We continue our new Q&A interview series, in which our co-founder Adam Goldberg speaks to founders about the most important inflection points in their careers, projects, and protocols.
Today we speak with Hilmar Veigar Pétursson, the CEO of CCP Games, the studio most well known for creating EVE Online.
Startups are a series of unknowable choices. All great startups have inflection points where builders are faced with difficult decisions that will compound for many, many years, for better or for worse. In this series, we will explore a range of inflections a project has encountered, some in the past with the benefit of hindsight, some in the present, and others with an eye towards the future.
This transcript has been condensed and edited for brevity and clarity.
THE PAST
Adam: EVE Online is one of the most well known MMORPG games of all time. But what really makes the game special is its extremely robust in-game economy. I recently saw a statistic saying EVE Online has a GDP of almost $10 million. Let's start off by talking about in-game economies generally, and then we can delve into some of the most important decisions in EVE Online’s past. Afterwards, we can explore how crypto is changing in-game economies and what inflections you expect to encounter in the future.
How would you measure an in-game economy as sophisticated and long lasting as EVE’s — is GDP even the right figure? I know you share really robust figures with the community every month, but if you had to distill it down to a single metric, what would it be?
Hilmar: We're not so much distilling it down to single metrics; we have a compounded metric internally to something like the Prosperity Index, which measures a lot of activities in the overall social economy. Without the context of what's happening in the social domain, it’s a lopsided way to measure things. You can measure GDP, you can measure inflation, you can measure economic growth — and while these are indicators of the temperature in the room, they are maybe not indicative of what's really going on in the room. The room can be warm because there's a fire in it, or the room can be warm because you're in a sauna.
The foundation of what is driving the metric is often more relevant and actionable, even when you're comparing economies. Elements like GDP are mainly invented to compare economies, rather than take action on economic prosperity. It's the competitive nature of GDP, I think. People have started to obsess about it in the abstract. So we have something called the Prosperity Index: it's broken down by activity, how many people are logging into the game, how much they are doing certain types of activities, how much gets produced, how much gets destroyed. EVE has a war economy, as it’s ultimately a game about spaceships and world domination. It's a lot about destruction. So we measure several factors but tend to not obsess on any singular one. When we have done that, I've seen a lot of unintended consequences, where the measurement tools become a goal in and of itself.
Adam: Let’s double-click on the durability of the economy a little bit later. But if we rewind, can you take us through the origins of what made you open up the economy to be so player controlled? It was really two decades ago, right? How do you design an economy that players have confidence in versus one that game developers overly control and manipulate?
Hilmar: Maybe it’s good to reason on the origin of economies in the first place: Why do we have economies? Then we can transpose that back into [the question of] how to replicate that process of building an economy on a totally different premise.
I think the economy is based on two elements. One is the specialization of labor, where humans specialize in something very niche and become orders of magnitude more skilled in that activity versus somebody that's trying to make everything. Imagine making a bow and arrow, where somebody makes the string, somebody makes the bow, and somebody makes the arrowheads. If you specialize, you end up with better outcomes, rather than a single person trying to do it all.
We live on a planet that has an asymmetric distribution of resources — not everything is evenly spread out. Some regions have more copper, some have more oil, etc. So we are trying to collapse asymmetries, either in skills as per the specialization, or in resource distribution. And then we have come up with ways to transfer these asymmetric resource distributions and asymmetric skill sets. So when you make an in-game economy, you need to start with these foundations. How can you make a game that has a rich-enough complexity level where people can specialize? And how can you make the foundations of the world such that nobody can just be happy in their corner of the world and doesn't need to either cooperate with others or not cooperate with others?
We thought a lot about that. How can we make a world where economic activity is the only reasonable outcome to be solved? So you're starting by pre-creating a problem, not the solution. The economy is the answer to the problems of labor specialization or resource distribution. The topology of the map in EVE Online was extremely thought-out; the nature of resources, the nature of jobs in the economy. You can mine, you can manufacture. You can be skilled in warfare. You can be skilled in leadership, communication, misinformation, espionage. Eve has all these very diversified roles, so that nobody can play all of the game, you can only play the game in cooperation with others. I think the longevity of the game has a lot to do with the fact that the problem statement of the game has not yet been solved by the community.
Adam: What really resonates is that participating in the economy really isn't a choice — which you could argue is actually much like human civilization, right? If you need food and shelter, there’s no option but to participate. I’d like to double-click on the second part of that question. How do you gain player confidence in a system like that? You've explained from a first-principles rationale why it works — but did players struggle to believe in a system in the early days, or was it just a natural transition?
