5+ years since the release of the first web3 games. Billions raised against the promise of digital property rights, liquidity, community governance and mass player/creator enrichment. What to show for it…?
I was inspired to write this essay after, on a recent trip to Iceland, I managed after 4 hard years of trying to convince my dear friend and colleague Michael Fan (co-head of gaming at Galaxy Interactive) that web3 gaming would be a thing. I encourage you to read his take here, as it is likely representative of what a lot of gamers will come to realize in the coming years.
Looking through the top blockchain games on DappRadar, some think this is Proof of Ponzi. Others argue the best is yet to come. My own view (a topic I’ve written about here, here and here) is that most web3 game studios will fail for one of three reasons:
· insufficient depth (focus on short-term engagement hooks like staking and speculation, without ever asking the most important question: why should players care?)
· overambition relative to budget, experience and suitability (“we want to build WoW on the blockchain” betrays a poor understanding of all three at once)
· unsustainable economy as the supply of virtual goods and currency overwhelms initial speculative demand (fomenting Ponzi-style price trajectories)
Despite this, I’ve never been more excited about the future of web3 gaming. Perhaps unsurprisingly, as I’ve always been a contrarian in this space. In 2018, blockchain gaming was a core part of our thesis, leading to early investments in companies like Immutable and Mythical Games. Then in 2020, just as the world was starting to get excited about web3 gaming, we soured on it. As other investors poured billions into the space, we stood on the sidelines. Having just attended GDC, the rhetoric now is that tradeable NFTs are interesting, but not tokens because they are too hard to get right. Once again, it’s time for an alternative take.
The “really big idea” of web3 games is not ownership of virtual goods, it’s creating a fungible virtual currency. Fungibility of virtual goods and currencies with real-world value (“fiat convertibility”) enables the affinity that gamers feel for a virtual world to be instantiated in financial form. We call this “joy premium”.
As I wrote in Thoughts at the Intersection of Web3 and Creative Culture:
My overarching thesis is that we are at the early stages of a multi-decade super-cycle of retail empowerment driven by the fact that “consumption, culture and community” are now tradeable assets. Consumption is no longer ephemeral, but persistent. No longer private, but communal. No longer limitless, but scarce. Consumption is, for the first time, collectable…. When I think about how to measure consumption value, I start by asking two questions: (1) What does the asset provide by way of patronage, status, access, exclusivity or utility within the community in which it is recognized? (2) Is that community likely to be around for the long term, such that “squad wealth” can be created from sustained engagement?
There is one critical caveat: financializing joy will be positive-sum only after the “real economy” has matured in the virtual world, characterized by the production output of intrinsically motivated players pursuing exploration, crafting, production and exchange loops. Until that point, joy premium must only be earned, not bought.
Joy premium is a commodity, unbounded by any measure of intrinsic value (e.g., discounted cash flows for equities). Properly constructed, its upside is enormous, and the redistributive wealth effects it creates will be without historical precedent. The key word here is “properly constructed” -- what we need is a playbook for how joy can be financialized as a positive-sum game.
With that, I present to you the Ten Commandments of Open Economy Games:
1. GDP before GMV (for avoiding speculative destruction)
2. Multiplayer environment designed to maximize player agency and social interaction (for meaning)
3. Violence at the narrative core (for lore and PvP sinks)
4. All virtual goods are player-crafted, not sold directly by developers (for a healthy economy)
5. All virtual currency is player-owned, on the date of fiat convertibility (for avoiding lumpy speculation)
6. Sound monetary policy: separation of central bank vs. Treasury, balanced sinks vs. faucets, no competing fiat economy (for sustainability)
7. Legally recognized digital co-ops (for property rights and governance structures)
8. Alignment of player/speculator archetypes (for social cohesion)
9. VAT-driven monetization model (for alignment between players/devs)
10. Proof of humanity and anti-cheat (for economic fairness)
Indulge me in a thought experiment.
