Gabe Newell on Valve’s new horizons

Gabe Newell’s up to old tricks. When Develop spoke to the Valve president earlier this year, his description of the company was outrageously utopian.

He claimed, if you can believe it, that no corporate hierarchy exists within the Washington studio. He said it was the employees themselves who decided Valve was suddenly going to support PS3. He insisted that Portal 2 work was not delegated by managers, but a project chosen by developers.

Newell, now speaking to Develop a few months later in a sweltering, windowless Gamescom booth, is at it again.

“I had nothing to do with initiating Dota 2,” he says in reference to the studio’s brave new project.

“The most important thing is to come into work and see a team of people all bleary-eyed because they love what they do. That was how Dota 2 came about – people were super interested in working on the game and so they got together and started making it.”

This time, to be fair, it’s easier to believe him. Valve has an inimitable reputation for building FPSes that sell in their millions. It also owns the Half-Life IP; one of the most lucrative brands in core gaming. Surely any company director, if the decision was down to them, would want the studio to play to these strengths. In fact if any of today’s publisher CEOs took control of Valve, the group would have already built Half-Life 3, Half-Life Black Ops, Half-Life 3DS, Half-LifeVille. The developers would have had no choice in the matter.

But Newell lets his studio vote by enthusiasm. And instead of reaching into its reserves of lucrative IP, Valve has chosen to work on the obscure Dota brand. Instead of building another FPS, the group opted for the abstruse PvP/RPG/MMO/arena-strategy-action genre.

These are not safety shots, but it’s not enforced ambition either. Valve is developing Dota 2, many miles away from its creative comfort zone, simply because it wants to.

“I haven’t seen Erik [Johnson] have this much fun in half a decade,” Newell says.

“He’s working twice as hard as he’s ever done, because he likes [the game’s original developer] Ice Frog, and he wants Dota 2 to work, and he loves these types of games.”

Today Nintendo is the only games studio in the world with a reputation for building pioneering, outstanding blockbuster games across not just one genre, but several. With Valve fleeing from its FPS roots and venturing forth into uncharted territory, I ask Newell if he wants to match that reputation.

“There wasn’t much discussion about deliberately doing something out of our comfort zone with Dota 2,” he says.

“This was about a group of people in the company that were more excited to build this game than anything else.”

In the Q&A below, we discuss Valve’s new project in more detail, as well as explore the progress Steam is making on PC and consoles.

When we last met in May, you said developers who make the same kind of game over and over again will “become the enemy of their next project”. You said they’ll be conditioned to do one thing and won’t be able to easily change direction.
Yeah, absolutely.

So how did you get your studio, which has been making FPSes for over a decade, to adjust to a project like Dota 2?
I think it helped that people who started working on the project were just huge fans of what Ice Frog had been doing. They were approaching it as enthusiasts, they had to answer to their inner fan when making this game.

In terms of development milestones, this project is a little different. One thing is that Ice Frog had a collection of testers who were being used since very early on in the development of Dota. They’ve been really invaluable to us in terms of shaping and reacting to what we’ve done.

Also, the need to run this game internationally is very much its own project development milestone. It’s forced us to address a bunch of issues and finish up on a number of services and features in the game.

That’s a completely different approach to what we did with, say, the Half-Life 2 project. With that we couldn’t just, half way through, decide to run a tournament that’s going to be watched by a couple of million people.

In that sense it’s a different mindset and approach to what we’ve done before. It feels right, and it’s been super useful. There’s already a bunch of stuff we’ve learned that will cause us to change the build we put out to our testers next week.

There are clear changes in the product too. Valve’s website is running a video of the $1 million Dota 2 tournament, complete with commentary and scoreboards.
With that, we got some people we really liked and respected within the community, and trained them up with the system we built, to provide commentary for the tournament. The reception to what they’ve been doing is fantastic.

You’ve hired from the community?
[PR guru Doug Lombardi: Yeah we’ve reached out to the community and have commentary in Russian, Chinese, German and English.]

Newell: I certainly wouldn’t want to be an announcer! I couldn’t announce a football match. I’d be all “oh golly! That was nifty!”

I take it with this kind of game, Valve is targeting the lucrative Asian market too?
We really don’t know how to do that kind of targeting. It’s really, really hard to make a good game, let alone make one targeted for a certain demographic.

I might be possible to build a game for one area and one market, but honestly we don’t know how to do that. We’re just trying to make a good game first. Gaming is too hard to predict. If you asked me to predict whether Left 4 Dead sold more on PC or console I’d definitely say PC, but it actually sold more on consoles.

So the plan is to make Dota 2 good. That’s just so hard in itself. We need to focus on that. Maybe one day we’ll be smart enough to say we’re making a game for left-handed bisexual women over the age of 40.

What are the commercial advantages of this kind of competitive game you’re trying to build? I get the impression that Dota 2 customers will invest far more time than, say, social network gamers, so does that give you an advantage with things like display advertising or perhaps microtransactions?
So, a game is a really distributed economic problem. You have all these people adding value to a game, but 99 per cent of them are not part of the system’s design.

One of the best things World of Warcraft has is its user interface mods, but that mod is being built entirely outside of their economy. And as much as I love World of Warcraft – I continually use it internally as a high bar for a lot of our own design – Blizzard needs to pull that sort of community-built content into their own design.

