Chapter 8
METAVERSE THEORISTS LIKE TO USE THE TERM “interoperable assets,” but this is a misnomer because virtual assets don’t exist. Only data does. And it’s here—at the very start—where the problems of interoperability begin.
Consider the “interoperability” of physical goods, such as a pair of shoes. The manager of an Adidas store in the “real world” could decide to prohibit a customer from wearing Nikes in their store. This would be a business decision, and obviously a poor one, and almost impossible to enforce. A customer wearing Nikes can get inside an Adidas store by opening its door. This is because physics is universal and thus atoms are “write once, run everywhere.” The fact that Nike shoes physically exist means they are automatically compatible inside an Adidas store. The Adidas store manager would need to create a system to block non-Adidas shoes, write a policy, and then enforce it.
Virtual atoms do not work in this way. For virtual goods from a virtual Nike store to be understood in a virtual Adidas outlet, the latter would need to admit information on these shoes from Nike, operate a system that understood this information, and then run code to operate the shoes accordingly. Suddenly, the admission of sneakers has changed from passive to active.
Today, there are hundreds of different file formats used to structure and store data. There are dozens of popular real-time rendering engines, most of which have been fragmented further through various code customizations.* As a result, almost all virtual worlds and software systems are incapable of understanding what each considers a “shoe” (data), let alone being able to use that understanding (code).
That such enormous variation could exist might surprise those familiar with common file formats such as JPEG or MP3, or who know most websites use HTML. But the standardization of online languages and media stems from how late “for-profit” businesses came to the internet. iTunes, for example, didn’t release until 2001, nearly 20 years after the Internet Protocol Suite was established. It was impractical for Apple to reject the standards already in wide use, such as WAV and MP3. Gaming is a different story. When the industry began to emerge in the 1950s, no standards for virtual objects, rendering, or engines existed. In many cases, the companies producing these games were pioneering computer-based content. Apple’s Audio Interchange File Format (AIFF) is still the most common audio file format used to store sound on Apple computers; it was created in 1988 and based on game maker Electronic Arts’ 1985 general purpose Interchange File Format standard. Furthermore, video games were never intended to be part of a “network” like the internet. Instead, they existed to run on a fixed and offline piece of software.
Virtual worlds today have so much technical diversity for this reason, but also because of the intense computational and networking demands of modern gaming—everything is purpose-built and individually optimized. AR and VR experiences, 2D and 3D games, realistic and cartoon-styled worlds, high concurrent users and low concurrent user simulations, high-budget and low-budget titles, and 3D printers—all use different formats and store data differently. Full standardization would likely mean underserving one application, falling massively short on another, and so on—often in unpredictable ways.
Image 2. From the web comic xkcd.xkcd.com
The challenge goes beyond file formats and approaches more ontological questions. It’s relatively easy to agree on what an image is—they’re only two dimensional and don’t move (with video files just being successions of images). But in 3D, especially with interactive objects, agreement is far harder. For example, is a shoe an object, or is it a collection of objects? And if so, how many? Are the caps on a shoelace part of the shoelace, or separate from it? Does a shoe have a dozen individual eyelets, each of which can be customized or even removed, or are they a single interconnected set? If shoes seem hard, just imagine avatars—would-be representations of real people. Forget trees; what is a person?
Beyond visuals, there are other attributes that must be examined, such as motion or “rigging.” The bodies of the Incredible Hulk and a jellyfish should not move in the same way, but this means the creator of these avatars needs to enshrine them with code detailing this movement and which another platform can understand. To permit third-party objects, platforms will also need data that describes a good’s appropriateness (e.g., nudity, penchant for violence, language style and tone). A game for toddlers needs to differentiate between a PG-rated swimsuit and an R-rated one. Similarly, a gritty war simulator will want to know the difference between a sniper wearing a tree-like ghillie suit and a sniper that’s actually an anthropomorphic tree. All of this requires data conventions, and probably additional systems, too. A 2D game will want to be able to import a 3D avatar, but restyle it accordingly. And vice versa.