· An experienced game developer takes the long view and undertakes to build a multiplayer game with mechanics designed to maximize social interaction, promote player-driven markets (with no competing direct sales of virtual goods), and embed sound monetary policy (faucets balanced by sinks, ideally through PvP destruction loops). The game is massive, social and collaborative at its core.
· The game is funded through some combination of crowdfunding, publisher/VC financing, unit sales and subscriptions.
· The game is released in Early Access and sells 1m copies at $20 per title.
· The virtual currency in the game (gold) can only be earned, not bought. Its convertibility to real world value is deferred until the economy has achieved a certain maturity of production.
· Once that date comes, players’ virtual currency balances are converted into fungible tokens on a 1:1 basis (supplemented by a one-time seignorage event to seed an ecosystem fund). The tokens are issued by a legal cooperative, and the players of the game are its legal members.
· Those players that do want to sell find that the fully diluted valuation (FDV) of the token is $1b. One million players find their gold balances are now worth an average of $1k each. With 100% of the virtual currency in the hands of players on the conversion date, the game is spared from the ravages of outside actors (dev team, investors) with no relevance to production. If they want in, now is their first chance to buy.
· Players are shocked and delighted at the notion of having fun and earning value for their time and energy. This attracts new players to the game, invigorates lapsed players, and retains existing players.
This, I would submit to you, could easily become a $10b+ FDV token….
1. GDP before GMV (for avoiding speculative destruction)
Why do most gamers hate web3? It’s not because they dislike recognition and reward; rather, it’s because all they see is speculative excess before there is even a game to evaluate! Put differently, the original sin of web3 games was “GMV before GDP”: relying on pre-product hype to stir up interest in tokens and NFTs (GMV) before player-driven production loops (GDP) have matured. It’s no wonder that these opportunities attract transient traders, rather than the intrinsically motivated gamers needed for lasting production in the virtual world.
Joy premium is therefore a double-edged sword: instantiated too early, the extrinsic nature of market forces will overwhelm the intrinsic motivators driving players to care about the world to begin with. Most web3 games to date have fallen on the wrong side of the sword, elevating fundraising needs and speculative fuel above the long-term interests of the player base. What is needed is a nuanced perspective on how to properly introduce open economy primitives in a way that is coherent when measured against the core game design.
Intrinsic motivators are deep and varied. Hilmar at CCP recently reminded me that words like joy and fun are woefully inadequate to describe the range of simulated experiences. It would be as if the only word for describing all the food and drink in the world were “tasty”. Perhaps less than the adjective, it’s the verbs these adjectives incite (learn, accelerate, frustrate, inspire, etc.) that matter, and how immersed we get in the process.
Instead of thinking about open economies as a binary on/off switch, it is instructive to think about it as a sliding scale progression across the “Who, What, When, Where, Why and How” dimensions. Each of these lines of inquiry represents different facets of a progressive evolution.
What is an open economy?
(Step 1) One-Way Conversion from Fiat to Virtual Currency
(Step 2) PvE Markets (Virtual Goods <> Virtual Currencies, NPC pricing)
(Step 3) PvP Markets (Virtual Goods <> Virtual Currencies, Human pricing)
(Step 4) Two-Way Conversion of Virtual Currency <> Fiat Currency
Who plans the economy, and how?
(Step 1) Autocratic, Free-Form (Dev makes decisions, with no restrictions on the decision making process)
(Step 2) Autocratic, Rules-Based (Dev makes decisions, on the basis of a strictly defined set of criteria)
(Step 3) Democratic, Rules-Based (Dev outsources decision making to the community, on the basis of a strictly defined set of criteria to be applied by the community)
(Step 4) Democratic, Free-Form (Dev outsources decision making to the community, on the basis of a constitutional framework for electing representatives to make decisions without restriction)
(Step 5) Democratic process breaks down for a particular set of intractable issues, enabling “emergency powers” authority for the developer to exercise autocratic, free-form control over a particular domain of problems. Repeat Steps 1-4 for the expanded universe of decision making, until history repeats itself.
