The Promise and Peril of Fully On-Chain Games

Fully on-chain games (FOCGs) have emerged as an exciting new frontier in the realm of blockchain gaming, captivating researchers, developers, and investors with their alluring value propositions, such as composable modding and smart contract-powered social contracts, among others. Yet, the technical implementation of FOCGs brings forth a distinct set of challenges that warrant meticulous exploration as we delve into the promises they offer.

Background

To grasp the intricacies of FOCGs, it is imperative to first understand the concepts of game state, game ticks, and tick rate.

Game State

Is a snapshot of all relevant information and data that describes the current state of a game at any given moment.
It includes various aspects such as positions and properties of game objects, player scores, health points, game progress and other variables that define the current state of the game.
The game state is typically updated and modified over time in response to inputs by players or other events.

Game Ticks

Also known as frames or updates, are discrete units of time that are used to measure the progression of a game.
Each game tick represents one iteration or cycle of the game’s main loop, during which the game logic and physics calculations are performed.
Each tick typically updates the game state based on the current input, processes AI behaviour, handles collisions, and performs any other necessary calculations to advance the game.

Tick Rate

Is the frequency at which game ticks occur per second.
It indicates the number of times the game logic and rendering are updated and refreshed within a second, often measured in Hertz (Hz) or Frames per second (FPS).
For example, a tick rate of 60 Hz means that the game logic is updated 60 times per second. A higher tick rate generally leads to smoother gameplay and more responsive controls, but it also requires more computational resources.

Web 2.0 Games

Under the Web 2.0 gaming paradigm, all game components including game state, logic and data are stored off-chain. It is managed natively by centralised servers with robust computational and networking capabilities to handle real-time gameplay and synchronisation.

Over the years, the combination of efficient game logic coding, hardware advancements, and optimisation techniques has enabled the development of high tick rate games.

Web 2.5 Games

Web 2.5 games represent a more decentralised form of gaming through the storage of in-game assets on-chain. These games often involve the launch of an ERC-20 game token and the tokenisation of in-game assets such as skins into NFTs. This addresses asset ownership issues in traditional games as players are free to own and trade these assets fully on-chain.

However, game state, game ticks, and tick rate operate isomorphically to Web 2.0 games. The lack of native game ticks on blockchains is one of many inherent structural and computational limitations of blockchains that impede the on-chain implementation of games. As such, computation heavy components are all handled off-chain in Web 2.5 games.

One issue that persists in both Web 2.0 and 2.5 games is the lack of transparency and censorship resistance. Game logic such as item rarity, drop rates, and true randomness are often not openly disclosed and can’t be provably fair. Players must trust the claims and decisions made by the game developer without the ability to independently verify them. Furthermore, a centralised entity (such as a gaming studio) possesses the authority to modify, censor, or control the game, and retains the ultimate decision-making power over the game’s termination, which can be technically executed at any time or as required.

FOCGs

FOCG involves putting the entire game logic and state and data storage on-chain. Blockchain is used as a decentralised game server where players participate and contribute to a shared game state trustlessly.

Client-agnostic FOCGs can be built with multiple frontend operators that all interact with the same game logic. This approach is analogous to Liquity’s decentralised frontends, enabling the same game logic to be displayed with various aesthetics and UI elements. Third party developers can build mods that monetise through creating characters as NFTs or bootstrapping economies with ERC-20s that interact with the underlying game logic.

By putting not only game assets but also the game state and loops on a credibly neutral blockchain, censorship and platform risks can be mitigated. This allows for on-chain games to scale safely, particularly those with high stakes.

However, as mentioned above, the absence of a native game tick on blockchains is a significant limitation. All blockchains operate with block confirmation times, which introduce latency issues. Depending on the blockchain, the time it takes for a transaction to be validated and included in a block can range from seconds to minutes. Consequently, early FOCGs were blockchain-based games with asynchronous gameplay elements, i.e. turn-based TCGs.