So we will need technical standards, conventions, and systems for an interoperable Metaverse. But that’s not enough. Think about what happens when you send a send a photo from your iCloud storage to your grandmother’s Gmail account—suddenly, your iCloud and her Gmail both have a copy of that image. Your email service does too. And if she downloads it from her email, there are now four copies. Yet this doesn’t work for virtual goods if they’re to hold value and be traded. Otherwise infinite copies will exist every time they’re shared between one world and another, or one user and another. This means that systems are needed to track, validate, and modify ownership rights to these virtual goods, while also safely sharing this data from partner to partner.
If a player buys an outfit in Activision Blizzard’s Call of Duty and wants to use it in EA’s Battlefield, how is that to work? Does Activision send the outfit’s ownership record to EA, which manages it until it’s needed elsewhere, or does Activision indefinitely manage the outfit and provide EA temporary rights to use it? And how is Activision paid to do this? If the player sells the outfit to an EA user who doesn’t have an Activision account, what happens then? Which company even processes the transaction? What if the users decide to modify the outfit in the EA game? How is that record altered? If users have virtual items scattered across multiple titles, how do they ever know what, altogether, they own and where what they own can or cannot be used?
The 3D standards to use (or not use), the systems to build and data to structure, the partnerships that need to be struck, the valuable data that must be protected but also shared—these and other issues have real-world financial implications. The largest of these considerations, however, might be how to manage an economy of interoperable virtual objects.
Video games are not designed to “maximize GDP.” They’re designed to be fun. While many games have virtual economies that enable users to buy, sell, trade, or earn virtual goods, this functionality exists in support of play and as part of the publisher’s revenue model. As a result, these publishers tend to manage in-game economies by fixing prices and exchange rates, limiting what can be sold or traded, and almost never allowing users to “cash out” into real-world currency.
Open economies, unrestrained trading, and interoperation into third-party titles all make creating a sustainable “game” much more difficult. The promise of profit naturally brings about work-like incentives for players, but these can erode fun—the game’s very purpose. And a level playing field for competition, also part of what makes playing a game fun, can be easily undermined by an ability to buy items that otherwise had to be earned. As many publishers monetize their games by selling in-game cosmetics and objects, they fear the moment their players stop buying their virtual items because they’ve bought them from a competing developer and then imported them. Given all of this, it’s understandable that many publishers would rather focus on making their games better, more appealing, and more popular rather than on connecting into a not-yet-formed virtual goods marketplace with unclear financial value and likely involving technical concessions.
To achieve even a measure of interoperability, the gaming industry will need to align on a handful of so-called interchange solutions—various common standards, working conventions, “systems of systems,” and “frameworks of frameworks” that can safely pass, interpret, and contextualize information from or to third parties, and consent to unprecedented (but secure and legal) data-sharing models that allow competitors to both “read” and “write” against their databases and even withdraw valuable items and virtual currency.
Interoperability Is a Spectrum
Reading about the difficulty of getting many virtual worlds to agree on a tree, or a pair of shoes, or the means of walking up to a tree to cut it down and sell it as a Christmas tree three virtual worlds over, you may be asking whether we can reasonably expect a meaningfully interoperable Metaverse to exist at any future point. The answer is yes, but it requires nuance.
Most apparel is interoperable in the real world. All belts, for example, are presumed to work with all pants. Exceptions exist, of course, but overall, most belts are compatible with most pants, irrespective of the year you bought the belt, the brand of belt, or what country you were in when you bought it. At the same time, not all belts fit all pants equally well. There are common standards for pants and belts, but a 30 × 30 pant from J.Crew fits differently from a 30 × 30 pant from Old Navy (dresses vary even more; European and American shoe size standards are entirely different; and so on).
Globally, many differing technical standards exist, such as those for residential voltage, and for measurements of speed, distance, or weight. In some instances, new equipment is required for a foreign device to be used (for example, an electrical outlet adapter), and in other instances a local government will require replacements, such as a car’s exhaust to be replaced to meet local emissions regulations.
