Justin (Nightcabbage): Here we go. Alright. Recording.

Gabe Newell (Co-founder and Managing Director of Valve Software): Hi, my name’s Gabe Newell. I’m President and co-founder of Valve.

Justin: And this is Justin. Nightcabbage on the forums. I will be interviewing Gabe here with questions that you, the community, have given us. We’ve tried to select a few...

Gabe: Is there going to be a transcript of this?

Justin: One of the guys might transcript it. I think the other guys who run the podcast; they might edit it a little bit, clean it up and put it on the podcast or something like that so...

Gabe: Well, welcome to Valve, I hope you had a good time so far.

Justin: I did, I really did. Well, all right, I’m going to get into some stuff here and hopefully there’s...if there’s anything you can’t answer just me let know. First question. This is one of the most commonly asked questions that we had received and we tried to form it into something you might be able to answer.

You’ve kept Episode Three under incredibly heavy wraps thus far. We would like to know why you have chosen to adopt such a reclusive approach this time around as opposed to previous releases. Was it based on the reception you received about letting out too much info prior to Episode Two or something completely different?

Gabe: I think that what’s going on is that we’re sort of, you know, we’re sort of always experimenting. We’re always trying out different kinds of things and that has positive as well as negative consequences for ourselves and for the community. So if you look at our different products, we’re trying out these different rhythms. Right now Team Fortress 2 has this every six weeks or every eight weeks where we are going to do a release and we see what people like and what people don’t like and how it impacts gifting, and how it impacts player minutes. Those kinds of things are sort of quantitative ways of measuring response and then there’s qualitative methods used by reading forums and seeing what people think of what we’ve done. Do they like the rate of change and the kind of changes we’re making there? With Left 4 Dead, you know, we did the first release November of last year and now we’re going to do a second release a year later so it’s a lot slower than the Team Fortress release schedule but for some people they’re upset, you know, because that it’s so much faster than say the release schedule between Half-Life 1 and Half-Life 2.

Right now, the Half-Life 2 Episodes themselves are on a third sort of ‘rhythm’ and, you know, we think it makes sense for the product and for what we’re trying to do there and the reason we’re not talking about it is mainly we don’t have anything to say. It’s not like we decided we released too much information, it’s just that if we had information that we were in a position to deliver to people we would and right now we don’t have anything to say about it. And it really is a consequence of these different sorts of rhythms to release schedules which we’re trying out.

Up until the emergence of the internet release schedules were pretty much the same for every product and every publisher. Every couple of years you do a big release and the reason that you do it is that there’s a bunch of friction associated with doing one of those releases, you know, you’ve got to get advertising and marketing and the retail and it’s four months between doing an interview with a publication and having it appear in print. Compare that today where we’re doing this podcast on a Friday and I assume that within a fairly short period of time it’s going to be out to a whole bunch of people who can then ask me a bunch of questions or follow-up questions or ask you follow-up questions so there’s this big difference between this old method where everybody knew exactly the right way to release their projects and today we’re sort of wanting to make a bunch of experiments and try a bunch of things out, some of which will work well for people and for us and some of which won’t. And so Episode Three is sort of this victim of our willingness to experiment but as soon as we have stuff that we’re ready to talk about, we will.

Justin: Ok. Fair enough. This is an interesting one. One of the staples of Valve as a developer is that you have great on-going support for your games post-launch in the form of new updates and new content – we’ve seen DLC in Source multiplayer games back as early as CS and Day of Defeat: Source – and the trend continues at an even greater pace with Team Fortress 2 and Left 4 Dead. We don’t exactly see the same approach to your single-player titles such as the Half-Life series or Portal. Some of the hurdles might be obvious, but could you elaborate on if you’ve been toying with the idea of how to include some sort of on-going support for single-player games and what ideas or challenges might have been presented in that?

Gabe: Well I think...I tend to be a little bit sceptical of these industry terminologies but I’m going to have to invoke one. I think that we had lived in this world where everybody thinks of entertainment as sort of this blockbuster phenomenon. Entertainment as blockbuster where the whole point of producing a movie was to spend two to three years doing it, 150 plus million dollars building it and then drive as many people into the theatres on opening weekend as you possibly can and once you’ve created a product, it’s pretty much done. This really was the industry focus for a lot of entertainment companies starting with Jaws. Jaws, back in 1973 when it was released, really had a huge impact on how everybody thought about tent-pole marketing or blockbuster mentality. It had a huge impact on all of that.