Where can users trade?
This will be determined by platform constraints. Today, trading of virtual goods and currency on Mobile/Console is largely limited to outside marketplaces, with limited functionality in game. Over time, the trend will be towards full functionality and integrated marketplaces cross-platform, so long as platform gatekeepers get their cut.
See here for a more detailed discussion of the mobile platform rules for blockchain games, from Pillar Legal.
When will we see mainstream web3 gaming?
I believe this year and next should be incubation years for the player ecosystems that are going to matter in 2025 and beyond.
In summary, I ask myself these questions for any web3 game:
- Is the game focused on developing GDP (collect, mine, craft, build, create, destroy) before GMV (exchange, speculation)?
- At scale, will I be able to deploy financial capital directly into the vehicles of production in a virtual economy (GDP), not merely its superficial speculative layer (GMV)?
This last point is interesting to ponder, especially as an investor. Instead of just investing in passive financial instruments (equity, tokens, NFTs), will I be able to start a productive enterprise directly within the virtual world? If the next Silicon Valley is going to coalesce in the virtual world, how can I be the Amazon or J.P. Morgan of it? I don’t know yet, but I’m going to be on the front lines to find out.
2. Multiplayer environment designed to maximize intrinsic motivations for play and social interaction (for meaning)
What makes GDP “real” in a virtual world is the intrinsic value we extract from the game’s core loops – exploration, crafting, exchange, production, combat, destruction, etc. In games, this requires a careful attunement of interaction loops and skill chains to present players with optimal challenge, in the context of a world narrative worth caring for, while leaving room for emergence (simulationism vs. stagecraft).
In the Chemistry of Game Design, Daniel Cook writes:
Mechanics and aesthetics are certainly important pieces of any model of game design, but in the end, such analysis provides little insight into what makes a game enjoyable. You end up with a set of fragmented pieces that tell you almost nothing about the meaningful interactions between the game as a simulation and the player as an active and evolving participant. Games are not mathematical systems. They are systems that always have a human being, full of desires, excitement and immense cleverness, sitting smack dab in the center. To accurately describe games, we need a working psychological model of the player.
I find the predictive processing model elucidated in Mastering uncertainty: A predictive processing account of enjoying uncertain success in video game play to be a particularly compelling framework for understanding player motivations.
Predictive processing (henceforth ‘PP’) is an increasingly influential neurocognitive framework of cognitive processes and how they are materially realised in the human biological system….
PP revolves around sub-personal cognitive uncertainty conceptualised as expected prediction error. In playing a game, a player wants to reduce uncertainty in the form of (a) pragmatic risk (getting closer to their goal by choosing the actions most likely to get them there), (b) epistemic ambiguity (sampling more information to get a grip on the current game state, especially in games with hidden information), and (c) novelty, exploring the game to discover new entities and actions….
Humans continually seek to reduce uncertainty or expected prediction error in its various forms (pragmatic, epistemic, novelty-related), and experience positive affect when they do so more efficiently than expected, which leads them to preferentially choose actions that bring such faster-than-expected reduction about. In everyday terms, we seek out doing well and seek out and enjoy doing better than expected – in attaining goals, understanding the world we are in, and learning how to improve both.
Entire books have been written on the topic of game design. The point being, all the same rules still apply. The mistake most web3 gaming investors made was in assuming an open economy would trump any concerns over how insanely difficult it is to get this balance right.
3. Violence at the narrative core (for meaning creation and PvP sinks)
Nobody cares about a pure sandbox. Meaningful virtual worlds must be bootstrapped through intelligent design and an incisive understanding of human nature.
As CCP states in its mission statement, virtual worlds should create more meaning than real life, not merely be a reflection of it. Some guiding principles from Hilmar (founder of CCP) on how to do so are instructive:
· Violence is everything.
· Constraints enable meaning.
· Simplicity drives purpose.
