Here’s the link to this second part of the interview. In case you haven’t read our first article on the subject, covering the first part of the interview, then here it is.
Part 2 starts off with a letter from the Steamcast hosts:
Backstepper: ‘Dear Mr. Newell, we greatly appreciate and thank you for taking the time to answer a handful of questions from the community. We speak for the community when we say that you are an inspiration to us all in the industry. Your combined efforts with the rest of the fine people at Valve have graced us unforgettable moments and memories in our gaming lives. Your interview with us in August 2009 has really helped our little Podcast get it to where it is today. We are in great debt to your generosity and kindness. Hopefully one day we will be able to thank you more appropriately. Please accept this letter as a token of our immense appreciation for your time.’
To which Gabe says:
Gabe: That’s very nice, thank you very much. One thing that everyone should remember is that Valve got it’s start through the support and community mindedness of id Software. So if it hadn’t been for John Carmack, Jay Wilber, Michael Abrash supporting us and encouraging us in what we did, Valve wouldn’t have gotten it’s start.
So our expectation is, and certainly our hope is, when we work to provide people with the same level of enthusiastic support that we received, that we’ll benefit years down the road by getting to play your games or to be entertained by your websites or content.
We really are all in this together, and the more effective we are as a community in gathering and supporting talented and smart people to build the kinds of entertainment we love the better of we’re all going to be. So we really mean that, and view our activities in this area as being of long-term benefit to all of us.
First question in Part 2 is from dcxs: ‘Hi guys, I’d love to hear Gabe talk about what he thinks of the current state of the gaming industry because there are more sequels made and if there is more and more repetition and less innovation for the industry or if you think that’s not the case.’
Interesting. Let’s see what Gabe thinks about that:
Gabe: I think everybody says there’s not enough innovation. I mean, ever since I started to be in the games industry people have always said we have too many sequels, we have too many movie derided projects, there’s not enough innovation. And that complaint always seems to be constant. I don’t have a good way of measuring that. It seems to me that the ability of small independent projects to reach a wide audience is much greater now than it used to be. So you’ve got all the Indie game scene, where they’re just developing crazy, crazy stuff.
I think for a while there was this period, this really broken period of a couple of years, where the analysts were only looking at a really narrow slice of the gaming business. So they didn’t understand PopCap, they didn’t understand Zynga, they didn’t understand Nexon, they didn’t understand World of Warcraft. And so what they looked at was that there didn’t appear to be anything new happening especially on the PC, and instead they saw a bunch of stuff happening on the consoles. And the console business was much, much easier for them to measure. They didn’t have to understand Korea, they didn’t have to understand free-to-lay, they didn’t have to understand distribution that didn’t go through stores that were publicly traded that had to report earnings and gross revenue. So to all of these analysts, it looked like the business was becoming boring and entirely console focused.
Well if you look at all the interesting stuff that was happening in the industry, it was all of the not that stuff. All of the interesting things that were happening were happening in the places that were hardest for analysts to keep tabs on.
Zynga just came out of nowhere, and now they’re like, you know, in after-hours or whatever the trading places that their shares get traded they’re valued at like five and a half billion dollars, which makes them in the top three independent game software companies in the world. And that just came out of nowhere as far as they were concerned.
All of the stuff Nexon does, they’re just trying new stuff every single day. Super Meat Boy or all of these things were outside the radar of the traditional print media and the traditional analyst community. So if you ask me, it actually seems easier for people to try new and novel stuff than it was ten years ago. I’m not sure how to measure it though, so it would be impossible of me to defend either side of the position without coming up with a more analytical model for measuring innovation or difference. I certainly know that if you have a great idea, and you can get it done, it’s way easier today than it was ever before to reach a world-wide audience and get your product into customer’s hands so, that certainly is something I can state with a lot of confidence.
Next question is from StrongSoul, who asked: ‘What, or who, inspired you to start working in the video game industry?’
Gabe: For me, it was definitely, like I said previously, getting the sort of encouragement of people that were already in the industry. When someone like John Carmack says ‘Yeah, I’m actually really interested to see what you guys do with the Quake engine”, you know, that gets your attention. When Michael Abrash says, ‘Yeah, I know you Gabe, we worked together, and you have skills that will be valuable in this space’, you can come into this not thinking you aren’t bringing anything to the table. You actually have stuff that you’ve learned and can do that will be valuable in game development.
Those are the things that made it easy to jump off a cliff and start a video game company when people who you respect very highly, who have in fact been very successful in this space, encourage you to give it a go; that makes it pretty easy.