Right now we’re moving towards something that we call ‘entertainment as a service’ and with that you try to answer the question of how we can do something every day, how can we be entertaining on an on-going basis and how can we create value for people? How can they say I’m glad I’m a fan of this because I read this interesting post or they added a new weapon or there’s this new map, and that’s very much where our customers are. I think if you look back at the late 90s your average Quake player was so much more sophisticated in his or her understanding of where the marketing and distribution of games were than your typical game publisher. If you took all the publishers in the industry and put them in a room, your average group of gamers knew a lot more about what the future was going to look like and I think the same thing is true today. You know, you’ve got all of these people and these publishing, distribution and marketing roles who are used to thinking in a blockbuster approach but your average gamer knows that each day I go to Kotaku or listen to my podcast or whatever and are much more connected on an on-going, daily or even sub-daily of consuming my entertainment and that’s where we need to go as entertainment companies.

I think if you look at where we’re furthest along at doing it right now is Team Fortress 2. That’s where we’re closest to where we think customers already are in terms of things like cross-media orientation. They don’t just want games, they want movies and they don’t want the movies to be generated by some shabby licensing deal with the highest bidder who tries to find the lowest production costs to jam out some crappy movie to take advantage of popularity of Tomb-Raider or whatever. The same thing is true the other way around. You’ve got these great movies that get turned into really shitty games. But our customers are like, “yeah that doesn’t really work for us, yeah we’re cross-media but we want the same feel, the same understanding of the characters and situations of the universe and the story to be informing both of those. So those are some of the characteristics this frequent updates, this cross-media consumption, this sophistication around that forums are, in many cases, just as important as content from the game producer.

Another aspect of this is that you’re in partnership with your community. Where a movie director can pretend that the movie experience is something that goes on in their head and that film actually does violence to this vision that’s only in its purest form inside of their head and it’s up to the viewer to approach that vision. This is the auteur theory that has a lot of credibility and academic film circles and as a game company you can’t get away with that. You’re in partnership with...if you don’t have a good idea of what the player has going through their mind that’s important, you’re going to be a terrible game developer. If you don’t recognise that for a multiplayer game that the community of people playing that game has just as big an impact on the experience of somebody coming into that game as the game rules themselves do, then you’re going to be a terrible game designer. And then if you’re looking at something like service provision. We had this console vendor came to us once and said “we’ve this awesome plan, it’s mind-boggling, and we’re going to have 150 game services running your game”. And we said that’s going to be great and I’m sure it’s going to be really expensive for you to do that but by working with our community, by working with fans and so on, we have 150,000 game servers and they’re located around the world and not just in North America. So that recognition that you’re in partnership with your community to create these entertainment experiences is another characteristic of what’s going on.

Now, like I said, Team Fortress 2 is where we’re furthest along in terms of understanding those kinds of issues but whether it’s Portal or Half-Life or any of our other single-player games where you need to be able to do these things; regular updates, the partnership with your community, the cross-media generation – that’s just as applicable to those as it is to our multiplayer games. We’ve just gotten further down the road with our multi-player games but we need to do exactly the same kinds of things with our single-player games as well. I also think that over time you’ll see there are lots and lots of distinctions between single-player and multi-player experiences. Your single-player experiences are going to have to scale up to include you and your friends because there’s such a strong social component. Whatever’s fun by yourself, it’s probably going to be more fun if you can drag some of your friends along into it. I don’t think you can continue to have the super-simplified world and story-structures that multi-player games have and be able to get away with it. We need to combine the richness of the worlds and authored experiences of the single-player games and allow for more people to participate in that and that’s a big design challenge. Either way, you’re going to need to figure out how to update it really frequently, pull the audience into the participation and creation of it and then also figure out how to make movies and comics around that.

