Valve Has a Secret

In this video, we'll look back at how Valve built the foundations of the metaverse, DAOs and even NFTs a decade ago, all while creating some of the most fascinating economic research ever. When he was at Microsoft in the early 90s, Gabe Newell commissioned a survey of what programs people had on their home computers. The second most widely installed software was Windows Number one, Doom. Gabe became transfixed by the idea that a small company that shipped games via shareware could get its software onto more computers and than the thousand pound gorilla that was Windows. After he and his colleague Mike Harrington left Microsoft, Gabe knew that he wanted to build a new kind of gaming company. Valve was founded in 1996. The company formation papers were signed on the same day as Gabe's wedding. The studio's first three games quickly became the stuff of legend. First Half Life, released in November of 1998. Then Team Fortress in 1999, which began its life as a mod for Quake before Valve acquired the team behind it. Finally, Counter Strike, released the following year, quickly became the world's most popular online shooter. Following a wild couple of years, including Mike's amicable departure and the theft of Half Life 2's source code, Valve released Steam in 2003. But the revolution really began in 2005 when the first third party games were released on the platform. This is when people realized that Steam was more than just a storefront for Valve's games. It gave developers an alternative to conventional retail for shipping their products. Although digital distribution didn't match retail volume, the lack of physical products enabled Valve and developers to achieve greater profit margins. By 2007, big publishers like id Software, who created the massively popular Doom, had started distributing via Steam. By then, Steam had 15 million active users who could choose from more than 200 games. Many of these games had already developed vibrant communities. Remember that by 2007, the first generation of MMOs had really arrived. Around 10 million gamers were actively playing World of Warcraft. So this idea of playing with other people and interacting via in game chat was already well established. But players weren't just planning raids and hanging out in game chat also enabled the exchange of items. If player A had a cool item which player B wanted, and vice versa, the two could coordinate to exchange them. In effect, they began bartering. Then, when the demand for an item, say a hat, in Team Fortress 2, which were very popular, rose sufficiently, the players would step out of the game, meet up on ebay, trade the item for real dollars, and then come back to the game where the seller would give it to the buyer. Valve noticed this and decided that they wanted to facilitate these transactions while also taking some of the revenues. So in 2007, Valve launched the Steam Community Market. The community market didn't invent the concept of spending real money on virtual items, but it solved an important economic problem. Barter markets need to overcome what's called the double coincidence of wants to. This means that it wasn't enough for people to want to just buy and sell. They had to agree to exchange exactly what one of them was offering for exactly what the other person had. By introducing real world money to serve as a medium of exchange, Valve made it much easier to trade digital assets. Around this time, Steam was basically a monopoly. It didn't have any meaningful challengers in digital PC game distribution until 2007 when Microsoft released games for Windows. So it did more than any single platform to popularize the idea of exchanging virtual items for real money. Over the years, this has led to some pretty wild stories. You can still find old YouTube videos and questions on sites like Quora asking if it's possible to make a living trading on the Steam community market. In one answer, someone named Kevin Fry said that a buddy of his made enough money from the community market to buy a Ford F350 truck, a beach house in North Carolina, and even pay for college. Steam users in places as far away as Kazakhstan were earning a living by providing services and creating items. It's no coincidence that Kazakhstan is now one of the world's leading bitcoin mining locations. It also blurred the once solid boundaries between producers and consumers. According to Gabe, there were employees of other companies who were actively working at independent game studios that made more money producing Steam assets than the salaries they made doing their day jobs. Suddenly, anyone could be a creator. Global Offensive even birthed an entire skin gambling industry where people would bet skins which had real world value on third party websites. A couple of years ago, Valve investigated and halted trading of some in game items after discovering that nearly all of the trading was linked to money laundering networks. It didn't end there though. In May 2013, Valve launched Steam Trading Cards, digital collectibles which could be crafted into game badges and subsequently exchanged for items such as profile backgrounds and coupons for game discounts or dlc. After the project left beta, players needed to purchase the few final cards to complete their set for a supported game by buying them on the Steam community market. Demand for trading cards initially rose to the point where, according to Valve, it became significant enough that there was an economic opportunity worth taking advantage of and that was when the fake developers arrived. According to a Valve blog post, these scammy developers would upload fake games to Steam and and generate Steam keys. The keys were then distributed to bots running Steam accounts, which passively accumulated trading cards. The problem became bad enough that Steam created a confidence metric to test if games were actually being played by genuine users. Cards would only be allocated to players of that game if it reached the required confidence metric. Today, most of the Steam trading cards sell for a dollar or two. There