Hilmar: The best way to gain confidence is consistency over time. I think that is the best way in any endeavor — do what you say, and say what you do. If something is inconsistent, it's hard to trust it. Over time, that trust started to evolve together between the company and the community, [and an agreement] of reasonable changes in the game. In the beginning we didn't have a super clear roadmap on how we were going to evolve. We had ideas, and some of them panned out and some of them didn't pan out. We learned together with the community how things should change. And we continue to learn together with our community. As the years progressed and the economy had a degree of stability and predictability through the actions of the players, people started to trust it more and more.
Adam: To offer contrast, can you tell us about something that didn't go as expected in the early economy design experimentation? What went wrong?
Hilmar: There's one very illustrative example where we were wrong. And we responded correctly when we found out to be wrong. As we were calculating the content progression in the sense of, like, you have a small frigate to start, and you want to have a big ship, a battleship, and a cruiser was the middle step. Even at the beginning we had about 30 different spaceships you could fly. (Today we have about 400 different ships just to give you a sense of how much the game has matured.) And we had calculated that the mineral fields are over here, the stations are over here, and a small spaceship can fly from here to the mineral field mines and mine minerals, take them back, and refine them. And this was all calculated: the size of the cargo hold, and how many minerals you could mine per minute. So what you could move back and forth was a known figure.
But players figured out [a hack] using a debugging feature that was left in the game and was not part of the design — which basically allowed you to eject a cargo container into space. And the cargo container was very, very big. It was like 50x what a single cargo hold could be. Players then figured out if they could mine into a shared container in the mineral field, and then use a transport ship to move back and forth. The transport ship had a very big cargo hold but very little ability to mine. So by pooling into the ejected container, the transport could materially increase the output by a factor of many multiples. So, in the beginning, we were thinking people would take years to get the battleship, but they ended up doing it in just weeks by collaborating.
We woke up to this fact a few months after release, because we were seeing everyone fly around in battleships. We were like, ‘Oh my God, how did this happen? This should be the earliest next year.’ In a way, this was the wrong way to play the game — this is not how we designed the game. But we thankfully had the wisdom to look at it that this was a way better way to play the game. Because the cargo containers were open, somebody could come and steal the minerals from the cargo container, you have to trust the person that was doing the hauling, they could just be taking it for themselves. So it was kind of a loosely coupled trust dynamic that was building between players, because you really had to trust the people that were together with you in the mining operation. So the players destroyed our assumption about how the game should be played. And on the other side, they invented a better way to play the game. We were this close to just taking out these ejectable cans, as they were called. But we left them in because this was a much better way to play the game.
THE PRESENT
Adam: It's really cool to see the extra dimensions that a community adds to the economy. We've spent most of our time talking about the economy so far, but let's shift and talk about communities. When we had lunch last, you said something that really stuck with me: EVE was always focused more on its social aspects rather than created as an addictive game. Which is a different approach than many games take today, especially in an era where we see casinos running next-gen slot machines with flashy, 50-inch 4k vertical screens that look almost like mobile games. So how did you see these communities in the early days? What worked and what didn't work?
Hilmar: We always had this notion that you should be playing the players not the game. There are many games which are more about playing the players than playing the game. Poker is a good example. Just understanding the math of poker will get you only so far. But understanding the social dynamics around the table will get you a lot further. And obviously understanding both will maybe get you really, really far.
So we had this idea you should be playing the players and not the game. The game is a landscape onto which you play the other players. We wanted to create a game with high stakes. And if you lost your ship, it should really sting. And it should sting even more if it was at the hands of another player. Meaningful human interactions are a serious matter. We very much believed that the dynamics in the social sphere will always be more important and more meaningful than the mechanics of the game. If you look at the game mechanics of human life, they are fairly simple. There are not a lot of unpredictable outcomes, like the [randomness] you find in a slot machine. EVE has almost none of these random elements because we outsourced all of these dynamics into the social sphere.
There's a way to describe systems as other L1 Chaotic or L2 Chaotic. So an L1 Chaotic system is like the weather: the weather doesn't change if you predict it. Whereas the financial markets are an L2 system. As soon as you make predictions about the market, you change the market. And this is true about economists: if you make predictions about the economy, if a central bank issues their predictions about the economy, they immediately change the economy just by creating predictions about them. And this is the part of the eternity machine of EVE Online: you can't really fully predict it because all of the players change the outcome of EVE Online. When you have an L2 Chaotic system, people can be endlessly fascinated by it, especially if they are the source of the [chaos]. So, this is, in a way, the foundation of the longevity and also the player engagement, because they really do feel that they are the game. The game itself is fairly predictable. The interesting part comes when you have two players.
