers tell us they want to buy our products at retail. I think that you have to be careful when you talk about a digital connection to your customers. So we did something that we couldn’t do through retail. We did a free weekend - play our game for free for a weekend. You can’t do that unless you have a connection to your customers, unless you have a robust authentication system. We did a promotion that required us to have that antipiracy authentication technology all the way to customers’ desktops.

The result of letting people pay for a free weekend is we sold a bunch of content, we did the billing of that content online for a whole bunch of people, but twice as many people who had never played Day Of Defeat before until that weekend went to the store and bought a copy and then used that to authenticate. They already had the bits on their machine. They just didn’t have the checkmark that said, “Yes, you can play this.” So we weren’t even fulfilling the product. We happened to have those bits on those discs, but that really was just the ticket that said, “Yes, you can keep playing.”

We have the flexibility now to do these kinds of promotions. The fact that those promotions cause people to go buy the product at retail, that’s great. If they want to go buy the product at retail, you make a bunch of money that way too. We’re not going to tell customers how they ought to buy our products.  As long as they want to rent them at a cyber café, as long as they want to buy them at retail, that’s great. As long as they want to have to flexibility of buying online, that’s great as well. Customers are the ones who get to vote.

GI: So are you most likely going to do another free weekend in the future?

Newell: We’re always going to be trying. It used to be if you tried to do marketing activities through traditional methods you have people who got fired at McDonalds doing stickers and trying to administer a program at your local EBX. And they tend to be very painful and very expensive and tend to give you bad data. Once you’re writing software and doing marketing programs and promotions that way you get much more precise data. You know exactly what’s happening and things that used to be sort of hearsay and anecdote turn into actual numbers. I can tell you exactly how many people bought Darwinia and then tried the free weekend and then went to a store. That’s a query against a Sequel database. It’s not talking to a bunch of field sales representatives. And so far everything that we’ve found is that every time we do one of these promotions they’re really effective. And they’re a much more effective way to spend our marketing bandwidth than through traditional means. So, yeah, we’re going to do a lot more of these. Will we do “every weekend’s a free weekend?” Probably not. But we’ll find other ways.

One of the things that we really want to do is help people to get their friends to try our games. That’s a high priority right now for us is to put all of the pieces in place. If you bought a copy of one of our games you can give five trials to your friends. You type in your e-mail address and then you’ll see the first time that they play you’ll have this little screen where you can say, “Oh look, he’s playing. Oh, he hasn’t played yet so I’m going to delete him and replace him with somebody else.” And give your friends a chance to play. It’s a multiplayer game. You get notified as soon as they start playing so you can jump in. That’s the kind of thing we think is going to be super valuable at making that connected customer experience more effective.

GI: What about stat tracking and things like that for Counter-Strike? Headshots, how long you’ve logged, etc. Is that something you guys are considering?

Newell: There are a couple of issues there. First of all people in our community are very smart. You can’t lie to them, you can’t kid them, and you certainly can’t f*** them over because there are consequences to doing that. An example of something I think we’ve done really well is the hardware surveys where we ask people “Is it okay for us to look at what your hardware is? And here are the results” I think if they thought we were taking this information for some internal nefarious purpose they’d say no. But instead they say, “Everybody can look at this, and it’s cool.” And hardware companies and operating system companies say, “Wow, you have far more detailed information about people’s PCs than we do because we don’t have this connection to our customers. We just send our PCs out in boxes that sit in warehouses someplace and we have no idea what our customers do with them afterwards. Whereas you guys can tell me exactly gamers have FX systems, how many of them are actually running dual threaded, how many of them have SLI for real?” They think that that’s super valuable. And we put our own information. But we see that survey [compared] with our own personal PCs.

And people respond to that. They say, “Okay, some people are evil bastards and I’m worried about my privacy, but so far I have confidence that Valve is doing things that I want them to do with this information. It’s just a combination of earning people’s trust and keeping their trust when you’re using that information. I don’t want anybody to ever think that we’re doing something behind their backs that they don’t want us to do with the information. I want them to always think, “They’re doing things to [help] them do a better job with customer service.” We’re increasing our service value to them. Then they’ll say, “Wow, that’s great. Go to town.” I would be horrified if people thought that we were starting to lose people’s confidence, that we were doing the right thing for them.

GI: With Darwinia and SiN rolling out soon, how many other games can we expect in the near future?

Newell: I think a lot. A whole bunch.

Doug Lombardi: Yeah, we’re pacing about two or three new ones a month.