Pants work everywhere, though not every location you wish to visit will admit jeans. Movie theaters allow almost any clothes and most forms of credit, but you can’t bring in outside food nor drink. One can carry a shotgun in much, but not all, of the American outdoors, but rarely in cities and almost never in a school. Cars work on all roads in the US, but to drive on a golf course you need to rent a golf cart (even if you own one). Not every business accepts every currency, but currencies can be exchanged for a fee. Many stores support some, but not all, credit cards and a few accept none. Most of the world now embraces trade, but not all of it, nor for all things, in all quantities, or for free.
Identity is even more complicated. We have passports, credit scores, school records, legal records, employer IDs, state IDs, and more. Which of them are used for what, which of them are available to outside parties or can be affected by outside parties, all varies—sometimes based on where a person is at a given time.
The internet is not much different. There are still public and private networks (and even offline ones), as well as networks, platforms, and software that admit the majority of common file formats, but not all. While the most popular protocols are free and open, many are paid and private.
Interoperability in the Metaverse is not binary. It is not about whether virtual worlds will or won’t share. It’s about how many share, how much is shared, when, where, and at what costs. So why am I optimistic that, given all these complications, there will be a Metaverse? Economics.
Start with the question of user spending. Many Metaverse skeptics pose some version of the question, “Who wants to wear Fortnite’s Peely skin while playing Call of Duty?” Now to be fair, a giant, comically styled anthropomorphic banana doesn’t make much sense in Call of Duty, or in a virtual classroom, for that matter. But it is equally obvious that some users want some items, such as a Darth Vader costume, a Lakers jersey, or a Prada purse, in many different spaces. And they certainly don’t want to buy these items over and over and over again. They might be reluctantly willing to do so today, but that’s because we’re still in the early stages of the shift to virtual apparel. In 2026, hundreds of millions of people will be sitting on numerous (effectively) duplicated outfits across their many previously played games—and will doubtlessly resist buying those outfits again. Liberating purchases from a single title will, the theory suggests, lead to both more purchases and higher prices.
Put another way, would Disney sell more or less merchandise if it could be worn or used only in its theme parks? How much would someone pay for a Real Madrid jersey that could only be used in Santiago Bernabéu Stadium? Or how much lower would user spending be on Roblox if a player’s outfit was limited to a single Roblox game?
It’s likely that consumer spending today is constrained by the very knowledge that no game lasts forever. Think of anything you might buy on holiday but don’t plan to bring home in your suitcase—a boogie board, a stainless steel water bottle, a costume for Día de los Muertos. Expected obsolescence always constrains our spending.
The utility of these goods is further limited by ownership restrictions. Most games and gaming platforms prohibit users from giving outfits or items to other users, or even selling them for in-game currency. The publishers that do allow reselling and trading typically place firm limits on this activity. Roblox Corporation only allows “limited items” to be resold (otherwise peer-to-peer trading would undermine the sale of goods from Roblox’s own shop)—and only Roblox Premium subscribers can sell these items.
What is more, although we might believe we have “purchased” these items, we’ve really just licensed them and the company can “repossess” them at any time. This isn’t a huge problem for $10 skins and dances, but no one will buy $10,000 worth of virtual property that could be taken away from them at any moment, with or without repayment.
Consider a case from early 2021, as reported by South China Morning Post’s Josh Ye. Tencent, China’s largest gaming company, “sued a game item trading platform to determine who owns the in-game currency and items.” Specifically, the company argued that these assets had “no material value in real life” and that in-game coins bought with real money were “effectively services charges.”1 The result was outrage, with many gamers feeling mistreated and/or demeaned.
Ownership rights are foundational to investment and the price of any good, while the opportunity for profit is a well-established motivator. Speculation has always financed the growth of new industries, even when it results in bubbles (a lot of America’s now cheap-to-use fiber optic cabling was laid in the run-up to the dotcom crash). If we want the greatest possible investment of time, energy, and money into the Metaverse—if we want to achieve the Metaverse—we need to establish firm ownership rights.