For other game genres, such as RTS games that operate synchronously with immediate or near-immediate feedback, scaling solutions become necessary. Some examples include:

Argus’ World Engine – a sharded layer 2 blockchain SDK with game shards that act like a high-performance game server purpose-built for handling in-game computations. This is achieved by decoupling game execution (game shard) from smart contract execution (EVM shard). Their first game shard implementation – Cardinal has achieved a higher tickrate, at 20 blocks per second, and can integrate with existing game engines like Unity directly.
Curio Research – building a customised ticking chain implementation on OP stack (using Caldera as RaaS). A high-throughput ECS based game engine that can support their upcoming RTS game.
Dojo – a “provable” game engine built on Starknet that enables game sessions to be run off-chain, reducing the need to record every player action on-chain. Instead, proofs can be generated at specific intervals to validate the correctness of previous actions within that interval, which are processed off-chain.

Lastly, the effective implementation of ZKPs (i.e. Dark Forest using zk-SNARKs) is also essential for incomplete information games, such as those with fog of war, to be built fully on chain while guaranteeing fairness.

The Promise of FOCGs

Composable Modding

We are currently witnessing a new era in gaming where players desire the ability to personalise, modify, and seamlessly integrate their own identity into games, requiring a level of customisation that can only be achieved through UGC or modding. The popularity of mods in Web 2.0 games like Minecraft and Roblox exemplifies this trend.

In these games, a new category of “players” emerge: modders who actively contribute to the growing library and ecosystem of mods. In Roblox, modders can monetise their content in Robux through game passes and/or in-game purchases. The returns earned from building and operating successful mods can be compared to those building a standalone game. To put it in perspective, Uplift Games, the independent studio behind Adopt Me!,employs roughly 40 people and earns 60 million dollars a year.

This dynamic creates a self-reinforcing flywheel with substantial network effects. As modders continue to create new games and content, players become more immersed, and new users are attracted through social networks. As the player base expands, the incentive and feedback system of mods encourages more players to become modders. The symbiotic relationship between modders and players, combined with intuitive developer toolings is driving the rise of distinct virtual worlds, experiences and communities within a single game.

However, mods in traditional games largely exist in isolation and fragmentation, lacking universal frameworks for meaningful interactions. While this may be intentional, new inputs and contributions from modders frequently remain confined within isolated virtual worlds. Even when cross-mod interactions do occur, the impact on gameplay tends to be minimal, such as the ability to wear the same skin (“allowed gear” in Roblox) across different mods.

The potential for composable modding stands as a significant value proposition for FOCGs. Upcoming FOCG projects are taking various approaches to their tech stack, but a common approach is to allow third-party developers to directly interact with the game logic through smart contracts, allowing them to:

Create new mods (with mod-specific token economy)
Extend games and mods with new functionality (such as adding new items, characters and maps)

It is conceivable that if a community in Mod A introduces a new resource, it should be directly compatible with Mod B’s world, allowing players there to grow, harvest and trade it.

This incentivises modders to consider not only a single mod but also all relevant mods when designing new features. Composable modding expands the potential for monetisation and the dissemination of best practices. However, for mods to be this composable on-chain while ensuring consistent gaming experience, modders would likely have to operate within a limited design space compared to traditional games.

Smart contracts as social contracts

In traditional games, players, groups and guilds relied on informal and unenforceable agreements for interactions that went beyond the predefined in-game mechanics. This approach posed several limitations, especially in high-stakes games where the consequences of broken agreements could be significant.

In 2014, a large-scale battle known as the “Bloodbath of B-R5RB” took place that involved thousands of Eve Online players. Prior to the battle, alliances had made informal agreements and treaties to cooperate against a common enemy. However, during the battle, one alliance unexpectedly turned on its allies, betraying the agreements and causing over $300,000 losses in game assets.

Treaties as explored by Curio Research, highlights the significant potential of smart contracts in facilitating complex social interactions among players or groups. With game-specific smart contracts like Treaties, agreements can be formalised between players and can be trusted to execute when predetermined conditions are met.

A treaty is many things. It can be a NATO-styled agreement where upon joining and membership payment you’re forbidden from attacking other committee members. It can be a national bank that issues USDC-backed tokens, forcing your trading partners who’ve signed to strictly use your national currency next time they swap in an in-game AMM.
Kevin Z from Curio:

The customisability and modularity of these in-game smart contracts could extend social dynamics beyond the boundaries of games and allow for direct interaction with other on-chain ecosystems, such as DeFi protocols.

