ServerFi: A New Symbiotic Relationship Between Games and Players
TechFlow Selected TechFlow Selected
ServerFi: A New Symbiotic Relationship Between Games and Players
A new understanding of GameFi from Yale University.
Navigating Web3 tides with focused insightsAuthor: Pavun Shetty
Translation: Bocai Bocai｜bocaibocai
Preface
When GameFi first exploded, Bocai invested real money into many Play-to-Earn games—only to lose almost everything. At that time, one thing was clear: early players reaped the biggest rewards, effortlessly profiting by simply mining, selling, and cashing out at the expense of later entrants. Newcomers had to buy in at inflated prices while constantly recruiting fresh "vegetables" (naive players) to sustain the cycle. This turned GameFi into Fi-only, with little actual gaming. While flawed tokenomics certainly played a role, this recent paper from Yale University proposes two entirely new token economic models, tested through simulations, offering fresh ideas for sustainable GameFi development.
Bocai has translated and excerpted key parts of the Yale paper for your reading:
Abstract
Blockchain-based games have introduced novel economic models that merge traditional gameplay with decentralized ownership and financial incentives, fueling the rapid rise of GameFi. However, despite their innovative appeal, these games face significant challenges in market stability, player retention, and the sustainability of token value. This paper examines the evolution of blockchain gaming and uses entropy theory to identify core flaws in current tokenomic designs. We propose two new models—ServerFi, which emphasizes privatization through asset synthesis, and another focused on continuously rewarding high-retention players. These models are formalized into mathematical frameworks and validated through agent-based simulation experiments. Our findings indicate that ServerFi is particularly effective in maintaining player engagement and ensuring long-term viability of game ecosystems, pointing toward promising directions for future blockchain game development.
Introduction As technology advances, the gaming industry continues to thrive on journeys filled with adventurers and outdoor enthusiasts [1]. Starting in the 1970s, Atari launched "Pong," an arcade-style table tennis game that captivated consumers during a turbulent decade and inspired numerous imitations. With more powerful microprocessors, dedicated graphics chips, and personal computers like the Commodore 64, it became possible to create complex, visually rich, and sonically immersive games. Following these pioneers, Nintendo quickly dominated the console market with its Nintendo Entertainment System (NES), releasing iconic titles such as Duck Hunt and Excitebike. Sega and Sony also emerged as strong competitors—Sega with Genesis and Game Gear, and Sony with PlayStation 2 and 3, which leveraged CD-ROMs for enhanced game storage and helped define the future of consoles after 1994. A final milestone came with Microsoft’s widespread adoption of the DirectX API, revolutionizing game development.
Online multiplayer games like World of Warcraft and Fortnite transformed how players interact, marking a giant leap for the gaming industry driven by internet advancements. These games became cultural phenomena, enabling millions to share virtual worlds and fully enjoy technological pleasures. The rise of cloud gaming platforms like Google Stadia and Microsoft xCloud is also notable—they stream games directly to devices, delivering high-quality experiences without requiring powerful hardware [2]. These innovations ushered players into a highly social and interconnected world, made possible by internet technologies, pushing the gaming industry into a new era. These visionary changes reignited public interest in decentralization and data ownership. In traditional gaming, player data and assets were stored centrally on servers controlled by game companies—even items purchased by players. Ownership of these contested assets never truly belonged to the players who bought them, constrained by legacy economic models. For decades, this traditional model revolved around player spending and corporate profits, offering minimal returns for players' investments of time and money. Often called “walled gardens,” these games host in-game items, characters, and currencies on developers’ servers, denying players true ownership over their accounts, content, and digital assets. This limited player rights despite significant time and financial investment, even excluding those who help maintain financial circulation and sustainability within the game from receiving any real economic value.