Every stakeholder in virtual worlds faces incentives and risks that point in this direction. It’s dangerous for any developer to build a business whose wares or services are limited by the popularity of a given platform or its economy (or economic policies). And anything that results in less investment and thus fewer and worse products overall doesn’t benefit the developer, the user, or the game and its platform.
Limiting the range of identity and player data is another impediment to the Metaverse economy. Toxicity in gaming is a significant concern for many, and rightfully so. Today, however, while Activision might ban a player from Call of Duty for abusive or racist language, that player can then go on to troll on Epic Games’ Fortnite (or on Twitter, or Facebook). The player could also just create a new PlayStation Network Account, or change to Xbox Live, and while that means fragmenting his or her achievements, some of these achievements are locked to a given platform anyway. Of course, publishers don’t want to make their competitors’ games better, nor are they usually inclined to share their play data. But no gaming company benefits from toxic behavior, and everyone is negatively affected by it.
Economics, then, will drive standardization and interoperation over time.
The Protocol Wars offer an illustrative example. From the 1970s to the 1990s, few believed that the many competing networking stacks would be replaced by a single suite, least of all one steered by nonprofit and informal working bodies. Instead, we’d contend with a “cyberspace divided.”
Banks and other financial institutions didn’t use to share credit data, either—it was considered far too valuable and privileged. But eventually, they were convinced that credit scores with better data and more coverage would be of collective benefit. Competing homestay marketplaces Airbnb and Vrbo are now partnering with a third party to prevent guests with a history of poor behavior from making future bookings. Although this harms the individual offenders, all other guests, hosts, and platforms benefit.
The best example of “economic gravity” comes from the game engines—the very companies pioneering the plumbing of the Metaverse. Although the opportunity in virtual worlds has never been as large as it is today, reaching the entirety of this market has never been so difficult. In the 1980s, a developer might make a game for just one console and in doing so, reach 70% of potential players. Two developers might reach all players. Today, there are three console makers, two of which operate consoles in two different console generations, plus cloud-based consoles that use their own technology stacks, namely Nvidia’s GeForce Now, Amazon’s Luna, and Google’s Stadia. There are also two PC platforms, Mac and Windows, that span dozens to hundreds of different hardware builds, and two dominant mobile computing platforms, iOS and Android, that span many more OS versions, GPUs, CPUs, and other chipsets. Every additional platform, device, or build requires code that is unique to a specific set of hardware, or that is written to work across many and without genericizing performance to the lowest common denominator. Creating and supporting all this code is costly, time consuming, and arduous. Another option is to just ignore a large part of the market, which is also expensive.
Network Gateways: A Cyberspace Divided. In this map, the major computer networks huddle in the mass of the Matrix, the term for the global collection of computer networks that can exchange electronic mail. The Internet serves as a common ground for much communication online, with commercial online services building gateways for electronic mail as well as other communication and data protocols to the Internet. Major national services such as France’s Minitel (http://www.minitel.fr/) now provide gateway communication from their services to the Internet.
Image 3. Telegeography Map This 1995 map and its caption reflect what many experts at the time considered to be the future of online networking: fragmentated networks and protocol suites. The internet, in this case, would not be a unifying internetworking standard, but more of a common ground for different collections of networks, some of which would be unable to directly communicate with one another. Most of these networks would exist in “The Matrix,” though some would forever sit outside of it. But this future never happened. Instead, the internet became the core gateway between all private and public networks, thereby enabling each network to communicate with any other. TeleGeography
This challenge, when combined with the ever-growing complexity of virtual words, is why cross-platform game engines such as Unity and Unreal have proliferated. They emerged as a response to fragmentation, and they don’t just solve it, they do it at a low cost and to everyone’s benefit—even that of the most entrenched platforms.