The Impediments to FOCGs

Undoubtedly, FOCGs have the potential to enable novel models of modding and social interactions on-chain. However, a series of open problems arise as we take a closer look at these promises.

1. Technical Constraints

Besides the absence of native on-chain game ticks, there are various technical constraints to consider:

High gas costs and low TPS – bottlenecks of most mainnet chains make it challenging to support synchronous games and concurrent game session/instances.
Vulnerability of on-chain randomness – validators can see the result of randomness before its confirmed on-chain, creating room for manipulation (requires VRF/ZKP solutions).
Botting – decentralised game entities have limited ways to regulate and enforce anti-cheating measures. Bots can interact with the games in the same way as human players, without any inherent restrictions.
Unintuitive UIUX for Web 2 gamers – using dApps often require wallet creation, transaction signing, bridging of assets etc. It is important to streamline this process (i.e. wallet account abstraction) to maximise the onboarding of non-crypto native gamers.
Security – when FOCGs have assets with real values that can interact with DeFi, exploits become irreversible and can permanently impact the FOCG ecology at large.

2. The Costs of Composability

The promise of composable modding introduces certain trade-offs, especially when we consider how the on-chain implementation could hinder the very success factors that made modding desirable and successful in the first place.

For Players

Successful Roblox mods like Jailbreak, Arsenal, and Adopt Me! have garnered a massive following, accumulating a combined total of over 40 billion visits. Each of these mods offers distinct gaming mechanisms, rules, environments, and in-game assets:

Jailbreak immerses players in a prison escape and police chase scenario.
Arsenal provides a FPS experience reminiscent of Call of Duty.
Adopt Me! revolves around the virtual adoption and care of pets.

From a player’s perspective, the appeal of these mods largely lies in their distinct gameplay mechanics, story telling, and thematic elements. Players can find playing Roblox comparable to logging onto Steam, as both offer a diverse selection of game genres and experiences to explore.

There have been community-driven attempts to make mods more interactive, such as in the case of “The Pixelmon Modpack,” where a bundle of pre-configured Minecraft mods can be downloaded. However, these mods still remain isolated from other modpacks and are run on local servers. Nevertheless, the current player demand for such interactive mods does not appear to be very high and the lack of official frameworks from games with major modding ecosystems reflects this.

For Modders

Physics and Game Balance

In a permissionless modding ecosystem, achieving composability requires careful consideration of game physics and game balance, which are vital for ensuring a fair and enjoyable gameplay experience.

In traditional games, modders have the freedom to modify game physics according to their desired game mechanics because the mods they create do not necessarily need to interact with other mods. For instance, while there are fundamental physics principles consistent across the world of Roblox, each mod has the flexibility to define and implement its own set of physics rules. Jailbreak has specific physics related to vehicle handling and collisions, while Adopt Me! has its own distinct physics systems related to pet movements. Roblox can attract and accommodate a diverse range of modders with expertise in various genres, such as FPS, horror and casual games because of this flexibility.

For mods to be truly composable, it would necessitate consistent game physics across all mods. As new items, economies, and systems are permissionlessly added into the game , game balance considerations bring additional inflexibility. For example, FOCGs would need to have careful calculations and frameworks in place to prevent new characters or strategies from becoming overpowered. Moreover, updating new inputs across the entire tech stack, including front-ends and relevant contracts, would entail complex coordination.

One can imagine that fully composable modules will create chaos in this regard and restrict new inputs to merely aesthetic variations, rather than allowing inputs that can impact gameplay outcomes. Such conformity could hinder the diversity of modders and converge to a homogenised modding ecosystem.

A permissioned modding ecosystem could alleviate some of these problems, but that would contradict the fundamental principles of the Autonomous World postulation that FOCG should adhere to.

Genre-specific Game Logic and Infrastructure

Different genres require different tick rates, which renders it impossible to build new mods for genres that rely on a higher tick rate than what the FOCG infrastructure can currently accommodate.

Hence, most upcoming FOCGs focus on initially building game-specific infrastructure before attempting to make it more general-purpose as they scale.

