KOSUNI CRYPTO
Korea's Crypto Pulse, in English
$BTC $ETH $XRP $SOL $DOGE
News
Neutral

Bitcoin Poker Protocol Proposed for Two-Player Texas Hold'em Without a Dealer

Published October 9, 2026 9:35 AM · 7 views $BTC
Bitcoin Poker Protocol Proposed for Two-Player Texas Hold'em Without a Dealer

A design has emerged that allows two players to conduct Texas Hold'em poker without a dealer while adhering strictly to Bitcoin's consensus rules. From generating private cards to betting and settling stakes when a player exits the game, all steps are handled through transaction and signing protocols. Stanford University doctoral researcher Robin Linus proposed a two-player limit Texas Hold'em protocol in his paper 'Bitcoin Poker.' Bitcoin.com News noted this design comes nearly 18 years after Satoshi Nakamoto left an unfinished poker interface in early software.

Linus' paper clarifies that while Satoshi's initial code included a poker lobby and table screen, it does not constitute a completed game protocol. It remains unclear whether Satoshi intended to build the actual game. The core of the design involves structuring possible game paths as pre-signed transaction trees. Two participants combine secret information to create cards, verifying validity and uniqueness through cryptographic commitments and zero-knowledge proofs. Instead of shuffling all 52 cards, only nine are generated per hand. According to the paper, the probability of passing duplicate checks is approximately 48.025%, requiring an average of about 2.08223 attempts to obtain a valid hand.

When betting or card reveals occur, corresponding transaction paths are followed. If one player fails to proceed in turn, the other can settle stakes via a predefined block delay timeout path. However, Bitcoin does not force card disclosure; players may end the game and settle agreed-upon stakes without revealing opponents' cards. Costs remain unresolved for such cases. The paper assumes 100 big blinds per participant with up to four bets per round. This requires 56,132 logical nodes, 54,855 regular signatures, and 1,306 card reveal packages in the transaction tree, with preparation data totaling 8.4MB.

Desktop browser tests showed preparing and storing the transaction tree took 26.39–27.26 seconds across three trials. The paper notes these measurements are from a test environment and do not guarantee startup times or real-world performance. The research limits participants to two, considering issues of secret card sharing and collusion between players. Formal proofs for all malicious participant scenarios were not provided, and independent review is needed for proof composition and adapter signature extensions.

The paper describes this proposal as a research design expressing non-public information games and dispute settlement using existing Bitcoin rules rather than a finished online poker service. Practical implementation would require addressing transaction tree preparation costs, off-chain message exchange, and independent security reviews.

Korean Source

This article is an English localization of a Korean-language crypto news report. Original headline: 비트코인 규칙 그대로 2인 포커 설계…준비 데이터 8.4MB