Imagine a developer decides to build a new game for iOS. Apple’s mobile ecosystem has 60% of smartphone share in the United States and 80% share among teenagers, and over two-thirds of mobile gaming revenues globally. In addition, a developer can reach nearly 90% of iOS users by writing to only a dozen iPhone SKUs. The remainder of the global market is split between thousands of different Android devices. Forced to choose between these two platforms, a developer would always pick iOS. But, by using Unity they can easily publish their game across all platforms (including web), thereby increasing their revenue potential by over 50% at little incremental cost.
Apple might prefer more exclusive games, and games fully optimized for their hardware, but it’s better for everyone, including iOS users and the App Store, that most mobile developers use Unity. By making more money, developers can build more and better games, thereby driving even more user spending on mobile devices.
The proliferation of cross-platform game engines such as Unity and Unreal should also make it easier to bring together the many fragmented virtual worlds operating today into a unified Metaverse. In fact, this has already been proven. For more than a decade after online console gaming emerged, Sony refused to support cross-play, cross-purchasing, or cross-progression between games played on its PlayStation and other platforms. Sony’s policy meant that even if a developer created versions of their game for both PlayStation and Xbox, and two friends bought copies of that same game, they could never play together. Even if a single player bought two copies of the same game (say, one for their PlayStation and another for their laptop), their in-game currencies and many of their rewards would remain siloed in one or the other.
Critics of the policy argued that Sony’s stance was a consequence of its dominant market position. The first PlayStation outsold the runner-up console, Nintendo 64, by 200%, and Xbox by over 900%. The PlayStation 2 sold 550% more than the Xbox and Nintendo GameCube combined. The PlayStation 3 barely beat the Xbox 360, largely due to the Xbox’s early innovations in online games, and lost to the Nintendo Wii, but by the mid-2010s, the PlayStation 4 had doubled sales of the Xbox One and quadrupled those of the Wii U.
As a result, it seemed that PlayStation saw cross-platform gaming as a threat. If users didn’t need a PlayStation to play with other PlayStation users—the majority of console gamers—they’d be less likely to buy a PlayStation in the first place, and PlayStation users might even churn away to competitors. Sony’s president of interactive entertainment tacitly admitted as much in 2016, stating “the technical aspect could be the easiest” part of opening up access to its PlayStation Network for cross-play.2 Yet only two years later, PlayStation enabled cross-play, cross-purchasing, and cross-progression. Three years after that, nearly every game that could support this functionality offered it.
Sony didn’t change its mind because of internal preferences, business models, or pressures. Instead, it did so in response to the success of Fortnite, which came from a company, Epic Games, that not coincidentally focused on cross-platform gaming.
Fortnite had a number of rare attributes when it launched. It was the first mainstream AAA game† that could be played on nearly every major gaming device globally, including two generations of PlayStations and Xboxes, the Nintendo Switch, Mac, PC, iPhone, and Android. The title was also free, which meant players didn’t have to buy multiple copies to be able to play on multiple platforms. Fortnite was also designed as a social game; it became better as more of your friends used it. And it was built around live services, rather than a fixed narrative or any offline play: the game’s content never ended and was updated as often as twice per week. This, plus superb creative execution, helped Fortnite become the most popular AAA game globally (excluding China) by the end of 2018. It was generating more revenue per month than any game in history.
Sony’s gaming competitors all embraced cross-platform services for Fortnite. PC and mobile had never blocked cross-platform functionality; neither Windows nor any mobile platform had ever bought exclusive games. Nintendo supported numerous cross-platform services for Fortnite from the start, too—but, unlike Sony, it had no real online networking business and didn’t prioritize it. Microsoft, for its part, had long pushed for cross-play (likely for the same reason Sony resisted it).
The lack of cross-platform integration meant that PlayStation not only had the worst version of Fortnite, but PlayStation owners had many better versions at their fingertips and didn’t need to pay a dollar to use them. This fundamentally changed Sony’s thinking. Denying such a capability for titles like Call of Duty might have had a modest impact on the number of copies sold by Activision, but with Fortnite Sony was missing out on most of the game’s revenue and driving PlayStation players to competing platforms. Sure, PlayStation offered a better technical experience than the iPhone, but most players considered the game’s social elements to be more important than its specs. And Epic “accidentally” activated cross-play on PlayStation, allegedly without Sony’s permission, on at least three occasions—thereby rallying even more upset users to petition Sony for change, and proving the impediment was policy, not technology.