· Struggle foments friendship.
· Scarcity drives solutions.
Violence forces people to band together to survive, which creates meaningful social connections, which enables new forms of interaction entirely. In Empires of Eve, Andrew Groen explores the politics, warfare and culture that defined the narrative history of EVE Online over the last twenty years. So much of what made these battles so epic was the trickery, traitors, thefts, half-truths and political chicanery that defined the “free for all” of the universe. To get to a vision of the world based on warmth, friendship and giving, you may first need to define the digital physics of the universe in a way is cold, dangerous and bitter.
You also need a keen understanding of socio-political systems and human history, at various levels of hierarchy in the social fabric. As Taleb wrote in Skin in the Game, “I am, at the Fed level, libertarian; at the state level, Republican; at the local level, Democrat; and at the family and friends level, a socialist.”
Note that PvP destruction is the ultimate sink for sustainable in-game economies (see point 6).
4. All virtual currency is player-owned, on the date of fiat convertibility (for avoiding lumpy speculation)
Players want to believe they are in control of their environment and know who to blame if they are not. The notion that their experience can be shaped by random, exogenous economic forces beyond the control of themselves or the developers is the issue, not price volatility per se. By ensuring that all the world’s currency is held by players on the conversion date, players are more likely to view the ensuing price volatility as agency-enhancing rather than agency-limiting. It also mitigates the risk of lumpy trading flows from early investors who are not active participants in the game world from adversely impacting the player experience.
This point is particularly important in the context of Predictive Processing theory and the idea of “consumable error” and “designer environments” as fertile playgrounds for intrinsic motivation.
In natural environments, this drive to reduce error at a better-than-expected rate lets predictive agents drift towards niches that are replete with consumable errors—that is, situations that agents expect and observe to be neither too complex to manage, nor too predictable and devoid of new information….
Humans live predominantly in cultural environments or ‘designer environments’ … that are often purpose-built to maximise enjoyable (and learning-supportive) consumable error. Along these lines, philosophical and empirical aesthetics have become increasingly interested in PP as a promising explanatory framework for the appeal and aesthetic experiences found in visual art, film, music, or literature … a growing body of evidence and argument supports that media and art are purpose-built to afford consumable error whose build-up and reduction generates aesthetic experiences of suspense, interest, surprise, insight, aha moments, and delight.
If players feel that open markets bring such randomness to their play experience that they can no longer progress through the skill chain, they are likely to churn out to a friendlier environment. Limiting trading dynamics to “in-world actors” is therefore key during the initial incubation period of a virtual world.
5. All virtual goods are player-crafted, not sold directly by developers (for a healthy economy)
One of the biggest issues with the first generation of web3 games is that developers try to have their cake and eat it too, selling a token or NFT at inflated levels, but making their usage optional to avoid user friction from interacting with crypto (and retaining fiat-based in-app-purchases as the primary monetization mechanism). This is the worst of both worlds.
A more subtle issue arises when developers sell virtual goods as finished products in competition with the player-crafted economy. If these goods touch only the cosmetics layer, they present less of an issue, although they raise questions of how to maintain the narrative immersion of a player-driven economy when “God” can come in and sell finished products directly. Once upon a time, this was extremely controversial – now, perhaps not as much.
A trickier issue is how to deal with cases where the resource being sold is necessary for the means of production. Lars Doucet covered this in the context of virtual land here. My own view is that war is the ultimate arbiter, and that scarce resources like land should be apportioned through natural selection. Balancing the benefits of conquering territory against the risk of invasion and loss is something deeply intuitive for humans.