Newell: And after the show it’s going to go up. I think you have to be careful. Digital distribution is not a portal. It’s not a different way for people to be a publisher. It’s tools for developers to connect with their audience. And I would be really surprise if, twelve months from now, any title doesn’t have some form of digital distribution strategy. It’s too valuable to developers and publishers and to consumers for people not to be doing it right now. I think there were a lot of fears about backlash from distribution. But now any concerns that people might have had about that pretty much went away when Microsoft was standing up and saying, “We’ve gotta do this. We’ve got Arcade.” And Sony is saying a key competitive feature of the PS3 is our better digital distribution strategy than Microsoft’s. Hey, we have a hard drive guaranteed in all of our boxes. That’s going to help us be a better connected gaming client than the Xbox 360.” I think that’s giving everyone a lot of comfort that this is a reasonable thing for them to be exploring and shipping to their customers.

GI: All three next-gen consoles are focusing on digital downloading. Have you expressed any specific interest in developing any sort of product for their structure?

Newell: No, we’re pretty happy with the space and the size of our audience. In the real world the console guys are but players in all of this. People who have been really successful are Real, who’s like orders of magnitude larger than Live Arcade. The guys who do Ricochet are just hugely larger than Arcade. Where this has been successful to date and the people who are learning interesting things right now are people in the PC space. So I think they’re going to be struggling for awhile to just make their consoles good internetwise, which is something that PCs have done for a long time now. I think the innovation is still going to be coming on the PC side for awhile. And the people that are going to be doing it are already large established players in the PC space.

GI: How’s development going on Half-Life 2: Episode 1?

Newell: Great. It’s a lot of fun.

GI: Any kind of release date?

Lombardi: May 31st.

GI: Is there anything you can say about the story, and how long the average player will be able to play through this content?

Newell: It took me seven hours to play through the game the last time I played through it. There’s a lot of variability of how long it takes people to play.

Lombardi:  $19.95 is the price point, and it’ll be at both retail and on Steam on May 31st.

Newell: A lot of it is about your relationship with Alex, and what the hell happened at the end of Half-Life 2. It was sort of a cliff hanger.

GI: Sort of!  (laughs)

Newell:  You’re on top of an exploding building, okay, so how do you not all die. It answers a bunch of those questions.  It also raises some questions about what the G-Man’s real role is in the Half-Life universe, so there are some surprises there for people.

GI: From the outset of design with Half-Life 2 were you guys planning that you’ll have a main game and then episodic content or were you planning on expansions?  Was there some reasoning behind this cliffhanger?

Newell: We thought it was a good way to structure the story. We spent a lot of time thinking about how the story needs to progress into the future.  We’ve got it mapped out for a long ways in terms of the story archs. There are issues that need to be resolved from Half-Life 2 and there’s a story arch that we’re going to propel you through the episodes.  And then there are longer term plot issues that the ground work is going to get laid. So we have a lot of storytelling to do. (laughs) We have more story to tell than we have ways of communicating them right now.

Lombardi:  Some of the reason why we are going episodic is our experience in multiplayer. Counter-Strike has arguably been episodically developed over the years. There was never a point where we said, “Oh this is it!” and then four years later you’ll get something else. Every couple months you’re getting stuff. After six years of building Half-Life 2 and watching Counter-Strike parallel from the other side of our office, you know, the light bulb starts going off at some point.

GI: How often do you think we’ll be seeing new chunks of story and content?

Lombardi: Episode Two is already in development – we’ll ship it when it’s great. We’re not saying our key strategy for episodic content every four months or every month, or whatever.  If the episodes are strong and people like them and sign up for Episode Two, and so forth.

Newell: They were supposed to be three months apart. (laughs)

Lombardi: We’ve learned that, once again, we’re better at making quality than trying to keep to a schedule. (laughs)

GI: Now with being able to deliver chunks of this episodically, you’re always increasing the technology, which is a cool thing that you can do with episodic content. Will that kill the need for having a Half-Life 3?