The emergence of GameFi reshaped economic production relationships by introducing real-world incentives. When discussing the unexpectedly smooth integration of “gaming” and “finance,” blockchain-based “play-to-earn” (P2E) games were perfectly positioned for a breakthrough. Blockchain games typically generate crypto assets in two main ways: tokenizing in-game items as NFTs and designating fungible tokens as in-game currencies [3]. By combining traditional gaming with on-chain assets, these games enable decentralized ownership, transparency, and tangible economic incentives for players. Yet, major challenges remain in market stability, player retention, and token value sustainability. This paper begins by outlining the development background and pioneering cases of blockchain gaming. Then, we apply entropy theory to analyze underlying causes of current challenges and clarify forces driving market dynamics. Based on these insights, we introduce two innovative tokenomic models: ServerFi, which enables privatization via asset synthesis, and a model that continuously rewards high-retention players. These models are formalized into mathematical frameworks and validated through agent-based simulation experiments. Our results highlight the potential of the ServerFi model in sustaining player engagement and ensuring long-term ecosystem viability.
Background: The Rise of GameFi Blockchain games create crypto assets primarily in two ways: representing in-game items as NFTs and granting fungible tokens the status of in-game currency. 2013 marked pivotal moments, including Meni Rosenfeld’s concept of Colored Coins, which drew attention to the importance of virtual asset ownership and enabled mapping real-world assets onto the Bitcoin blockchain [4]. Four years after Rosenfeld, Larva Labs launched the CryptoPunks NFT collection—a major milestone in NFT development. Its 10,000 unique, randomly generated character images inspired Ethereum’s ERC-721 standard for digital art and collectibles [5, 6].
NFT technology was clearly embraced by visionary founders. Dapper Labs released CryptoKitties, the first blockchain game on Ethereum, which briefly congested the network and caused significant transaction delays. In this game, players could buy, breed, and trade virtual cats, each with unique visual traits and varying rarity. The massive success of CryptoKitties highlighted the appeal of NFT-based gameplay. It tapped into psychological desires for true ownership and potential financial gain, attracting passionate collectors and savvy investors through in-game financial loops. Breeding and trading rare cats created speculative excitement. That year, discussions about CryptoKitties nearly went mainstream. This creatively designed GameFi game attracted millions of players who not only owned rare “cats” but also gained social identity and belonging through the CryptoKitties community.
Among NFT and play-to-earn (P2E) crypto games, Axie Infinity by Sky Mavis rose as a spiritual successor to CryptoKitties, quickly becoming a sensation thanks to its engaging gameplay loop, with players often playing late into the night. Axie Infinity allows players to collect, breed, and battle fantasy creatures called Axies [7]. Each Axie is backed by an NFT with unique attributes and abilities that can be enhanced through strategic breeding and gameplay [8]. This enjoyable GameFi game offered economic incentives similar to CryptoKitties but introduced more complex mechanics and a robust in-game economy. Its far-reaching design philosophy attracted a broad player base, setting new standards for its era and benchmarks for all future blockchain games.
Challenges in Tokenomics and Our Solutions Faced with intense competition from traditional online games running on centralized infrastructure, blockchain games are adapting to storing digital assets on-chain, allowing players to own items they can sell, transfer across games, or use in specific DeFi applications. Incentive models are gradually maturing alongside the mass adoption of blockchain technology. This opens an entirely new path for cutting-edge production relationships between players and developers. Times have changed—these innovations aim to reconstruct digital society, holding transformative potential for the post-gaming era. Against this backdrop of progress, we must ask: why, amid the leap forward in Web3, would game developers adopt a new production relationship originating from GameFi—one where demand for assets varies among players and traditional, relaxed gaming experiences are de-prioritized?
Most games have a lifecycle, and CryptoKitties was no exception. Within its core mechanics, the breeding system allowed players to produce new “cats,” inadvertently increasing supply and gradually reducing individual “cat” rarity and value over time. As more players joined and bred cats, the secondary market quickly became oversaturated. The scenario was novel and initially exciting, but the dilemma was familiar: how to maintain circulating token prices? Without sufficient active players, demand cannot keep up with growing supply, exacerbating depreciation. Individuals who invested substantial time and resources into breeding may find their efforts increasingly unrewarded. As the collective game progresses, initial scarcity gives way to abundance, leading to declining player interest and participation.