6. Sound monetary policy: separation of central bank vs. Treasury, balanced sinks vs. faucets, no competing fiat economy (for sustainability)
The other big issue I see with the first generation of web3 games is that developers confuse their role of Treasury vs. Fed. These roles need to be carefully delineated. A model I see working well might look as follows:
· The developer (Treasury) monetizes by taking a fixed percentage of every marketplace transaction (VAT). It sells only title access (premium monetization), and does not sell tokens, NFTs, virtual goods necessary for production (e.g., land), or any form of “pay to win” progression or power buffs, as such actions would interfere with the player-driven economy. This “take rate” is periodically converted back to fiat via transparent reverse Dutch auctions, similar to how the US Treasury solicits bids for Treasury bonds. These funds allow the developer to continue servicing the game through live ops, and invest in R&D for the content universe.
· The foundation (Central Bank) endeavors to pursue a stable monetary policy, one where sinks and faucets remain broadly aligned as the player base contracts and expands. Following the conversion event, this will be a deeply dynamic system, dependent on the degree to which changes in the value of the numeraire (fiat currency) impact the “real economy” (denominated in virtual currency). All net new issuance of virtual currency or virtual goods are distributed directly to players. In success, the reliability of the Central Bank attracts investors who believe the continued growth of the game will be proxied by the token.
Again, this is much easier said than done, and will undoubtedly require significant experimentation to get right. This is why developers building open economy games should plan their economy from day 1 as if their virtual currency had a floating exchange rate with fiat – this in turn requires a careful focus on sinks vs. faucets. One thing I look for is the existence of destructive sinks, whether in the form of skill-based gaming / betting loops, or the existence of conditions likely to lead to frequent disasters and war. Is it any wonder that the game with some of the most powerful sinks over the decades (EVE Online) was made in a country at constant existential risk from the next volcano eruption?
One way to kill your token economy for sure: by making its usage optional (usually under the guise of minimizing user friction) and allowing users to pay for progression in the virtual world via in-app purchases in fiat. This “leaky bucket” will ensure that sustainable value will never accrue to your virtual currency, and can only coexist with an “NFT-only” model (a pattern I see with many mobile developers building web3 games, and deserving of a separate analysis entirely).
7. Legally recognized digital co-ops (for property rights and governance structures)
The term ‘co-operative’ is defined by the International Co-operative Alliance as “an autonomous association of persons united voluntarily to meet their common economic, social and cultural needs and aspirations through jointly owned and democratically controlled enterprise”.
In theory, co-ops can be used to redistribute profits of a digital collective back to its members, in the form of patronage refunds linked to network activity. The key is to structure this arrangement so that economic and governance rights can be distributed back to players, without running afoul of securities laws (because of the active nature of the participation by co-op members to earn their membership interests). I believe co-op law is ripe for further study as a means of distributing fractional ownership of a network back to its participants in a legally compliant manner.
Creating a virtual economic and governance structure with real-world legal enforceability unlocks many other second-order primitives. For instance, guilds can evolve from being a loose social construct, into a real-world co-op with programmatic allocations of production resources and quest loot. Operating entities within the game can be created with full time employees, balance sheets and income statements, with all the upside and downside potential of real-world companies. And new financial primitives are just around the horizon—imagine the equivalent of financing a 15th century sailing expedition and having an algorithmic security interest in the fruits of the journey. In the data-rich environment of a virtual world, the sample space of possibility dramatically expands.
8. Alignment of player/speculator archetypes (for social cohesion)
Once the virtual currency becomes convertible, it will be important not to create two glasses of gamer—one that trades and the other that HODLs. Artificial incentives to constrict tradability will only serve to divide the player base over time. Mechanisms like staking serve little purpose in a virtual economy with a real GDP—in fact, they are usually discouraged in favor of consumption. The point should not be to restrict token supply, it should be to create organic demand.
9. VAT-driven monetization model (for alignment between players/devs)
Note the historical evolution in the business model of gaming:
- Coin-op (pay per play)
- Premium (pay per title)
- F2P (free to play, sell cosmetic virtual goods, ads, progression)
- P2W (free to play, sell power)
- P2E (free to play, raise token/NFT capital, propagate Ponzi)
- Digital Co-op (pay per title, bring fungibility to virtual currency once GDP has achieved “liftoff” (the point at which the ability to buy currency will not overwhelm the game), monetize via take rate on in-game transactions and periodic batch auctions of virtual ccy to fiat ccy
This last stage represents the pinnacle of long-term alignment between the interests of players and developers.