Newell: It depends on what you mean by having a Half-Life 3, right?  There are a couple of different elements. Technology is probably going to be more interivally developed now.   You don’t have to develop that in lock step with single player content release.  But the HDR in Lost Coast was really valuable. Essentially when you’re doing something that is significant, it’s like you’re breaking everything and then you’re trying to put it back together. HDR touched every element of the engine – the renderer, physics – there’s all these different pieces and there are consequences to that. If you could only do one of those at a time, having being broken in two places is four times as painful as being broken in one place, because all of a sudden it’s a lot harder to tell where you did the damage. That sort of steps on the technology side. It’s way more efficient to build that way. On the storytelling side I think right now we’re having a lot of fun – chapter, chapter. chapter. I assume at some point we’re going to say, and our fans are going to say, “That was really great for awhile, but now we want the big one.” I think that we’re having the option now of having the same choices of – tv show, tv show, tv show, movie. How do you make that decision? Fans like them both.  But, some things you want to do with a movie-like piece, and some things work better as individual episodes.

It’s taken The Simpsons over ten years to finally get around to a feature film. As a fan of The Simpsons I wish they had done the movies sooner. As a fan of Battlestar Galactica, I would love it if they would take the time to do either another mini-series – they had an interesting choice, which I think was smart, is they launched with a big thing and they’ve gone episodic, and I think at some point they’ll come back and launch another mini series or another three hour theatrical release.

But I think for developers like ourselves, we now have that choice. Where before we had no choice. You did that movie, you did it every 24 months, or in our case six years (laughs) that was really painful. But now we have the flexibility. The technology side, innovations are going to come at more bite sized chunks.

GI: How many episodes can we expect?

Lombardi: We’ve just announced the second one.  That’s all we’ve announced so far.

GI:  Can you say if you’ve got the gameplan laid out so far, or are you seeing how popular it gets?

Lombardi: We know how many episodes there are, and we know where this round of episodes will leave off.

GI: One of the games that I still play is Team Fortress 2…

Newell: I’m not going to say anything about Team Fortress 2.

GI: Awww…that hurts me. (laughs)

Newell: It hurts me too. (laughs)

GI: Well can I at least ask if you would ever do a Source enabled Team Fortress Classic?

Newell: (pause) I’m not going to say anything about Team Fortress right now.  I love that game. Everybody at Valve loves that game.

GI: I know those maps better than I could drive around Minneapolis [ed: Where GI is located]. (laughs)

Newell: At some time we should do an article of all the weird things that have happened in Team Fortress development land. There are completely unknown chapters of that history that have not seen the light of day.

GI: I think you guys have one of the best physics setups in gaming, and with things like Aegia and what Havok is doing with NVidia, how will you implement that in Half-Life?

Newell: There are a couple of different issues going on there. With our physics, the player can collide with objects. This is a good thing. The stuff that is being shown by NVidia right now, the latencies are too high. Their physics is essentially to make prettier pictures. It’s like you can have a bunch of different things bouncing around so long as they don’t actually touch anything that matters. If you don’t actually have to read the data out – if your AI system ever needed to know about whether or not one of those objects had collided with something else it would run slower by running on the GPU than having it run on the main CPU. So physics that matter is different than physics that makes pretty pictures.

The presentation quality is fine. I think we went through a round of that with graphics acceleration. It wasn’t until it enabled different types of games, games that involve the game logic being able to ascertain the state of the physics system I think that that’s the really important long term direction from everybody. So having pretty explosions is cool, but you stop seeing it pretty quickly. You only see the stuff that actually matters. There’s still a lot of work to be able to figure out how to get an order of magnitude or two orders of magnitude improvement in that kind of physics.  That’ll be a big challenge for everybody.

GI:  Valve did an Xbox version of Half-Life 2.  Were you happy with that? Are you thinking about doing more things with consoles or staying on PC?

Newell: That product was the first big retail disappointment we’ve had in the history of the company. So that sort of caused us to scratch our heads a little bit. We’ve never missed our forecast on any significant product before, so the fact that that product under performed by somewhere between the factor of two to three was really surprising to us. We still haven’t figured out what the hell that meant. There’s certainly certain technology investments that we’re interested in, and we’ll continue to make, but right now it’s sort of hard for us to get excited about the Xbox 360s install base, and the PS3 slipping so that’s where we are.

GI: You’re obviously talking with game developers here at GDC about the source engine, how do you pitch your engine versus id’s stuff, or Unreal technology? What do you think of what other people build their games with?

Newell: I think that developers are pretty efficient with determining which engines do better at various things. I think that our strongest track record is in areas of the games that get built on our tech and the sales of those games. I think Epic is really attractive to people who want to develop for the PC and the PS3. That’s something where they’re a good solution. I think developers are pretty smart at looking at what their requirements are and what’s most important to them and then using that to select an engine and core technology.

-Billy Berghammer, Bryan Vore