Applying entropy theory to tokenomics offers a professional and insightful perspective on understanding token flow and value fluctuations in blockchain projects. Rooted in the second law of thermodynamics, entropy theory states that in a closed system, entropy (a measure of disorder) tends to increase over time. This concept can be analogously applied to economic systems, especially tokenomics, enhancing our understanding of token distribution, usage, and market volatility. In tokenomics, initial token distribution is typically orderly—tokens are relatively concentrated, prices stable, and player expectations high [9]. Over time, more tokens enter the market through game mechanisms. Increased player trading and token circulation raise market entropy (disorder). During this intermediate phase, internal chaos surges, causing high price volatility. Challenges in tokenomics include inflation caused by excessive token supply and price instability due to speculative influxes. Without effective market regulation and incentive mechanisms, the system may reach a high-entropy state where token values decline broadly and player engagement drops. To maintain long-term system health, it's crucial to implement mechanisms that link new incentives and regulations—actions that can slow entropy growth, preserve relative market order and stability, and sustain player engagement.
We often view tokenomics as isolated events—single-point failures caused by specific reasons and effects. But from this angle, the story is less about any single company and more about the global entropy increase in token circulation. Certain factors are inherently disruptive; certain gameplay patterns inevitably fail. Take Axie Infinity: its tokenomic design has several drawbacks from a player perspective. First, Axie Infinity’s economy heavily relies on continuous generation of new tokens (e.g., Smooth Love Potion, SLP). As more players join and breed Axies, newly minted tokens flood the market, rapidly expanding supply. This imbalance leads to declining token value over time, devaluing players’ holdings. Second, during the Token Generation Event (TGE), many players and investors rush in to profit quickly from trading. Such speculation causes sharp price swings, undermining market stability. In the long run, early speculators exiting with profits may trigger a crash, harming regular players. Third, Axie Infinity lacks sustained incentives to retain players post-TGE. Once initial novelty fades, player enthusiasm wanes due to limited economic rewards. Solving in-game flaws helps attract new users and potentially boosts token demand. However, playing Axie Infinity requires purchasing Axies—an expensive upfront cost. This high barrier limits accessibility and mass adoption. Moreover, prices for rare Axies can become astronomically high, making them unaffordable for average players.
Based on the above discussion, we propose two recommendations to improve GameFi tokenomic models:
ServerFi: Privatization Through Asset Synthesis
In line with Web3 principles, players could be allowed to combine their in-game assets to ultimately gain sovereignty over future servers. This concept, called “ServerFi,” involves players accumulating and merging various NFTs and other digital assets within the game to gain control over game servers. This form of privatization not only incentivizes deeper investment in the game but also aligns with Web3 ideals of decentralization and community governance. By granting players ownership and control over game servers, we foster a more engaged and loyal player base, as they hold real stakes in the ecosystem. For example, we could design a game where players earn daily lottery entries based on their contribution value to the server. These entries allow them to draw fragments. Once a player collects all required fragments, they can synthesize an NFT. By staking this NFT, players can share in the contribution value generated by others on the same server.
Continuous Rewards for High-Retention Players
Another approach is for project teams to continuously identify and nurture high-retention players to maintain token vitality and ensure ecosystem health. Using advanced algorithms and data analytics, projects can monitor player behavior and engagement, offering targeted rewards to those showing strong commitment and high activity. This method ensures the most loyal players stay involved, promoting ongoing interaction and supporting overall tokenomic stability and growth. For instance, we could design a game that airdrops a portion of server revenue daily to top contributors based on system contribution value. This creates a dynamic “play-to-earn” model that rewards participation and contribution.
Experiment
To evaluate the effectiveness of our proposed tokenomic models, we conducted agent-based simulation experiments for each. These experiments aimed to compare and analyze differences in value capture capability between blockchain games built on two distinct tokenomic frameworks. For more accurate modeling, we first formalize the definitions of these tokenomic mechanisms as follows.
-
ServerFi: Privatization Through Asset Synthesis
-
Let vi denote player i’s contribution value to the system in each iteration.