All of these factors forced Sony to change its policies. This has obviously been for the good of all. Today, a number of hit games can be accessed by nearly all computing devices globally (and thus could be played by anyone, anywhere, anytime), without users needing to repay or fragment their identity, achievements, or player networks. Furthermore, cross-platform play, progression, and purchasing means every console competes on hardware, content, and services. Sony still thrives, too: PlayStation drives over 45% of total Fortnite revenues (and the PlayStation 5 has outsold the Xbox Series S and X by a ratio of more than 2:1).3
Crucially, Sony’s decision to open up its closed platform also provides a view into potential economic solutions to the challenge of interoperability. In order to avoid “revenue leakage,” Sony required Epic to “true up” its payments to the PlayStation store. For example, if a Fortnite player spent 100 hours playing on PlayStation and 100 hours on Nintendo Switch, but spent only $40 on PlayStation compared to $60 on Nintendo Switch, Epic would have to pay Nintendo a 25% commission on its $60, but then pay PlayStation 25% on its $40 and the $10 its share of time would suggest it was owed. In other words, Epic pays twice on that $10. It’s not clear whether this policy is still in place—the public only knows it exists because of Epic’s lawsuit against Apple. Regardless, the model is an example of how the proliferation of cross-platform gaming aids all market participants.
The success of Discord is another good example. Historically, gaming platforms such as Nintendo, PlayStation, Xbox, and Steam tightly protect their player networks and communication services. This is why someone on Xbox Live cannot “friend” someone on PlayStation Network, nor speak to them directly. Instead, users on other platforms are only available inside cross-platform games, such as Fortnite, and via their game-specific IDs. While this approach worked well enough when two players knew which single game they wanted to play before logging on, it didn’t work well for unplanned hanging out, or enjoying ad hoc play. The more central gaming was to someone’s lifestyle, the less this solution suited them.
Discord emerged to meet this demand, and it has offered gamers numerous benefits. It operates across all major computing platforms—PC, Macs, iPhones, and Androids—meaning every gamer can access a single social graph (and non-gamers can join, too). The service also presents gamers with a rich suite of APIs that can be integrated into other games and even quasi-competitive social services, such as Slack and Twitch, as well as stand-alone games it doesn’t distribute or otherwise operate. Discord has been able to build a gamer communication network larger—and far more active—than any single immersive gaming platform.
Importantly, there was no way for the platforms to stop users from using the Discord apps on their phones and using its chat features, in particular. Discord’s success drove both Xbox and PlayStation to eventually announce native integrations of Discord into their closed platforms—a move that created a new “interchange” solution for their player networks, communication services, and socializing online.
Establishing Common 3D Formats and Exchanges
The standardization of game engines and communications suites is fairly complex compared to how 3D-objects conventions will emerge.
Witness the current universe of 3D assets. Billions of dollars have been spent on non-standardized virtual objects and environments across film and video games, civil and industrial engineering, healthcare, education, and more. There are no signs that this level of spending will do anything but increase in the near future. Constantly remaking these objects for a new file format or engine is financially impractical and often wasteful; the greatest attribute of a digital “thing” is that it can be endlessly re-used without additional cost.