are more valuable ones, though, which usually require a significant achievement within a game. Trading and speculating on digital items didn't start with NFTs. It didn't even technically begin with Steam. But Steam did way more than any other platform to popularize the concept of creating and exchanging digital items for real cash. Don't be surprised if many of the people boosting NFTs today were buying trading cards or selling hats and skins on the Steam community market in the early 2010s. By late 2011, Steam had more than 40 million active users, which they had attracted through a growing library of games and and a unique experience that no other company could match. Steam was an immersive virtual community. Actually, it was a community of communities. Gamers were initially drawn in purely by the prospect of playing the games they bought. But they hung around to trade items and share mods they'd made for those games. Before long, they began creating meaningful experiences with each other. They made friends, met partners, watched movies together, talked about the news. The place on Steam where you hung out became like the bar where everyone knows your name. You go there because that's where your friends are and because you know you'll have a good time. Add this all together and it's clear that whether Valve knew it or not, they had succeeded in creating an ecosystem which closely resembled what people today call the Metaverse, with its own economies, cultures and norms of conduct. Gaming was the ideal first use case for the proto Metaverse, because gamers were already spending hours of their lives online, in reality adding immersive social and economic elements, or was only one step further down a road they were already traveling on. Gabe Newell understood that the communal experience on Steam was a vital part of Valve's brand, as well as an important part of why their user retention was so strong. But he also understood that he had his own version of a very common problem faced by leaders of successful tech how to smoothly scale up and meet the challenges created by growth. The most popular games on Steam quickly created economies that were so large they simultaneously Excited and spooked, Valve items that had originally been distributed for free were sometimes trading for thousands of dollars on ebay. What kept Gabe up at night was how to ensure that the prices of these suddenly lucrative items wouldn't just crash. It was a tricky problem, but he realized it wasn't a novel one because he knew that when he turned on the TV, he saw a similar problem unfolding in Europe. The European debt crisis is really complicated. The ultra short explanation is that some countries with big economic problems, massive recessions, high unemployment, slumping investment, were prevented from devaluing their currency because they were locked in a monetary union with the other countries on the euro. This meant they couldn't realign their economic fundamentals with the value of their currency, which in turn massively increased their cost of borrowing and the chance that they would default on their debt. Gabe was afraid that something like the Eurozone crisis could emerge on Steam. As certain games and in game items became more popular and valuable than others, he needed the advice of an expert. He found one in the form of Giannis Varoufakis. Giannis would go on to become Greece's Finance minister and remains in the country's parliament to this day. But at the time Giannis was an associate professor of economic theory at the University of Athens. He was also one of the most sought after experts voices on the debt crisis and in frequent media appearances strongly criticized the role played by the European Union and the European Central Bank. Gabe thought Giannis could bring a unique perspective to the problems Valve was facing. So with all the optimism of someone who had built an industry leading tech company from nothing, he sent him an email. At first Giannis thought the message in his inbox was from yet another crackpot with a secret plan to solve the crisis. But he was intrigued enough to read on. The email continued. I have been following your blog for a while. Here at my company we were discussing an issue of linking economies in two virtual environments, creating a shared currency and wrestling with some of the thornier problems of balance of payments when it occurred to me this is Germany and Greece. A thought that wouldn't have occurred to me without having followed your blog. Rather than continue to run an emulator of you in my head, I thought I'd check to see if we couldn't get the real you interested in what we are doing. Giannis was no gamer. He was a professor who had published books on game theory and political economy. He often referred to himself as an errant Marxist. But he was open minded and interested in Gabe's offer and Given that he was about to travel to the States to promote his latest book, he decided to make an extra stop and visit the Valve headquarters in Seattle. Giannis met Gabe and some of the team. He didn't really know what to expect, but he was definitely impressed. He immediately understood that he was talking to smart people who came cared about what he had to say. The more Giannis learned about Valve, the more excited he became because in addition to helping them solve their growth problem, he realized he had stumbled onto a hidden paradise. It's often said that the thing preventing economics from being a real science is that we don't have a second earth that economists can use as a control in experiments. And the economy we do have is just too messy to measure relationships with complete certainty. Giannis understood that within Steam, he could perform the kind of experiments he had always dreamed of, but could never actually execute. Every action left a digital trail. Every item could be accounted for and every exchange was instantly and reliably recorded. He could dial up or down different settings and observe changes in prices and behaviors. He was hired as Valve's first ever economist in residence and got