Erik Johnson: Mine’s a little bit different. Most of my professional career has been at Valve. I think it’s kind of the people that work here are the reason why I stay in this industry. It’s kind of sappy, but it’s definitely true. If I couldn’t work with the people at Valve I’m not sure I’d work in this industry, is another way of saying it.
Gabe: Used cars call you. (laughs)
Erik: I’ve always got that to fall back on. (laughs)
Next question was from our very own Emild00d: ‘If you could re-do HL1 with todays technology, would you add some of the cut content?’
A bit of a cliché’d question, but an interesting answer nonetheless:
Erik: The cut content from Half-Life 1?
Gabe: No, I think we’d just invent new stuff. I mean, I think we’re much better game designers and developers than we were back then. It’d be fun to go back and do a remake…
Erik: It makes me laughs just thinking about this.
Gabe: …There’s an unlimited amount of opportunities. I think that we would try and move forward. The cut ideas were cut because…
Erik: The Stukabat still wouldn’t make it.
Gabe: (laughs) Yeah! So, you know, it’s fun to think about that stuff but almost always you’re better moving forward than you are looking back.
Somewhere, out there, the Stukabat is crying. But through its tears, it sees only revenge.
The next question is from Mimaz, regular poster on the Steam Powered User Forums (SPUF), and co-host of Steamcast. He asks a great, great question, and asks it better than anyone else could have:
‘A little while ago, during a period of Portal 2 coverage, you made mention that it ‘never hurts to communicate better’ with your audience. It’s interesting because in terms of a certain ‘tabooed’ subject, this doesn’t seem to apply. With TF2, L4D and, indeed, more recently, Portal, Valve have found consistently new and interesting ways to maintain interest in these respective franchises. Outside of a moderate flow of free content for TF2 and L4D, TF2 also has its various character shorts and comic stories, L4D had it’s Sacrifice Comic, and Portal had its ARG. Without touching upon the next HL game, why has that franchise been excluded from such treatment? Why hasn’t interest been maintained in any form? Interest doesn’t have to consist of revealing previews, screenshots or videos after all.’
I think you can already expect what the answer will be:
Gabe: Ahh… (laughs)
Gabe: I don’t have anything to say.
Gabe: It’s a good question. What the hell are we doing?
Backstepper: Maybe we find out later?
Erik: (laughs) We’re fine with this uncomfortable silence going on for as long as you want.
hahahahahahahahhahaa i kill myself
Next question is from Blood Muffin. I don’t know how, but in each part we’ve had someone with a badass name. First it was Killsmith (or nightcabbage), now it’s Blood Muffin.
If and when Steamcast’s third Gabe interview comes around, my question will be asked under the pseudonym… goddamn it, everyone takes the badass pseudonyms. If you have any good names, feel free to send them at awesomenamesforvic(at)gmail.com. I might forget about it entirely, but hey, why not? Why not… send them in. Not “why not forget about it entirely”.
Moving on, Blood Muffin’s question was: ‘Looking forward to the future. What are some of your long-term goals for Valve in this new decade? Do you think you will be able to achieve any of them?’
Gabe: The way I tend to think of it is, it’s been a great place to work and create content that no other company in the world can. That really is how we’ve tried to think about what we’re doing for a long time. I think I’m super interested in how much we can bridge the world between the professional content developer and the ‘fan’.
How we can make that as less…we want to make it a smoother ride between people moving up that value chain. I’m really interested in that. I’m really interested in the economic nature of what we’re doing. I’m really interested in how our sense of the identity of the player is evolving. There are a lot of sort of narrow things but it’s hard to sort of elevate them. If we fail at any one of those things I just listed it’s not that big a deal, because we’re still doing the thing that’s most important which is entertaining people and creating a great place for people to work.
Next up is ChristopherD, who wants to know: ‘Do you plan on entering the portable market? Maybe with the PSP, for example?’
Wait. PSP? The NGP is the future, baby.
Gabe: So a couple of years ago, we didn’t have any presence at all in the living room and now we’re doing a lot better there. We’re way behind in the mobile space. You know, I think that the company…I’m super impressed by what Tim Sweeny and what the guys at Epic did with Infinity Blade, not just graphically but there were a set of decisions they made around it that I thought were really interesting in terms of how the design related to the positives and negatives of that platform.