People are creating content for you, so you have to build a system that allows them to create and publish their work, and then be paid for it.

When you look at competitive players, they add a huge amount of value to Dota 2. When I was watching the tournament earlier, I was learning something about how to play every ten seconds.

We think the way people design their own interesting hats in Team Fortress 2 is creating value, we think these Dota 2 players are creating value. We’re trying to design a system that we, and other developers using Steamworks, can recognise and allow the community to value this kind of contribution.

Yesterday the Gamescom trailer was the number one thing on YouTube, and that’s an indication that we’re on the right track in terms of designing in the value of these kinds of players.

I’m really interested to see how well people respond to the gameplay videos being streamed right now, because it will tell us a huge amount about the game.

The first impression I got with Dota 2 is how demonstrably versatile the Source engine is.
Well what we’ve always felt is that you can move technology faster through regular updates to the tech, and we obviously think that’s working so far. Dota 2 is of a genre that didn’t even exist when we first built Source.

I know the headline of the Dota 2 contest is how you’re spending $1 million, but really there’s far more to it than that. This is about Valve TV, and about a community facing itself. What does Dota 2 say about the future of interactive entertainment for Valve?
So, I try to think of each game we make as a building block. Dota 2 wouldn’t have been possible without a bunch of stuff we learned from Portal 2, and Portal 2 wouldn’t have been possible without a bunch of lessons and we learned from Team Fortress 2 and Left 4 Dead.

You build something, find out all the mistakes, then try to extend it and find out if it’s still working. With Team Fortress 2, for example, we should have changed the economy support so it would work with Portal 2.

With Dota 2, we think the technology we’re building will be valuable to our other games, and we’re going to migrate the tech into Steamworks for our Steam partners to use.

So if your studio is building a game for Steam they can use the same tech you’ve used for the Dota 2 competition.

You mentioned Portal 2 there and I recall in May you explained how you want many of your games built under the credo of ‘entertainment as a service’, in that you’ll continue to evolve games to adapt them to customer tastes. Is Portal 2 going in this direction?
Oh absolutely. I think you’ll see us pushing very hard in increasing the overall play value of Portal 2.

It used to be incredibly expensive to continually engage with your customers about what you can offer them. That was the hardest part, to make sure your customers were aware of what you were doing. You used to have to book SuperBowl adverts.

These days I think it’s way more important to think of how we provide an ongoing significant value to our customers, and how you get your community to work with you to do that, and how you can get your community to care about what the community is doing.

With Team Fortress 2 it’s been a huge success pulling the community in to build assets for the game. I mean, they build ten times as many assets as we do. When they do something that’s valuable, rather than having people give them a thumbs up, they have $40,000 in their PayPal account. It’s things like that that indicate to us we have a better handle on designing an online system built for the community.

How is Steam handling on PS3?
Yeah it’s going well. With the release of the first Portal 2 DLC people will really see the value of Sony allowing Steam to be released on their system. We can patch and update regularly, and I think not only us but other developers are benefiting from Sony’s approach.

I think Sony will start to benefit from what it’s doing. They’ve done the scary thing and I think it’s up to us as developers to make sure Sony and its customers are rewarded.

We and other developers have to show we’re creating value to Sony for opening up its network. We have to show them it’s a smart business move.

Dust 514 [CCP Games project that unites PS3 and PC via online gameplay] needs to be a success to show that an opened console is a smart business move for the platform holders. The same is true for Portal 2, would you agree?
Yes it’s absolutely important. We need to reward Sony for making this decision. We need to prove Sony is right.

I wanted to discuss the conflict of interests with EA. Should Steam customers expect more EA-published games disappearing from your online store?
So, we’ve created a set of services, and they need to be valuable to customers. Personally, I hate having to install things on my PC, it’s like a huge tax. If something’s being installed on my PC it’s using up memory, it’s using up network bandwidth.

The point being, we have to earn the right to be installing content on a regular basis. The same thing is true of game developers and Steam. We have to prove we are creating value on an ongoing basis, whether it’s to EA or Ubisoft or whoever.

We really want to show there’s a lot of value having EA titles on Steam. We want EA’s game on Steam and we have to show them that this is a smart thing to do.

I take it this issue hinges on EA’s wishes to sell DLC directly to Steam customers.
It’s a whole complicated set of issues. I don’t think Valve can pick just one thing and think the issue would go away if we fixed that.

We have to show EA it’s a smart decision to have EA games on Steam, and we’re going to try showing them that.

I think at the end of the day we’re going to prove to them they have happier customers, a higher quality service, and will make more money if they have their titles on Steam. It’s our duty to demonstrate that to them. We don’t have a natural right to publish their games.

Finally, what do you want Valve to be doing the next five years?
The focus is always based on three constituencies. One is making sure everyone at Valve is happy, and having a great time. The second is building things they couldn’t build anywhere else. The third is creating value for our customers and our partners. That’s what we’ve done in the past and that’s what we need to do in the future.

There are of course so many opportunities, and we’ll have to pick and choose. But as long as we stick to those three principals I’ll be happy. When we stop doing those things, that’ll be the beginning of the end [laughs].

[Additional questions provided by Lewis Tyler]