Justin: All right. Many people are still wandering about the extra content provided to 360 owners that PC owners have gone without, mainly being the engine update and achievements for Half-Life 2 and Episode One. I think that originally people expected this because somewhere floating in the forums somebody had a screenshot of an email supposedly from you saying that, and this was shortly after the Orange Box release, where somebody noticed that “hey my 360 version looks different from my PC version”. So I don’t know if that was true in the first place or where that stands now.

Gabe: It’s something we said we would do and something we intend to do. We just need to find...you know there are trade-offs, right, spending time getting the engine enhanced and also doing commentary for Half-Life 1. One request I get a lot is having commentary tracks back on the original Half-Life. So, we’d like to do it. We just have to fit into a schedule when we also feel a lot of pressure to do Counter-Strike Source updates and do more Left 4 Dead content. It’d be a tough call, getting a new monster or new special infected out for Left 4 Dead 1 versus getting the engine updates out for the original Half-Life 2. You know we can look at that and see right now that a lot more people are playing Left 4 Dead than are playing Half-Life 2 and we know that if we actually did the updates a lot more people would play Half-Life 2 so it’s just a balancing act. But it is something we said we’re going do and we’ll get around to doing it. It’s just managing all of the things that people would like us to do, all of which we want to do.

Justin: It seems a lot of new features in Left 4 Dead 2 are aimed at beating cheap survivor tactics like closet-camping and player stacking. While these changes in this sequel are welcome, can we expect any sort of fix to prevent or discourage these tactics in Left 4 Dead 1.

Gabe: I think Left 4 Dead is really important to us and we want to continue to support that game and support that community and grow that community. So I think you should expect to see us continue to address these and other issues. There are lists of bugs that people want us to fix and we need to get those fixed and clearly there are lists of new content that people would like us to do. We’ll get around to it as quickly as we can.

Justin: Shortly after unveiling the Source Engine, Valve was promoting its modular nature saying it would extend the life beyond other game engines. We’ve certainly seen the benefits thus far. Are you starting to see any limitations with Source for the near future and how long do you think the Source Engine will last as the platform for your games?

Gabe: I think the key thing is that there is a lot of value to releasing regular updates to our technology base. It’s a super useful thing to do because the benefits then are going to accrue to all of the different games that we have. So I think that as products like Team Fortress 2 continue to iterate and add new capabilities to those games we want to see those capabilities and technologies being applied to Left 4 Dead or future titles in our other game categories. So, I think that, we would switch up when there was a sort of a discontinuity, when there was a group of features that were just going to take too long to be able to capture them in a single update. That’s one we would need to switch out of the rapid iteration and move towards a bigger chunk of changes.

One of those changes would be something like Larrabee. To date we’ve seen traditional CPU architectural styles best exemplified by Intel’s CPU’s and then you’ve had a different architectural approach to what is essentially a CPU on the GPU side, which is something that we...in the various names for it Massively Multi-Core is one name for it or Through-Put Computing is another approach to it. I think at this point it’s pretty clearly demonstrated that there’s a limit to the sort of out-of-order execution approach that’s been used and that the through-put architecture is the right approach in the long run. The good news for that is that the same gigantic performance games that we’ve seen over the last few years and 3d rendering are going to now apply to every aspect of a game. You know your physics will scale the same way, your graphics have scaled, and your AI will suddenly have a huge amounts of opportunity to improve and the cool thing is that if your AI isn’t running fast enough all you have to do is reduce your resolution and now suddenly you have extra CPU cycles available to spend on AI and in the past you haven’t had that scalability knob that a user can control and set themselves. It’s like you would sort of allocate a fixed amount of CPU to these various tasks because you ended up making a bunch of least-common denominator decisions about a lot of the things that makes games, games as opposed to movies and when you move towards these other architectures that goes away. So I think you get this huge scalability improvement and we also get much more granularity and allocating those resources which is good news for all of the other pieces.

But that also requires you to go back and rethink a whole bunch of architectural decisions. You need to, in a very detailed way, go back and make sure that you’re optimised for that kind of approach. Scaling up to a couple of cores is one thing but scaling up to, you know, 34 or 64 and greater threads of execution require you to have a different way of thinking about the work-flow inside of your engine. That would be a kind of example where you’d need a fairly large jump in your engine architecture to take advantage of it. We’re big fans of Larrabee. We think Larrabee is the logical response to everything that’s been learned over the last few years about the importance of through-put computing and we’re really excited to see that architecture sort of proliferate, you know, not just as a graphics architecture but as a general computing architecture.