You know, there’s only so many hours in a day though, right? We have a bunch of people here with DS’s and PSP’s and we have a bunch of people who are hardcore social networkers and we don’t really have any presence in the Facebook space or on iPads or DS’s. The problem is in those spaces takes away time from doing Half-Life 2 Commentary. We recognize there’s huge value there, and our customers are already there, and we’re late in getting anything to them on those platforms. If our customers are there then we need to be there, and our customers are clearly highly social and highly mobile.
Move over, inexistent Valve in-jokes! “Half-Life 2 Commentary” needs some room!
Strange answer, but fairly interesting nonetheless. We hope we eventually see something from Valve on the NGP.
Next question was from gugoslav: ‘Can you allow people to buy items from the Valve Store from the Czech Republic or other countries, and if not, why?’
I sure thought the Valve Store ships outside of the USA, but whatever. Let’s see what Erik and Gabe said:
Gabe: It’s just a question of getting a facility set up and being able to fulfill it. There’s nothing philosophical about it. It just takes work. The guy who runs the store really wants to do it, so I’m sure we’ll get there eventually.
Next, Beryl77 asked: ‘At the moment, Steam is the biggest digital distribution platform for PC’s and is getting bigger with more games that use Steamworks. Do you think Steam Cloud could become a monopoly some day? Would you even want that or do you think it would be something better?’
Gabe: Well, monopoly is the term in legal terms so I’m not really in a position to answer that question. You should ask an attorney. I think the key question is we’re really at the beginning of these sorts of things like how do you do a better job of connecting creators and consumers and how do you blur the line…
This is a point that just seems to escape a lot of people which is the difference between being a creator and a consumer is becoming very thin. The idea that we’re going to enter a period of anything remotely resembling stability is both technologically and historically myopic. There’s just so much work to do and so many ways in which these systems are going to evolve. You know, I can just see it over the next couple of years much less over a decade or two.
Our focus really is to just get our arms around a few small problems and solve those. That’s what we’re trying to do. Everybody is trying to figure this stuff out really fast. You look at stores on mobile devices and what’s happening there, I’d be really surprised if Windows 8 didn’t try to address some of these issues and opportunities as well. If you look at what Facebook is trying to do on the identity side and with Facebook credits. You know, you look at what is happening with Nexon and Counter-Strike Online. Apple, Itunes and what they’re doing. Amazon has its own take on how this stuff should be done. So we’re all learning, I think, really fast about what this is but I don’t think we have a clear idea what the most interesting opportunities are, much less the sense that these systems are going to be locked in place for any length of time.
I would be really surprised if the people who are apparently leading in certain areas aren’t completely out of the market five years later. I would be surprised if that didn’t occur. The boundaries between what are apparently disparate systems now but which end up being parts of something larger later on are going to blur constantly.
I think the main thing to do is to focus on, you know, pick up the phone, call the people who are using Steamworks and say how can we make it better? Go to your people who use Steam and say how can we make it better? That’s really the most important thing to do is just pick another piece of work, go and get it done, get it out to people, and move onto the next piece of work. You can very easily get distracted in a way that is bad for you, bad for your partners, and bad for your customers if you don’t just sort of keep thinking of how to move stuff forward in a way that’s valuable.
Veeery long, but very great.
Next was AirDecade. ‘In a PC Gamer interview, you mentioned you were thinking of Thin Client Architecture. Does that mean a service similar to OnLive?’
Gabe: So Onlive’s interesting. The thing I sort of struggle with a little bit is, you know, everyone knows bandwidth is becoming free. The problem is the bandwidth is becoming free slower than client bandwidth or client performance. So you end up looking at these two numbers, like a Calculus problem, as everything becomes infinitely fast and free what is the right architecture?
So right now, in a lot of ways, what you want to do is figure out how to optimize client performance to reduce your bandwidth and not go the other way. That’s certainly been the direction for the last 25 or 30 years which is networks get fast but clients get faster and have more storage.
I mean, the thing on the end is getting really fast, and if you’re looking for a Thin Client architecture for a lot of games, Flash is a great solution. It’s both lightweight in terms of delivery footprint but it’s also lightweight in terms of how much demand in puts on the network.
For highest quality-lowest latency obviously, you want to cash everything on the client and that’s what a game looks like today as a distributed application with a lot of client side cashing and it puts input latency as close to the client as possible. I think the Onlive guys are really, really smart and have solved a lot of interesting problems.