Interchange solutions are already emerging to tap into the “virtual gold mine” of previously created and fragmented asset libraries. A good example is Nvidia’s Omniverse, which launched in 2020 and enables companies to build and collaborate in shared virtual simulations built upon 3D assets and environments from different file formats, engines, and other rendering solutions. An automotive company might be able to bring its Unreal-based cars to an environment designed in Unity and have those cars interact with objects made in Blender. Omniverse doesn’t support all possible contributions, nor all metadata and functionality, but because of this, it provides independent developers a clearer reason to standardize. Collaboration, meanwhile, leads to formal and informal conventions. Notably, Omniverse is built on Universal Scene Description (USD), an interchange framework developed by Pixar in 2012 and open sourced in 2016. USD provides a common language for defining, packaging, assembling, and editing 3D, with Nvidia likening it to HTML, but for the Metaverse.4 In short, Omniverse is driving both an interchange platform and a 3D standard. Helios, the proprietary real-time rendering engine used by the visual effects services company Industrial Light & Magic, is another good example, as it is compatible with only select engines and file formats.
As 3D collaboration grows, standards will naturally emerge. By the early 2010s, for example, globalization had led many of the world’s largest corporations to mandate English as their official corporate language—this included Rakuten, Japan’s largest e-commerce company; Airbus, an aerospace giant which counts the governments of France and Germany as its two largest shareholders; Nokia, the fourth-largest company in Finland; Samsung, South Korea’s largest company, and more. A 2012 poll by Ipsos found that 67% of individuals whose work involved communications with people based in other countries preferred to conduct this work in English. The next closest language was Spanish, at 5%. Crucially, 61% of respondents said they did not use their native language when working with foreign partners, thus the alignment on English was not a reflection of the fact most respondents were primarily English speakers.5 Globalization has also led to de facto standards in currencies (namely the US dollar and the Euro); units (e.g., the metric system); exchange (the intermodal shipping container); and so on.
Crucially, as Omniverse showed, software does not need everyone to speak the same language. Instead, think of it as comparable to the system within the European Union, which has 24 official languages represented but three (English, French, German) “procedural” languages which are prioritized (also, much of the EU’s leadership, parliament, and staff can speak at least two of these languages).
Epic Games, meanwhile, is working to pioneer data standards that allow a single “asset” (really, a right to data) to be repurposed in multiple environments. Not long after acquiring Psyonix, Epic Games announced that the publisher’s hit game Rocket League would go free-to-play and move over to Epic Online Services. A few months later, Epic announced the first of several “Llama-Rama” events. These limited-time modes enabled Fortnite players to complete challenges in Rocket League that would unlock exclusive outfits and achievements that could be worn in and across either game. A year later, Epic bought Tonic Games Group, makers of Fall Guys and dozens of other games, as part of its investments “in building the Metaverse.”6 It’s likely Epic will extend its Rocket League experiments across Tonic’s titles, as well as those coming from its Epic Games Publishing, which finances and distributes games from independent studios.
With its cross-title assets and achievements model, Epic is likely looking to set a similar precedent as those the company established in cross-platform gaming. Epic clearly believes that there are benefits—that is, profits—to reducing the friction to accessing different games, making it easier to bring your friends and items across these games, and giving players a reason to try out new games. Players will then spend more time gaming, with more people, across a greater diversity of titles, spending more money along the way. If so, an ever-expanding network of third-party games will want to connect to Epic’s virtual identity, communications, and entitlements systems (i.e., parts of Epic Online Services), thereby driving standardization around Epic’s various offerings.
Alongside Epic are a series of other socially focused software giants looking to use their reach to establish common standards and frameworks for shared virtual goods. One clear example comes from Facebook, which is adding “interoperable avatars” to its Facebook Connect set of authentication APIs. Facebook Connect is better known to the public as “Log in with Facebook,” which allows Facebook users to substitute their Facebook log-in for a website or app’s own account system. Most developers would prefer that people create a bespoke account, as it provides the developer with greater information about the user, and means the developer controls this information and account (and not Facebook). However, Facebook Connect is far simpler and faster, and therefore it is the preferred solution for most users. As a result, developers benefit from more registered users (versus anonymous ones). A similar value proposition will exist for Facebook’s avatar suite (or perhaps, those of Google or Twitter or Apple). If customized avatars are essential to user expression in 3D space, then few users will want to create a new, detailed avatar for every virtual world they use. The services that accept the avatars a user has already invested in will be able to offer a better experience for said user. Some people even argue that the inability to use a consistent avatar means that no avatar can truly represent the user—just as we wouldn’t say that Steve Jobs had a uniform if he could only sometimes wear jeans and a black mock turtleneck, and occasionally needed to wear chambray pants and a gray turtleneck depending on the venue. That’s an aesthetic, rather than a uniform intended to reinforce your identity. Regardless, the establishment of cross-title services such as Facebook will serve as another de facto standardization process (in this case, based on Facebook’s specifications and advanced by its AR, VR, and IVWP initiatives).