Justin: Do you plan to expand upon your current physics system further? Are you happy with your current physics system? Are you starting to eye something like PhysX?

Gabe: I think we’re pretty comfortable with, you know, we’re happy to use third-party technology if it solves problems for us and for our customers. I think some of the interesting problems to solve that need to be solved for some future game requirements haven’t really been addressed by any of those packages and I do expect that we’ll be moving our technology forward to address those concerns. If somebody can help us get there that’s great but right now I suspect we’ll end up having to write a lot of that code ourselves.

Justin: In interviews, you guys describe how basically you get quite a few emails from the community once a title, extra content or a patch is released. Many of them, I imagine, are quite temperamental and vulgar along the lines of “I hate such and such about your new game, this shit sucks, I hate you Valve” and yet it seems as though you take that into stride, extrapolating what you consider to be needed and valuable feedback. How hard is that for members of your team to do when you basically have sometimes what seems to be ungrateful and jaded fans slandering your hard work? Don’t you take it personally?

Gabe: I think that it helps that we’re also all fans ourselves. We understand that fundamentally the impulse behind it is a positive one. We’ve all had the experience of being a huge fan of something and then seeing it completely fucked up. I personally will never forgive George Lucas for what he did to Star Wars. So when people are that way, they think that they have to turn everything up to 11 to be heard. So whenever I response to somebody who starts off with an email by saying, “hey, you fat cocksucker”, that’s the opening line, I respond and say “hey I’m actually reading your emails”. They usually become pretty polite when they realise they’re actually talking to a person. It’s just something that we’re all pretty used to at this point. We’re actually all on the same side. We’re just as interested in building great entertainment products as people are in having us build them. I think the fact that we are on the same side as the fans make it’s a lot easier to separate the signal from the noise.

Justin: Do you think games are still making the leaps and bounds they were ten years ago? What do you think the next big thing in the game industry will be?

Gabe: Well I think we’re starting to see more and more freedom from...you know you really can build a bunch of interesting games you weren’t able to in the past. I mean if you look into the Indie game scene, what’s possible there, is a lot more interesting than what would have been possible ten years ago for a smaller group of independent producers. I think that that’s like the way of taking the temperature of what’s possible. I think Portal was interesting just in terms of showing that novel game mechanics are still worth an awful of a lot. I think we’re very much at the beginning of an industry of discovering interesting entertainment forums and interesting ways of building those entertainment forums. Where it’s going to go...anybody whoever makes those kinds of pronouncements ends up looking really stupid two or three years out so I tend to be a little bit reluctant to play prognosticator. But I do expect that there is this deep characteristic about being part of the entertainment experience. When I first started, certainly during the early 90s, there was this perception that entertainment was fundamentally a passive experience. That you hear the same phrase repeated over and over; “after a hard day at work I just want to sit down and turn my mind off”. And this was something you sort of thought if you weren’t in the flip-on NBC and turn-your-brain-off camp. It seemed like there was this conventional wisdom that people wanted to be passive, that they wanted to get the same experience that everybody else was.

I think that there are deep reasons that psychologists talk about and entertainment companies tend to ignore about why people actually value experiences that they are participants in and that that’s a key driver behind where the entertainment is going to go. That if you’re a part of it, if you’re having an impact on it, if you view yourself as having agency in that experience that you’re low’s a lower and your high’s are higher, your memories about it are greater and that it’s just a more valuable experience to you in the long-run. That is an engine that will drive our industry forward. Years and years and years from now people will laugh at the idea that games were the red-headed step-child of the movie industry or the TV industry and people will say that of course everybody would much rather be a part of the action rather than a passive spectator. That will drive a huge transformation where people will spend their time.

Justin: In a recent interview, you mentioned exploring the possibility of the community investing into titles through steam for developers to use the seed money to get going. When you mentioned that, and obviously this is sort of forward looking, are you thinking more along the lines of indie or start-up titles or do you think that might be something you implement into Valve’s development structure?