At the end of the day the features that you build out of this are things like rapid-demo experiences. Obviously it makes [spectating] easier. The thing I really love about Onlive is it makes video essentially a free object in your system. So if I want to take a piece of video and composite it into my game world. If I want to put TV Screens with CNN or ESPN on them inside of my game world or to take some arbitrary video input and use it as the basis for some informational or gameplay mechanic I can. And that’s really pretty powerful and seductive. So I think there’s a lot of interesting opportunities and I think it really helps to have a longer-term perspective on how these things have evolved.
IBM was really pushing in the late 80’s and early 90’s to treat the PC as a terminal, which is sort of the Onlive approach. […] They couldn’t have picked a worse time to do that because everything that was happening in Silicon was going in the other direction and fairly rapidly you got into this perverse situation where PC clients had higher energy throughput than the mainframes they were being connected to. Whether or not we’re in the situation reversing itself now or if this is just another variation on that time will tell. But we certainly have some smart companies that are taking a stab at figuring it out.
Next, Mikhail Popov wanted to know: “Epic is really pushing its UA3 to mobile devices. Do you see Valve trying to make Android and iOS versions of Source? If not, how about Steam on mobile devices? Plants vs Zombies is already on iOS and would benefit greatly from Steam Play functionality.”
Good question. Let’s see what Gabe had to say about that:
Gabe: Yeah, there’s zero doubt that mobile clients are important. The first thing to do is to get Steam Community support up in a mobile space just because it’s easier, not because it’s more valuable to customers.
Like Erik said earlier, you’re always looking at how much work is it versus how many customers benefit? You have to look at both sides of that. And then getting the engine up and running and then getting games up and running. The thing that worries me a huge amount is to fully take advantage of the capabilities of those devices. Getting Source up and running on a mobile device is actually not that hard but creating a great experience that lives within the constraints of the device and takes advantages of its upside is a lot more work and I don’t think we’ve made any interesting strides in that direction at all.
Not because we don’t think that there’s opportunity there but simply because we haven’t managed to get a critical mass of people at Valve to tackle the problem. I keep thinking that you should be using the camera. If you’re not using the camera on one of these devices you’re making a mistake. If you’re not using GPS you’re making a mistake. If you’re not taking advantage of the fact that a person always sort of being connected to that device or having a reasonable expectation that they’re connected to the device you’re probably making a mistake in terms of missing what will end up being the super valuable opportunities to do cool entertainment.
So those are pretty nebulous ways of talking about the problem but that’s the kind of thing you have to wrestle with when you start saying, ‘so if we’re building something on an iPad, why is it going to be worth our customers’ time?’ Simply taking one of our existing pieces of content and getting it to run on the iPad is a small step. It’s a necessary step, but it’s not a sufficient step to doing useful work to take advantage of.
Erik: It’s not a strategy.
Gabe: It’s not a strategy, right. Getting stuff to run there is fine but…
Erik: We wouldn’t do anything in the mobile space until the thing we were doing would teach us a bunch about the space. We recognize we’re behind on that.
It’s great to hear that even if Valve will never move to the mobile or portable spaces, they’ll be doing so for their customers.
Next question was from Jeromy, who asked: ‘What do you think of competitive multiplayer?’
Interesting question, even if it wasn’t put very well.
Gabe: Ah, I like it. It’s kind of an open-ended question. It’s fun to compete against other people.
Erik: It sucks to lose. It’s really awesome to win.
Gabe: John Marello can be made to have his entire head unscrew and fall off on the floor when we play. Do you have a better understanding of what the question is? Do we think that the competitive multiplayer scene…
Erik: Yeah, are we talking about professional gaming, maybe?
Backstepper: Not in the question, but I would guess. Maybe it’s that direction.
And those things are all pretty tractable, and the nice things about professional gamers is they tend to be sophisticated in their descriptions of what it is that they would like. It’s pretty easy to get a laundry list from them and say, okay, we can do three of these right away. I think the harder problem is how to make what they are doing, which is awesome…every time I see one of the people at the elite levels playing any game whether it’s WoW or Dota or Counter-Strike it’s just stunning how good it is.
But I think what’s also stunning is the terrible job we as an industry have done so far in terms of making it entertaining to watch and appreciate how talented and skilled those people are. So if I had to pick one problem to tackle in that space it would be to make it more fun to be a spectator. Is there anything that you know about what’s going on with the competitive scene…?
Erik: I think Dota 2 is going to be the product where we are going to focus on this problem. It presents some advantages over some of our other games in terms of viewing just because of the type of game it is. It has a very well established mature professional scene. We’re going to push on this. I think it’s way undervalued. There’s a lot of room there.