right to work. It's amazing to go back and read the blog posts he published about his work. He built an index to see if players could make profits buying and selling items in Team Fortress 2 without any risks. It turns out that you could, especially when sales brought an influx of new players. He ran randomized control trials on different pricing strategies. He studied in game behaviors to see if an item emerged as a reserve currency that the value of other items were measured against. Kind of like a virtual US dollar. Yes was the answer to that hypothesis. It was hats in Team Fortress. It was an amazing story. Some of the most exciting economic research of the decade was a being conducted at a gaming company. Giannis and the rest of the team at Valve were learning tons about the Steam community. But Giannis was also learning about what a remarkable company Gabe had built. Right at the top of this video, I mentioned that while he was still at Microsoft, Gabe commissioned a survey to learn what programs were installed most often on people's PCs. Those findings really stuck with him. Gabe came to believe that the true lesson of Doom's incredible success was that a startup could create much more value by focusing everything on creating creating innovative and diverse new products instead of delivering the same product with incremental improvements. So he resolved to build a company that optimized for the talent and creativity of its engineers while getting rid of all the unnecessary bureaucracy and bottlenecks. The Valve handbook for New employees is not your average onboarding guide. How can you not love something that calls itself a fearless adventure in knowing what to do when no one's there telling you what to do? That wasn't just a soundbite. It was the truth. What was the point, Valve asked, of recruiting the most intelligent, innovative, talented people on Earth, only to obliterate 99% of their value by chaining them to a desk and telling them what to do? That's a good way to drain innovation out of people and replace it with frustration. So Valve was flat, truly flat. No one reported to anyone else. Gabe was the founder and president, but even he wasn't anyone's manager. Projects were greenlit by an employee proposing them and persuading colleagues of their value. If your colleagues liked your project, they joined it. Leads emerged informally, and once a project was done, you could just ship it without seeking permission. The handbook even made a big point of explaining that the reason all the desks had wheels was so that they could be moved around to create new configurations of workers and new opportunities for collaboration. Just as it had blurred the boundaries between consumers and producers, Valve was blurring the boundaries that had traditionally separated workers from managers. This was all pretty crazy, however, and there were several reasons why it could only really work at a company like Valve. First, and most obvious, was elite talent. Valve employees were trusted to ask the right questions and apply the right solutions. Another reason was Valve's reasonably small size. Today Valve is thought to have around 1,000 employees, but a decade ago it was much smaller, just over 300. That kept quality high and confusion low. Valve was also entirely self funded. Without outside investors occupying board seats and influencing business owners strategy, Valve could become the kind of company its workers wanted. Giannis saw that Valve wasn't just internally democratic, it worked like a free market. He had long been curious about the fact that the vast majority of companies, which are these huge engines of capitalism, actually behaved much more like planned economies. Thanks to the insights of Ronald Coase, we know why this is the case. Traditional firms emerged as a way of reducing transaction costs, which would grow too high if people had to continually bargain and contract with each other. But Valve was different. Decisions weren't made from on high. People weren't listened to because they had a more impressive title. They were listened to because others could see that their ideas were going to generate value for the company. The only boss who really mattered was the customer. This was audacious. And because Valve was so committed to its principles, it was successful. So to recap, Valve was radically decentralized. It was highly autonomous, and by adhering to a totally flat organizational structure and increasing encouraging Steam users to create and share in game assets, it almost totally erased traditional hierarchies. You see where I'm going with this, don't you? It may have existed half in the real world and half in the virtual world, but when you put Valve and Steam together, you get a three letter word dao. It's incredible to look back and realize how many of the foundations of Web3 were actually built by Valve a decade ago. The main concept underpinning the Metaverse, people creating meaningful experiences in virtual communities, was established in large part by Steam. Virtual items and trading cards helped establish norms of digital ownership and user generated revenue that echo NFTs. Valve and steam together functioned like the first DAO, where workers could spontaneously come together to do a sprint for an individual project and then part ways beyond these individual elements. The spirit of Web3 immersive, decentralized virtual environments that allow horizontal collaboration and value creation owes so much to Valve, the products they built and the communities they nurtured. Neither Valve nor Steam were doing anything as advanced from an engineering perspective as the people making top notch defi protocols today. But Valve, its workers and its users built a new type of Internet. Together, they built the foundations for Web3. Even though Valve familiarized millions of people with the core concepts that underpin Web3, there are still tons of skeptics out there. The biggest web3 skeptic of all is none other than Elon Musk. So to find out why Elon hates web3 so much, you have to watch this video next. Thanks a lot.