In addition to driving asset interoperability, Epic is also driving the “interoperation” of competing intellectual properties, which is a philosophical, not technical, problem (cross-platform gaming reminds us this is the harder of the two challenges). As virtual platforms like Fortnite, Minecraft, and Roblox grew into culture-driving social spaces, they’ve become an increasingly necessary part of consumer marketing, brand building, and multimedia franchise experiences. In the past three years, Fortnite has produced experiences with the NFL and FIFA, Disney’s Marvel Comics, Star Wars and Alien, Warner Bros.’ DC Comics, Lionsgate’s John Wick, Microsoft’s Halo, Sony’s God of War and Horizon Zero Dawn, Capcom’s Street Fighter, Hasbro’s G.I. Joe, Nike and Michael Jordan, Travis Scott, and more.
But to participate in these experiences, brand owners must embrace something they almost never allow: unlimited-term licenses (in-game outfits are kept by players forever), overlapping marketing windows (some brand events are mere days apart or overlap entirely), and little to no editorial control. In sum, this means it’s now possible to dress as Neymar while wearing a Baby Yoda or Air Jordan backpack, holding Aquaman’s trident, and exploring a virtual Stark Industries. And the owners of these franchises want this to happen.
If interoperability truly has value, then financial incentives and competitive pressure will eventually solve for it. Developers will eventually figure out how to technically and commercially support Metaverse business models. And they’ll use the Metaverse’s larger economy to surpass “legacy” game makers.
This is one lesson of the rise of free-to-play game monetization. In this business model, players are charged nothing to download and install a game—or even to play it—but are presented with optional in-game purchases such as an extra level or a cosmetic item. When it was first introduced in the 2000s, and even a decade later, many believed free-to-play would, at best, lead to lower revenues for a given game and at worst cannibalize the industry. Instead, it proved to be the best way to monetize a game and a core driver behind video gaming’s cultural ascendance. Yes, it led to many non-paying players, but it substantially increased the total number of players and even gave paying players a reason to spend more. After all, the more people you can show a customized avatar off to, the more you’ll pay to do so.
Just as free-to-play led to new products to sell to players, from dances to voice modulators and “battle passes,” interoperability will too. Developers might bake degradation into an asset’s code—this skin works for 100 hours of use, or 500 games, or three years, during which time it slowly wears out. Alternatively, users might have to pay an additional fee to bring an item from one publisher’s title to a competing publisher’s (just as many goods have import duties in the “real world”) or pay more in the first place for an “interoperable edition.” Not all virtual worlds will move to widely interoperable model, of course. Despite the prevalence of free-to-play multiplayer online games today, many titles are still paid, single-player, offline, or all three.
Web3-focused readers might be wondering why I’ve not yet addressed blockchains, cryptocurrencies, and non-fungible tokens. These three interrelated innovations look likely to play a foundational role in our virtual future and are already operating as a sort of common standard across an ever-expanding series of worlds and experiences. But before examining these technologies, we must first examine the role of hardware and payments in the Metaverse.
* In Unity, the y-axis in an x/y/z coordinate system for a virtual object refers to up/down, while Unreal uses the z-axis for up/down and maps the y-axis to left/right. Converting this information is easy for software, but the disagreements on such foundational data conventions help us to understand how different conventions are between engines.
† “AAA” is an informal classification for video games with large production and marketing budgets and which usually come from the largest video game studios and publishers. It is similar to the “blockbuster” designation in the film industry. Neither term means the title is a financial success.