Adam: I also want to also talk about monetization. How do you think about the tradeoffs between designing a durable economy like the one we've been describing — to let people play a gamers game — against simpler models that have proven to be really profitable, like pay-to-win. Does the most intellectually deep model win out, or does “fast food,” so to speak, make more money?
Hilmar: If you break down the distribution of where the money is coming from in your typical free-to-play game, compared to EVE Online which is ultimately a subscription game, you have a very asymmetric outcome in most free-to-play games. Most free-to-play games have a curve of 1% or 5% of your players who are responsible for maybe 80% of your revenue — it is a fairly well understood phenomenon. There's a lot of people that are not really paying anything. The challenge with this model is that companies generally just follow the money. Like who is your customer? Where does your revenue come from? When your revenue distribution is this asymmetric you really tend to focus on the people that are paying for the whole show. Usually, at the cost of the other ones.
So if you have a more “even” model, where people are more evenly contributing to the revenue base, the company has a more fair overall look at the customer base. So I think that due to the fact that EVE’s business model is largely subscription revenue, we have a fairly even take. You can also play the game for free, like a freemium model, but most people turn to subscription after playing the game for a while. And I think that makes us look more across the board to all different cohorts that are playing the game. And I think that is a part of the longevity. Free-to-play games with exaggerated asymmetric distribution have a bit of less longevity.
THE FUTURE
Adam: What you're saying is subscription revenues smooths over the need to go really asymmetric on the player distribution. So let's pivot to crypto being in the picture. You're working on a new crypto title. How do you expect gaming economies — especially EVE Online types — to look differently with crypto in the picture?
Hilmar: EVE Online is multiplayer online role playing game: we’ve had many of those over time. They always have people who really want to sell the stuff they make in the game. And a lot of the players that play the game, they will not buy the stuff, either from the company or from another player, without putting in the effort to make it. And this happened Day 1 with EVE. On Day 2, we saw spaceships for sale on eBay. And even if we send a letter to eBay and ask them to please don't list the spaceships, life finds a way. It's impossible to stop people from finding a way to trade on a gray aftermarket. We solve this to some extent with our PLEX system, which introduces the concept of putting extra money into the economy. But it doesn't allow you to take money out of the economy. It has long been on our roadmap, as an idea: is there a way to have bi-directional flows? Not only putting money in, but being able to take money out.
Thousands of people have played EVE Online for 20 years, and they have accumulated a lot of value inside the game in an economic sense. And they have provided a lot of value in a value sense to themselves and other players. Some have even taken up developing software for EVE Online: we have an API gateway, where people write third-party software. You go to the GitHub repository and can get a massive list of other software that has been granted for EVE in the past 20 years. Half the transactions in our database originated from code not written by CCP. So it's pretty massive.
People have spent a lot of time and a lot of energy, a lot of creativity into the game. And a lot of them also want to trade off the game, even if we have to fight it, and we do fight it, because it's against our licensing agreement. I have always thought, is there a better way to solve for this? Generally you don't want to be fighting the behavior amongst your customers. And the people who are making all the third-party software, they don't really have a business model. They often resort to Google ads just to pay for their AWS bills. I doubt anyone is making any money really on running the massive amount of infrastructure for EVE Online.
So when I started to be curious about crypto around 2015, I was actually looking into storing spaces on the Bitcoin blockchain. A little sort of toy example that I was playing around with — is there a way to make a game where you could have the premise, just different? And, and I mean, obviously in 2015, maybe not so much. It really only became practical with some of the scaling solutions that we are seeing. So we decided, Okay, let's see if we can build something that utilizes the affordances that come from blockchains, which I think is programmability, and employs a different business model.
As we think about the experience we're doing now — it's called Project Awakening — then you could look at it as an MMO where you can mod the server. It’s very common in MMOs that people are modding the client, but in this case you can mod the server through smart contracts. I believe we can build a more inclusive business model, where the people are building the whole world with us through smart contracts. They’ll have a foundation in the overall economic outcome of the whole endeavor. Not just having to do Google ads to pay for their AWS bills. And we started really looking into this almost three years ago. We have it up and running, and had about 2,000 people play it for two weeks over Christmas. And we saw some phenomenal results.
Adam: In some ways, it's like you previously had this open social service and now you're actually combining that with the game surface to create something that's really socially programmable, for lack of a better word.