-
Define function f(v) = λv as the number of lottery entries a player earns through contribution value v, where λ is a scaling constant greater than 1.
-
Assume there are k prizes in the lottery, each with a probability of 1/k.
-
Assume the number of new players on day one is n, and considering game growth dynamics, define the number of new players at iteration i as n/α^(i−1).
-
We assume all players are rational. Thus, if a player calculates that the cost of synthesizing an NFT exceeds current staking rewards, they will exit the game. Specifically, for a new player, the expected cost to collect all fragments is λΣ(1/k). When this cost exceeds the staking reward of a single NFT, no new players will join.
-
The total value of the system at iteration i (day) is Ti = Σvi, where n is the number of players at iteration i.
Continuous Rewards for High-Retention Players
-
Let vi denote player i’s contribution value to the system in each iteration.
-
We specify that the system will reward the top 20% of players with 80% of total revenue based on cumulative contributions over the past five days.
-
We assume all participants are rational. Each player has a randomly initialized tolerance threshold; if they fail to receive rewards multiple times consecutively, they will choose to exit the game.
-
The total value of the system at iteration i is Ti = Σvi, where n is the number of players at iteration i.
Given the inherent randomness in real-world scenarios, our simulation experiments introduced stochastic noise from multiple angles, including individual behavior and population growth. For example, we included mutation operators in individual modeling to capture random productivity fluctuations among participants. To ensure fair comparison between the two strategies, both experimental groups used identical parameters, such as maximum iterations and initial population size. Each economic model ran for 500 iterations, repeated 100 times. Results are shown in Figure 1. The x-axis represents iteration count, and the y-axis shows total player contribution value per iteration. The light band indicates the range between maximum and minimum values, and the dark line represents the mean.
In the asset-synthesis privatization model (left), we observe a consistent upward trend in total player contribution as iterations increase, indicating the model effectively sustains player engagement and drives long-term value growth. In contrast, in the continuous-reward-for-high-retention model (right), player contributions rise significantly at first but then sharply decline. Although this model shows high initial contributions, the subsequent drop suggests challenges in maintaining long-term player participation.
Based on the modeling results, we conclude that while continuously rewarding high-retention players may boost early engagement, this approach inherently intensifies player stratification over time. Specifically, it risks marginalizing tail-end players due to insufficient positive feedback, eventually driving them to quit. This stratification also raises entry barriers for new players. Consequently, fewer new players combined with departing tail players reduce rewards for existing top players, triggering a downward spiral.
In contrast, the ServerFi mechanism, based on fragment synthesis, introduces randomness through the lottery process, enhancing social mobility within the player community. For existing NFT holders, the need to continuously synthesize new NFTs ensures that even top players cannot “coast”—they must keep contributing to maintain their status. For new or low-contributing players, ample opportunities remain to synthesize NFTs and share server rewards, promoting upward mobility. Thus, the ServerFi model more effectively fosters social mobility among players, activates the entire system, and cultivates a more sustainable ecosystem.
Conclusion
In this paper, we examined the tokenomic challenges present in current blockchain-based games. Analysis reveals that traditional economic models often lead to market instability, declining player engagement, and unsustainable token value. To address these pressing issues, we proposed and analyzed two promising tokenomic models, with particular emphasis on the ServerFi model based on privatization through asset synthesis. Through extensive agent-based simulation experiments, ServerFi demonstrated significant potential in maintaining player engagement and ensuring long-term ecosystem sustainability. Unlike traditional models, ServerFi fosters social mobility by creating a dynamic, competitive environment where continuous value contribution is necessary to maintain status. This model not only nurtures a more vibrant and inclusive community but also provides a scalable and resilient framework for future blockchain games. As the industry evolves, the ServerFi approach may represent a pivotal shift in tokenomic architecture, offering a more sustainable path for integrating decentralized technology into gaming.
Join TechFlow official community to stay tuned
Telegram:https://t.me/TechFlowDaily
X (Twitter):https://x.com/TechFlowPost
X (Twitter) EN:https://x.com/BlockFlow_News