Gabe: Well Valve is in this situation where we don’t need funding that comes from outside of us and that’s sort of a luxury, which means we carry all the risks and so on. I think it’ll end up being something that starts more with the Indie side, with people who can build sort of a grass-roots audience and have that direct consumer appeal because the people who are able to fund triple A titles today tend to be fairly conservative in their response to...you know they can afford to sit on the sidelines and watch the internet develop or new game styles emerge and then they panic after a while when they suddenly see how you can know that they’re taking off and you end up in a situation where everybody suddenly realises they have two quarters to ship a Guitar Hero clone that everybody copies and races after that. So I think you’ll see some stuff like that happen first with people who own their own IP and look more like they’re the indie scene than they do the traditional side. I could be wrong. It seems like, in a lot of ways, the rate of change in the game industry, the old methods of doing stuff are clearly being replaced with new techniques and somebody might come along, especially if some studio heads really gets into the idea and views it as not only a funding mechanism but an information channel. When people have an opinion its one thing but when people have an opinion where they’re placing bets they’re vastly more reliable so if you look at the speculation markets where people were bidding on whether or not Obama or McCain would win, there’s a big difference between people voting based on preference and voting versus putting their dollars at risk.

So if you look at one of these predictive markets as a way of doing funding decisions, your audience will help steer you to make better decisions about your product if they have a financial stake as well. It’s just one of the characteristics of those kinds of votes. So people might realise that’s it’s actually a super valuable way to let’s role and if we add this feature and this feature which one is going to cause more people to be willing to invest in that project. So it’s something we’re actually looking at. There is a whole bunch of incredibly scary legal issues associated with it like are you going to become like NAZDAC. Are you offering securities to what are called non-sophisticated investors? So there a whole bunch of issues that has to be worked through to make this possible but it is my personal project right now to try to figure out how we can do this, simply because I think connecting customers and using...in the same way that you can create a service provision for game services that’s vastly greater by working with your community. I also think that the same impact will happen on the financing side that better decisions will be made about financing. That there’s nobody better to have as an investor in a project than somebody who’s a fan of the outcome of the project because they’re highly motivated for you to be successful above and beyond their hopes of making a return. I think that there are lots of good reasons for why this is a good direction go in and so we’ll keep going down this route until we’re either successful at it or convince ourselves that there were things we didn’t understand that sort of precluded from being an interesting source of funding for games and for entertainment.

Justin: What games are you playing right now? What are some of your favourite games of all time, a part from your own?

Gabe: I’m actually going to be at BlizCon so I started playing Diablo 2 again. I found the old multi-res hack so I can run it at 1600 by 1200 and I’m jumping around being a barbarian again. And then I’ve gone super retro and I’m actually a precursor to Rouge, I mean to NetHack called Rouge. I managed to find an old copy of that. It’s sort of the distant ancestor of Diablo actually. So that’s where my head is right now. My favourite game of all time still would have to be Super Mario 64. I think it was a super brilliant game. I think that it’s done stuff with space and the re-use of space and locations that’s still an education to any game designer to go back and play that game. I still find new things that impress me about the choices that the team made. I’m interested organisationally and what happened with Mario. I thought that Mario Sunshine and Mario Galaxies didn’t have those characteristics. In fact, they seemed to be completely unaware of those factors. Mario Galaxies, in my mind, is going in the wrong direction because you spend even less time with these spaces and they have less significance. I go back and I start up Mario 64 and I walk in through one of those pictures and I have this whole range of feelings and emotions and expectations and memories about those spaces that a game like Mario Galaxy and it’s approach to space and location could never possibly ever create. It makes me want to ask them about those decisions because the thing, to my mind, that was most special about Mario 64 would totally fight with the design of Mario Galaxy and so I would be curious about why they went in that direction.

Justin: Well, thank you very much Gabe. I appreciate this opportunity very much. I sure hope the community does as well. We’ll have this up here probably in a few days for the community to be able to hear. So thank you very much for everything you and Valve has done. So thanks.

Gabe: Thanks for the interview.