Dota 2 is starting to sound quite promising.
Next was spider323, who asked if Valve has learned anything from the early release of L4D2.
Gabe: I think it was interesting. I mean, I don’t how much inside baseball people want to know. But for us, there was sort of this argument internally about a couple of different things. So Tom Leonard, who led the L4D2 project, said that it would be useful for the company as a whole to be able to prove that we could ship a product on time, or rather than constraining resources or rather than constraining quality that we constrained time. He proved that we could ship a product on a very specific schedule. It’s not something we’re necessarily going to do very often.
It’s good to know that we can do it, and that there were a bunch of lessons we learned about how do you do Triage, how do you think about quality, how do you think about features. That was all really useful from a development perspective. So people who have worked in development situations know what I’m talking about, about the organizational lessons you have to learn in order to do something that way.
I also think it was interesting to me personally to realize how much angst it caused the community when we did do that, and that surprised me and I’m sort of annoyed with myself that I was surprised because it would have been easier to do a better job than we did at talking about that initially to customers. It really was very much like, ‘Look, we’re going to do something in a year’. And it was very much, ‘Oh my God, they’re only taking a year to do something’.
And it was that second reaction that we were ill-prepared for and it was stupid, really on my part than anybody else’s, not to anticipate that and to do a better job of explaining what was going on. We’re not going to do that very often. It’s nice to know we can do it. Nobody’s like, ‘oh, let’s ship a new L4D every year until the end of time’.
Erik: There’s some unique factors there to that…
Gabe: There were a group of people who really wanted to do it which surprised me.
Erik: Yes! That’s the tiebreaker. Internally, on anything, the people that are going to do the work want to do the work, then no one’s going to get in their way.
Gabe: I thought everybody was sick and tired of L4D after we shipped it. I thought it would be like two or three years before we would go back and do anything major in that space. Instead, they were like, ‘we want to go into crunch mode next week!’. It’s like, ‘Oh my God!’.
Erik: Are you sure? Yes? Okay. (laughs)
Gabe: ‘Are you sure? Can I talk you out of that?’ One thing that may or may not be obvious to people from the outside is that a lot of our decisions internally are heavily affected by people’s interest and excitement in working on it. And it’s not because we’re self-indulgent employers. It’s that there’s a huge amount of information encoded in that kind of voting. So essentially creating a marketplace internally for ways that people can invest their time. The biggest danger for the company is that any one person starts making decisions for everybody else. Because even ‘super-genius Gabe’ is full of hideously bad ideas.
The easiest way for us to know I’ve had yet another really bad idea is that no one wants to work on it. When a bunch of people really want to work on something there’s probably going to be a lot of value.
The PS3 for Portal 2 really came out of that. I mean, we were right at the sort of tipping point of ‘do we, or don’t we’? I personally was really excited to work with Sony to get Steam running on the PS3 because I think there’s a lot of benefits to us and a lot of benefits to our partners if we do that. But the tipping point was that guys like Vapally and Sergay totally wanted to work on the PS3 again because they’d worked on it before they came to Valve. So that was the thing that made it easy to make the decision. I don’t know if that marketplace of where you spend your time has really been discussed at Valve in the past and the public but it really influences a huge amount of our decision-making.
Erik: If you want to know how project’s are doing at Valve you ask people, ‘are you having fun’? And are you having fun means a whole bunch of things and it’s different for different people.
Gabe: And it’s a term of art inside of Valve.
Erik: Yeah, I mean, ‘are you having fun’ could mean that you’re working like crazy but that’s what you think is the right thing to do. Fun for people here usually means the time I’m spending right now feels like it’s the right thing to do.
Next up was Slipperywhenwet, one of our readers and a regular on the Steam Forums, who asked: ‘What are some new IPs Valve would like to do in the future? I guess if you could choose anything maybe.’
Gabe and Erik said:
Gabe: Seattle Seahawks football simulator.
Erik: Our current obsession, and the problem with our obsessions is they tend to lead us into weird places, our current collective obsession is with the Seattle Seahawks (Johnson laughs).
So we’re all trying to figure out…so our current strategy is that Erik should make friends with Paul Allen and get adopted as his son so that when Paul dies, that Erik can inherit the Seahawks, so we can all own the Seahawks. So we’re all obsessed with football this year. So if we could figure out how to do a football game we’d be all set.
Erik: A Seahawks football game.