10. Proof of humanity and anti-cheat (for economic fairness)
As real money trading (or the anticipation of it develops), it will be particularly important to ensure robust checks for proof of humanity and anti-cheat. This is more problematic with some genres than others, such as first person shooters (for instance, 1 of every 6 people working on Valorant were dedicated to anti-cheat, and that wasn’t even a RMT game). Issuers of virtual currency are required to comply with KYC requirements at certain materiality thresholds, and so federated sanctions screening and transaction monitoring services are likely to be in high demand. To the extent non-documentary approaches to these problems (e.g., utilizing zk-proofs to verify certain challenges, such as age or citizenship) can be compliant with regulations, that will be a step forward, but regulatory changes are likely to take some time.
In the meanwhile, the biggest issue will be user friction: how many gamers are willing to scan their passports to keep playing? Do developers even want to know how told their users are, or would they rather remain blissfully ignorant? (Think Discord or Roblox.) Considering these frictions, perhaps the most powerful approach of all will revolve around social consensus geared towards identifying bots as a core part of the game design itself. “Proof of humanity” checks will become much more top of mind than they are today.
Conclusion
In a July 2022 blog entitled How can the next wave of Web3 games become sustainable?, Javier Barnes wrote:
The core of the problem with Web3 games sustainability is that it is not possible to fulfill the player expectation that these are positive-sum games. It is not possible to have a game where most players earn money by playing. That money has to come from somewhere. From an economic perspective, and taking into consideration the system as a whole, ultimately every videogame is negative-sum … To become sustainable, the premise of Web3 games for most players has to do a 180º flip: It has to go from “this is a game where you come to make money”, to “this is a game where you come to spend (lose) money, to have fun”.
Javier is a shrewd analyst of the gaming space. For the last 9 months, if you asked my honest feelings on the matter, I would have agreed with Javier’s take in its entirety.
It may be the Icelandic volcanoes messing with my brain, but I’ve now once again managed to convince myself that there is indeed a bright future for web3 gaming. But first, as the great Yoda once counselled: “You must unlearn what you have learned.”
Richard, very well articulated.
I really liked some of the ideas —GDP vs GMV, the one time seigniorage event (akin to IPO). The main thing is maintaining a community’s consistent trust in the team/entity to deliver a solid open economy without letting them succumb to insider behavior.
Some thoughts…
VAT-driven monetization requires a lot of faith. Devs are incentivized to push variety and velocity of goods, and develop a wide economy, but it’s not obvious you’re able to plan out a UA campaign, for example. (less risk for investment underwriting, for one)
The main challenge is that your monetization is contingent on creating valuable, deep and rich worlds, (which is a function of time but also luck), rather than tuning and optimizing for funnels and retention. Secondarily, you might find that the market size is small, much too late, as you build your game and its community. So we’re back to hit-driven Movies…
One challenge is the working capital requirements for a company to build open economy games. Not only to build but to market. F2P/P2W scaled more easily. Even $20 1M presales / sales is non trivial to achieve without some significant marketing / buildup. This is why the token sale / presale meta was promising, but it broke because well, GMV.
Second challenge for taking and estimating corporation profits is VAT evasion (lol) and alignment (I hold all these tokens. When do I sell them? Will the supply overhang scare people? Can I print more? Why/why not? etc).
Always a pleasure to read your writing!
Many thanks for this article. Have you looked into the game Rust (https://rust.facepunch.com/)? This seems an extremely intense example of the violence/survival mechanism you mention. In the early days, they tried out several ways of releasing personalization skins (ended up doing it through steam, though if they did it today maybe they would consider making them nfts on a chain where transactions are cheap gas-wise). This would seem to be close to what you think can happen with web3 gaming? What do you think?