There is a noticeable pattern amongst these teams – nearly all of them are focusing on building infrastructure to support strategy games (Dark Forest, Treaties, Primodium etc.). On one hand, the strategy genre can be relatively easier to implement and realise more benefits from being fully on-chain, as demonstrated by Treaties. On the other hand, a genre-specific tech stack could limit player and modder diversity, curtailing the full potential of composable modding.

Should Games be Fully On-Chain?

In an ideal scenario where the value proposition of FOCGs, such as composable modding, proves highly valuable for players and modders, traditional game studios could incorporate these elements into their off-chain or hybrid systems, benefitting from them without being restricted by design space limitations.

For instance, tokenising in-game assets as NFTs and incorporating smart contracts for enforceable social contracts. With the remaining components to ran off-chain, including a modding ecosystem redesigned with proper frameworks in place to guarantee composability.

The Case for FOCGs

While that is indeed plausible, I remain open-minded that FOCG could out compete these hybrid games in specific niches. I believe the strongest case for FOCG are high stake games, driven by a player base that demands not only in-game assets and credentials but also the entire game logic to be on-chain.

To illustrate this point, take Marvel Heroes shutting down as an example. It was a free-to-play online action RPG featuring characters from the Marvel Universe. In November 2017, the game suddenly announced its shutdown and received a wave of demands for refunds. However, with clever legal technicalities, consumers found themselves without a straightforward entitlement to compensation nor access to purchased in-game assets.

“You agree that neither Gazillion nor any of its affiliates, licensors, agents, or employees is liable to you or any third party for any revision, supplement, suspension, or discontinuation of the System, and termination of any license. “Gazillion, in its sole discretion, may remove or delete any of your Content when such Content has exceeded a pre-determined time restriction and/or quantity, or when the System, or any component thereof, requires maintenance or upgrades.”
@GIBiz

Merely having assets on-chain does not completely solve this problem in the event of a game studio rug pull, as the game state and progress would still be lost, rendering those assets useless within the game. Furthermore, in such an event, the studio is under no obligation to open-source the game. FOCGs solves this problem.

In-game mechanics and outcomes being provably fair is also significant. Imagine a group of players engaged in a battle against a common enemy, where a high-value item drop is at stake for the highest damage dealer. Or a 4x RTS game where spawn location and proximity to resources are randomised. In these scenarios, FOCG can guarantee the outcome of the battle or randomness to be provably fair. This trustless environment would not only benefit the players but also speculators who place bets via smart contracts on game outcomes. Ensuring unmanipulated game results and permanence (as game state and logic can be forked) would allow for high stake game ecosystems and assets to truly scale without any platform risks.

I believe several other elements need to be further explored and aligned. What I am excited to see:

Further exploration of the practicality of on-chain true randomness. If not, the speeding up and scaling off-chain randomness solutions.
Effective implementation of ZKPs for provably fair and verifiable events, sequences and data.
Architectural optimisation that better enables concurrent game sessions/instances without sacrificing composability.
Game mechanics and social dynamics that can interact directly with adjacent on-chain ecosystems. Economic systems within or derived from high-stakes FOCGs would greatly benefit from composability with DeFi functionalities.
Tokenomics that better align incentives for players and modders. As the line between developers and players becomes more blurred, new business models may emerge, similar to how the rise of mobile games led to the widespread use of the free-to-play model.
Decentralised game balance governance structures to accommodate permissionless modding.

I am hopeful that new primitives can emerge from games built fully on-chain, enabling a new paradigm of gaming that harnesses the complete range of benefits offered by blockchain – instead of just partial ones.

Acknowledgements

Thanks Erik Lie, Richard Yuen, Jonathan Yuen, Adrian Chow and Harry Lam for help with drafts. Special thanks to my little brother, Nathan, who is a Roblox enthusiast and answered all my questions about the game.

References

Disclaimer: the inclusion of links to third-party websites is solely intended for convenience purposes. Such links should not be interpreted as an endorsement, approval, investigation, verification, or monitoring of any content or information contained within or accessible from those linked sites. The opinions expressed in this post are solely those of the author and the views expressed herein may not necessarily align with the opinions of Hailstone Labs, its affiliates, or individuals associated with Hailstone. The writer’s perspectives are provided for informational purposes only and should not be construed as financial or investment advice.