Gabe: So we’re even at the point where we’re designing half-time shows and allowing the fans to interactively fire players for particularly bone-headed plays on the spot. The firing mascot would come out and drag the offending player from the field…
Erik: And people could call plays from the stands like with a controller. We’d crowd source…
Gabe: We’d crowd source coaching. Or, at least, play calling. So, you know, in our fantasies running the Seahawks is pretty high in our collective fantasy lists right now.
Erik: The serious answer of how we look at new IPs may be interesting for people.
Gabe: Go ahead. I didn’t know we had a serious way of looking at it.
Erik: I’m about to get serious. Yeah! (laughs)
Gabe: Ok. Erik has his serious face on! (into voice recorder)
Erik: (laughs) We think that there’s nothing magical about new IP. In a lot of ways for products it’s this extra challenge you put on top of a product. I think a lot of times people use the term ‘new IP’ they mean new experience or a new way of being entertained. The difference between HL1 and HL2 is pretty extreme in terms of the experience that a player is having. It’s in the same universe and, I guess, the same IP. It feels like a lot of the times they ask a question about a new IP they’re asking about the thing we did between those two products.
Gabe: What are you talking about? Character design? Are you talking about art direction? Are you talking about game mechanic? A lot of times people assume that you do all those things at exactly the same time and it’s almost never the right answer. You’re almost always making a mistake to attempt to expand in all directions simultaneously.
Erik: L4D1 was new IP. There we had a game which we felt pretty great about the amount of entertainment it would generate for people but it carries with it this problem which is you got to tell everyone…like, explaining what kind of game L4D is, is kind of hard like it is with any product. You can have somebody sit down and play L4D and they’ll tell you I’m having a bunch of fun…
Gabe: L4D started though as terrorists with knives. In that sense, we had a really tractable way of thinking about the problem for a long time. Is that a new IP? What we could have set it in. Why wasn’t it set in Counter-Strike land?
Erik: That’s a fair question. I don’t know…
Gabe: Well it’s hard to explain why the terrorists only have knives. (laughs)
Erik: Yeah. I mean there’s definitely some value in the collective lore about zombies which we learned as we were shipping L4D.
Gabe: It would have been an interesting discussion internally if we had asked ourselves that question. So the Boomer could have been an IED strapped. You could have actually said let’s not try and innovate in character and world settings and instead innovate around the game mechanic.
That’s the kind of question that would have been a super legitimate conversation internally. We might have ended up with the same answer it at least would have been a way of talking about it that would have been stimulative internally.
Alright, throw Fart Cops out, throw Unnamed Submarine Game out, hell throw everything out. Valve’s Seahawks is where it’s at.
Next, ThatDudeFromPortugal asked: ‘Does Valve plan any collaboration with other companies in the near future?’
Good question. I would be particularly interested in hearing about a possible collaboration with Minh Le on Tactical Intervention, or if Counter-Strike Online (perhaps Neo and HL2: Survivor as well) will ever be released, in any form, in the West.
Gabe: I’m not sure what collaboration means because we work with our Steam partners all the time. We do stuff with the Sam and Max guys and the PopCap guys and looking into do that. We’re working with Nexon on Counter-Strike Online. So yeah, we love to work with other companies. It’s a lot of fun.
I still want to do Counter-Strike versus World of Warcraft. I keep telling the guys at Blizzard that we should do that. I think that’d be awesome. You know, bring in your WoW character with all of its stats and try to have a bunch of SWAT guys take your character down. I think that’d be a hoot. But it hasn’t been green-lit yet.
Counter-Strike versus World of Warcraft? Alright, screw everything, man. This is where it’s at.
Next up was k00pa, one of our collaborators: ‘Why doesn’t Steam have a speed limit option?’
Erik: Oh, like, control the rate of your download?
Backstepper: Yeah, for example, when he plays it he’s not downloading super fast…
Erik: So he’s not choking out his…
Gabe: In general, we should probably have a bunch of features. Like you should be able to metre how much your usage is and stuff like that. Those are features, they’re good features, and we’ll add them as they make it up the priority list. But there’s a whole bunch of features related to download management that would be good to have. There’s no question of if its useful, it’s a question of how do we prioritize stuff? And how do we prioritize stuff that is the most bang for the widest audience?
So many little features, so little time!
Next was our very own David-B-737, one of our contributors and collaborators, with a great question: ‘Dear Gabe, How does the management of developer teams work at Valve? Are there teams who work on multiple games at the same time or is each team working just on one project? Are there teams who swap between games?’
