Zero-Knowledge Proof x Secret Computation Technology Enables Game Progression Without a "Trusted Third Party" - "Trustless Werewolf" Released on Ethereum
Yoii Inc. has released "Trustless Werewolf" on the Ethereum testnet, a game developed with their proprietary "zk-mpc" technology that combines Zero-Knowledge Proof (ZKP) and Multi-Party Computation (MPC) to enable complete game progression without a trusted third party.
📋 Article Processing Timeline
- 📰 Published: April 2, 2026 at 19:10
- 🔍 Collected: April 2, 2026 at 14:02
- 🤖 AI Analyzed: April 21, 2026 at 06:49 (448h 47m after Collected)
Yoii Inc. (Headquarters: Shibuya-ku, Tokyo; CEO: Masaharu Uno; hereinafter 'Yoii') is pleased to announce the release of 'Trustless Werewolf' on the Ethereum testnet (Sepolia). This game has been developed using our proprietary cryptographic technology 'zk-mpc', which combines Zero-Knowledge Proof (ZKP) and Multi-Party Computation (MPC), enabling complete game progression without the need for any trusted third party (game master).
This project was realized after approximately two and a half years of research and development, having been selected twice (February 2024 and February 2025) for the Ethereum Foundation's Ecosystem Support Program (ESP). All source code is publicly available as open source, with applications envisioned beyond just the Werewolf game to a wide range of fields.
### ■ Background of Development
The Werewolf game is a prime example of an 'imperfect information game' where each participant holds secret information such as roles (Villager, Seer, Werewolf, etc.) and proceeds with actions like discussion, voting, divination, and attacks while keeping this information hidden. In traditional online Werewolf games, functions such as role distribution, vote counting, victory determination, and divination result notifications were all handled by the server operator (game master). This mechanism presupposed that players would disclose their secret information to this third party and 'trust' its handling.
However, when games are played between anonymous participants online, there was no means to detect fraud or collusion by the game master. 'How to achieve fair game progression in an environment without a trusted third party' is a challenge common not only to games but also to various aspects of society, including financial transactions, voting, and identity verification.
### ■ Overview of zk-mpc Technology and the Mechanism of 'Trustless Werewolf'
'Trustless Werewolf' replaces all functions of a game master by combining the following two cryptographic technologies:
### Secret Computation - Multi-Party Computation (MPC)
Each player's secret information (role, voting target, divination target, etc.) is split into multiple shares using a technique called 'secret sharing,' with each share held by a separate computation node. Unless the nodes collude, the original information is not revealed to anyone (including the server). This allows all calculations involving secret information, such as role distribution, anonymous voting, divination, attacks, and victory determination, to be executed while maintaining secrecy.
For example, in anonymous voting, the system can calculate only the person with the most votes while keeping who voted for whom secret. While conventional ZKP-only implementations would reveal the number of votes for each candidate, using MPC in conjunction allows the result to be outputted while keeping even the vote counts secret.
### Zero-Knowledge Proof (ZKP: Zero-Knowledge Proof)
This is a technology for 'proving that a certain computation was performed correctly, without revealing any details about its content.' Importantly, ZKP is less a technology for protecting secrets and more a 'mechanism for allowing a third party to verify whether a computation was performed correctly.'
In this system, a proof is generated that verifies the correctness of the computations performed by MPC. This proof allows anyone to verify that the entire game process, including the randomness of role distribution, the validity of each action (e.g., that a person attempting to divine is indeed a Seer), the accuracy of voting results, and the correctness of victory determination, has not been tampered with, without disclosing internal data.
This project was realized after approximately two and a half years of research and development, having been selected twice (February 2024 and February 2025) for the Ethereum Foundation's Ecosystem Support Program (ESP). All source code is publicly available as open source, with applications envisioned beyond just the Werewolf game to a wide range of fields.
### ■ Background of Development
The Werewolf game is a prime example of an 'imperfect information game' where each participant holds secret information such as roles (Villager, Seer, Werewolf, etc.) and proceeds with actions like discussion, voting, divination, and attacks while keeping this information hidden. In traditional online Werewolf games, functions such as role distribution, vote counting, victory determination, and divination result notifications were all handled by the server operator (game master). This mechanism presupposed that players would disclose their secret information to this third party and 'trust' its handling.
However, when games are played between anonymous participants online, there was no means to detect fraud or collusion by the game master. 'How to achieve fair game progression in an environment without a trusted third party' is a challenge common not only to games but also to various aspects of society, including financial transactions, voting, and identity verification.
### ■ Overview of zk-mpc Technology and the Mechanism of 'Trustless Werewolf'
'Trustless Werewolf' replaces all functions of a game master by combining the following two cryptographic technologies:
### Secret Computation - Multi-Party Computation (MPC)
Each player's secret information (role, voting target, divination target, etc.) is split into multiple shares using a technique called 'secret sharing,' with each share held by a separate computation node. Unless the nodes collude, the original information is not revealed to anyone (including the server). This allows all calculations involving secret information, such as role distribution, anonymous voting, divination, attacks, and victory determination, to be executed while maintaining secrecy.
For example, in anonymous voting, the system can calculate only the person with the most votes while keeping who voted for whom secret. While conventional ZKP-only implementations would reveal the number of votes for each candidate, using MPC in conjunction allows the result to be outputted while keeping even the vote counts secret.
### Zero-Knowledge Proof (ZKP: Zero-Knowledge Proof)
This is a technology for 'proving that a certain computation was performed correctly, without revealing any details about its content.' Importantly, ZKP is less a technology for protecting secrets and more a 'mechanism for allowing a third party to verify whether a computation was performed correctly.'
In this system, a proof is generated that verifies the correctness of the computations performed by MPC. This proof allows anyone to verify that the entire game process, including the randomness of role distribution, the validity of each action (e.g., that a person attempting to divine is indeed a Seer), the accuracy of voting results, and the correctness of victory determination, has not been tampered with, without disclosing internal data.