Hilmar: In a way you could argue EVE is socially programmable, it largely happens through memes on Reddit, where people are waving their propaganda and kind of cultural wars. But this is almost to the point where you can program the fabric of reality of the game. In our December play test, we saw people who implemented new features in the game. And I mean, this was only a two-week play test, people had very little time to catch up to the platform's capabilities. But some of them are third-party development veterans, probably some of them are blockchain developers. And it was staggering how much people could build in those two weeks.
Adam: What was the most surprising thing someone built during that two week period?
Hilmar: One person actually implemented a corporation task management system, which is staggering. That person was an extremely experienced EVE third-party developer, so they had a very good understanding of the most useful things to build. At first I thought they must be moving some of their old code they had written over to this. No, it wasn't like that — it was all fully persistent on-chain, there were no servers involved. It was shocking. And when I saw that... this is way beyond what is reasonable to expect right out of the gate. So we are now factoring in these discoveries.
Adam: What an impressive form of user-generated content — maybe the most extreme version we've ever seen. So let's pivot to go-to-market. Some crypto entrepreneurs have used crypto to pioneer novel acquisition strategies within gaming. I think about Limit Break and free-to-own NFTs, where you can build an audience by distributing NFTs free of charge to help build a community that is aligned. You once put it to me that you're a lot more focused on pioneering on the retention side than the acquisition side. We've already talked about the social layer and what that's looked like historically pre-crypto. But I'm curious how you'll experiment with crypto with respect to novel retention.
Hilmar: We are ultimately creating a development platform where players can build the game together. And that puts an enormous amount of rigor into something we colloquially call “digital physics.” How we orchestrate the immutable aspects of the world, you build a lot of immutability into the world so that the world can be built on, and become programmable. And I believe from seeing what happens in EVE Online that when you add that to your product development capabilities, you have a lot more brainpower to put to bear, to build the game in a way that engages a wider number of people, because you have a wider diversity of people building.
I'm a big believer that when you can partner up with somebody, or you can recruit people into your social structure — where you're literally building new experiences inside this virtual world — that it will very fundamentally increase retention as you are now literally building the world. And the world will be better, as it's more engaging to a wider audience. We're not only kind of decentralizing the technology foundation perspective, but also decentralizing the whole product management and discovery phase.
Adam: One more question. A common theme we've talked about today is, it feels like you're pretty big on letting players govern the ecosystem, the platform, and help decide which behaviors are okay and what assets are worth. Looking forward to the future, what do you think regulation should look like here? Given that, you know, a scam or a crime on Project Awakening could be a theft of hundreds of thousands of dollars? Should players govern what's okay, or should governments get involved?
Hilmar: So in the beginning of EVE Online, even if it's obviously play money, this was tested quite a bit. Now, it's largely normalized what is allowed in MMOs, and whether the police should get involved. And we have figured out with various governments around the world, where is the line that we call the police in the country in question. It's not many cases. But there are lines where you're crossing the line and this is now a police matter. We are not the police of the world, we cannot regulate all activity where we don't have any control or any jurisdiction. But we have also flagged cases.
There are always going to be boundaries. And I think it is not the feature of crypto that the boundaries aren't set. It is a bit of a problem that governments haven't really decided what is the interpretation in each of the jurisdictions. I think Europe is probably pretty far along with their electronic money regulation, when it comes to the financial aspects of crypto. But when it comes to the sort of the game aspect, and certainly a game of a considerable source of sophistication, like a game like EVE Online, it is largely a fairly unwritten area of what is allowed.
If you are exploiting somebody's spaceship, what does that really mean in that context? What kind of action is that? So when we are putting this experience to market, we want to make it very clear that here, anything goes, and this is for a comfortable adult to participate in. And there is no protection offered other than something you arrange inside the game.
All the strings of the human heart will be played as they have been in EVE Online. This is that kind of experience. Do take care. We are not the nanny of this world. Frankly the way we set it up, we actually don't really have any power. That's the whole premise of making the world immutable. Of course in the beginning, we will have to have backdoors to some of the smart contracts, because getting a game experience of this level of sophistication right out the gate is a pretty high bar. But over time, I think at some point there is no backdoor. And as the time passes, the world will be fully immutable.
We want to establish this kind of almost a free state where anything goes. And everyone should be agreeing to that. There is no recourse back into reality. In EVE Online we’ve even had cases where people are cutting power lines to each other houses. That is really just normal in the world of MMO. I believe that if your intent is to do good things and not bad, the world generally allows for that. And we are excited to work with governments around the world to create more clarity.