Gabe: So we don’t tend to actually think about things in terms of teams. The whole design of the company is built around people self-managing. So what we look for people who are very capable of doing their work, who are good communicators who also have experience managing large projects. The reason that they need to know how to manage large projects is that they need to know when something that they’re doing is irrelevant to other people and when what they’re doing has to be known about by other people.
In other words, they have to make the decision that I’m doing something now that I have to tell other people about because it’ll have an impact on support or have an imapct about how we talk about the product. So each person essentially has to know how to run their own company even though they’re not having to make those decisions. They have to know what the side-effects are and when they should get other people involved. So for those kinds of people we really want them to be managing their own time and the most important decision they make is what it is they should be working on. So we’re really good at invention and we’re not nearly as good as sort of a hierarchical command and control system at doing repeatability and measurability. Those are the consequences of the kind of organization you make.
We very much optimize towards the autonomous agent making its own decisions about what it does. So what that means is I don’t ever tell anybody what to do. I always ask people what it is that they want to do because there’s a lot of good data there. So we have people at the company who work on every product and we have other people who don’t work with anyone else at all and they’re off doing their own purely individual contribution. They’re a project team of one. And each person is expected to make their own decisions about that and to be making those decisions in the context of what can I do today that is of the greatest benefit to customers.
EJ, actually Erik, the person sitting next to me, tends to be very cross-group focused and also tends to spend a lot of his time in group contribution making other people more valuable as opposed to say somebody like Jay Stelly who probably splits his time 50/50 between group and individual contribution. Or as opposed to, say, Yan whose probably 90 percent individual contribution 10 percent group contribution. Everybody’s sort of this free-roaming agent. I mean Yan picks an interesting problem to solve and then he goes and solves it. Sometimes that involves working with the same group of people on the same project for a long time and sometimes it involves him working with 4 different projects simultaneously. So each person has to make their own decision about that. And that decision making autonomy not just applies to what they do but things like office-space. So these office-spaces are designed by and changed by the people who are working in them.
So we have this thing right now where everybody’s desk is on wheels and they have only two points of contact with their environment; power and networking. And so somebody can move their desk, and all the desks are designed to fit on the elevators, so people can move their entire desk and be somewhere else working with a different group in 15 minutes.
Erik: People move multiple times a week. Some people.
Gabe: So people are joking now we need to motorize them.
Erik: And have robots run them.
Gabe: (laughs) Yeah! Have Tom write the AI and have all our desks end up in a cluster some place. […] So the answer to that is going to change regularly for everybody. The reason that we do that is that it’s fun for really talented people…I mean, the average person here can pick up a phone and have a new job in five minutes. We have probably 20 people who could pick up a phone and have a 20 million dollar game deal signed within a day. And they’re here because of the environment and the colleagues that they work with. That’s what we optimize for. We optimize so that those kinds of people, which people either are or are becoming, have the best environment for what they do.
What that means is we’re terrible at certain other things. So a great place for Jay to work we’re a terrible place for a recent college graduate with no experience. We’re terrible at hiring people straight out of school. There’s certain personality types that we’re just not a good place for. People who like a lot of structure and a lot of predictability will be driven insane working here. And there are other things that we’re bad at.
Traditionally, we’ve been bad at schedules although I’d say we’ve been better at that in the last couple of years. But measurability? There’s absolutely no attempt made to keep me abreast of what’s going on in large pieces of the company. I get involved if something’s gone horribly haywire and that is an expensive way. But the reality is it doesn’t happen very often. We also tend to reinvent the wheel a lot. If we’re going to do something, ‘well, okay, let’s start from basic principles’. It’s not like we say, like the Portal 2 launch is not about taking the Portal 1 launch and doing the same thing again. There are costs and benefits from that. I think it’s useful for our fans to have more insight into how things work here because it helps make some of our decisions more transparent and makes us a little more predictable which I know can be frustrating.
One of the thing’s I’m sort of curious about is when I talk about this stuff I don’t know if we make any sense at all.
Erik: We make a bunch of sense to each other.
Gabe: Yeah, we all have been working together…I mean, you and I have working together for 14 years now? So we have all these memes and tropes and concepts that we’ve shared and we know exactly what they mean, and we know how to apply them, and what the consequences and trade-offs are. I’m worried sometimes that people outside will listen to us talking about this stuff and saying, like, ‘those words make no sense, I have no idea…’
Erik: ‘Now tell me the real story’.
Gabe: Yeah, now tell me the real story. What did you really mean? So I don’t know if this is even passable by people outside the company. So it’d be curious to hear…
Next is Zunekade: ‘I think it’s a really interesting way of working together. The next question is, what can we expect in terms of updates for the Steam Community and friends for this year if you can answer?’
Erik: I personally don’t know the details on that.
Gabe: I got nothing to say.
Then came ThatoneJeff, with his second question: ‘Will you ever answer questions for your Developer Wiki page again?’
Gabe: Oh, sure. I haven’t been getting updates on when I get questions. Yes, I will, and I need it to be more automatic rather then when someone reminds me to do it.
Erik: During an interview.
Gabe: During an interview. Oh, yeah, I should probably do that, should I? This is another example of repeatability.
Erik: (laughs) Yeah! Exactly!
And we’re close to the end! This question is: ‘So, Danater wants to know: ‘Will you maybe make a Half-Life movie in the future?’, asked by, you guessed it, Danater.
Gabe: Certainly it’s something we’ve talked about a bunch. The problem that we have is the same problem we have when we were having this conversations with people outside is we need to have an interesting movie in mind before we commit to it. There’s really so little value to making a bad movie about anything, I know it’s super exciting if somebody shows up and says ‘Hey we want to option your movie and we’ve got a director and a writer attached to it, blah blah blah’ – it sounds really exciting and then you sort of take a step back and say ‘There’re been a lot of really bad, bad, bad video game movies made and the world doesn’t need another one’, and I think that applying that test to anything we do is just as valuable as applying the test to any of the proposals that we ever say.
Erik: There’s been a couple of cases of really similar lines of thought, like there’s a pattern here that you can see where… there was something else that was similar. We’re just unwilling to do work at all unless we think our customers are going to get a bunch of value, it’s not worth the risk of…
Gabe: There are too many other…
Erik: We’re not going to set fire to anything, in that way, it’s too – we put a high cost on that.
Gabe: There are no cult third strikes.
Erik: Right. We’d rather not do the work.
Gabe: We’d rather not take work away from something else.
Erik: Yeah, not take work – that’s to be more crisp, yeah.
Gabe: Everything we do means we’re not doing something else, right? So I was working on this goofy story idea with one of the comic book guys, it was sort of like a take on — my idea was Superman would go insane, because every time Superman does anything it means people are dying.
Not that we’re Supermen – it’s a stupid analogy, it’s not an analogy, it’s just like an amusing thing – so, every time Superman sits down with Lois Lane and has dinner, it means a whole bunch of people somewhere died because he was having dinner. But he can’t be having dinner…
Gabe: …without somebody on a ferry in southeast Asia drowning, every choice he makes has consequences. If you can hear everybody in the world, you hear everybody dying. You hear everything going wrong, eventually it drives Superman insane. So, probably what I was doing was projecting my own issues with not getting, you know, developer commentary for Half-Life 2 out sooner…
Gabe: It’s like everything I do has a tradeoff! Erik’s laughing, he gets it!
Erik: I haven’t heard this story before but it’s pretty funny!
That’s actually a pretty cool story idea.
Final question: ‘Do you want to do anything about censorship and things like age-verification for countries that require it or want it like Germany or Australia?’, and I have no idea who asked it. Sorry.
Gabe: Are you saying will we be adding age verifications to address some countries concerns about censorship so that people who are above a certain age can then have access to unmodified content?
Backstepper: For example, yes.
Gabe: Yeah, that would be a fine thing to add. We’ve got Torsten is like a huge fan of that internally and has been pushing to get that done. Torsten is persuasive and effective so I’m sure sooner or later we’ll get that out on Steam, and certainly our view as a content company is that there should be no restrictions on content. Of course that’s a pretty self-serving view; anything we can do to help our customers get access to the content they want, we’re going to figure out how to do.
Backstepper: Good! That’s about it, I don’t have any more questions from the community. Do you want to say or maybe even ask anything?
Gabe: Well I’m really curious to hear if people enjoyed this interview, if it gave them useful information, if it made any sense! And I really want to thank you guys for taking the time to come all the way out here, it’s a hell of a long plane ride and everybody spent a lot of time and energy putting this together and I want to say that we’re all grateful for the support that we’ve received from everybody in the community, so you guys did a good job, thanks!
Backstepper: And you!
Well, that was a great series of interviews!
We’d like to thank Gabe, Erik, Backstepper, and of course, Brad, Sam and Saurabh from Steamcast, for giving everyone not only this awesome opportunity, but such a great and informative series of interviews.
Thanks for reading, but remember to check out the interviews themselves, if you haven’t already!