Cryptographic Rock-Paper-Scissors

A networked implementation of rock-paper-scissors that uses cryptographic commitments to prevent cheating in distributed games.

The Problem

In traditional networked games of rock-paper-scissors, there's a fundamental trust problem:

• If Player A sends their choice first, Player B can see it and choose accordingly (guaranteed win for Player B)
• If both players send simultaneously, network delays mean one player's choice arrives first
• Even with a trusted third party, you must trust that party won't leak information

How can two players play a fair game over a network without trusting each other or a third party?

The Solution: Cryptographic Commitments

This implementation uses a commit-reveal protocol based on cryptographic hashing to solve the cheating problem:

Phase 1: Commitment

Instead of sending their actual choice, each player sends a cryptographic hash of their choice combined with a random secret (called a "nonce").

Hash = SHA-256(Play + Nonce)

The hash acts as a cryptographic commitment:

• Binding: Once committed, the player cannot change their choice (the hash would no longer match)
• Hiding: The opponent cannot reverse the hash to discover the choice

Phase 2: Reveal

After both players have committed (by exchanging hashes), they reveal:

• Their actual play (Rock, Paper, or Scissors)
• Their secret nonce

Phase 3: Validation

Each player verifies the opponent's commitment:

1. Recompute the hash using the revealed play and nonce
2. Compare it to the hash received in Phase 1
3. If they don't match → the opponent cheated
4. If they match → the game is valid

Why This Works

• No premature information leak: The hash reveals nothing about the actual play
• Tamper-proof: Changing the play after commitment would require breaking SHA-256 (computationally infeasible)
• No trusted third party needed: The mathematics guarantee fairness
• Verifiable: Both players can independently verify no cheating occurred

Game Protocol

Player A (Server)          Network          Player B (Client)
─────────────────────────────────────────────────────────────
1. Choose play + nonce
2. Compute hash            ───────>         Receive hash
                                            Choose play + nonce
3. Receive hash            <───────         Send hash

4. Reveal play + nonce     ───────>         Receive revelation
                                            Verify hash matches
5. Receive revelation      <───────         Reveal play + nonce
   Verify hash matches

6. Determine winner                         Determine winner

Quick Start

Prerequisites

• Rust toolchain (rustc, cargo)

Installation

git clone <repository-url>
cd crypto-rock-paper-scissors
cargo build --release

Running the Game

Terminal 1 (Server):

cargo run
# Select 's' for Server mode
# Choose your play (r/p/s)

Terminal 2 (Client):

cargo run
# Select 'c' for Client mode
# Choose your play (r/p/s)

The game will:

1. Exchange cryptographic commitments
2. Reveal plays and nonces
3. Validate both players played fairly
4. Display the winner

Cryptographic Properties

Security Guarantees

• Collision Resistance: SHA-256 makes it computationally infeasible to find two different inputs that produce the same hash
• Pre-image Resistance: Cannot reverse the hash to discover the original play
• Nonce Randomness: 256-bit random nonce ensures the same play hashes differently each game

Attack Resistance

• Replay attacks: Random nonce prevents reusing old commitments
• Brute force: Only 3 possible plays, but 256-bit nonce makes brute force impractical
• Timing attacks: Both commitments exchanged before any reveals
• Man-in-the-middle: Would be detected during validation phase

Protocol Details

Message Format

Messages are JSON-encoded:

Commitment:

{"HashedPlay": "a3f5d8..."}

Revelation:

{"Explanation": {"play": "r", "nonce": "7c8f3d..."}}

Limitations & Future Work

Current State:

• Single round per execution
• Local network only (localhost)
• Basic error handling
• No reconnection support

Potential Enhancements:

• Best-of-N tournament mode
• Remote play over internet
• Persistent game statistics
• GUI interface
• Multiple simultaneous games
• Configurable network settings

Educational Value

This project demonstrates practical applications of:

• Cryptographic commitments in distributed systems
• Zero-knowledge protocols (hiding information until reveal)
• Byzantine fault tolerance (playing fairly with untrusted parties)
• Mental poker protocols (the same concept used in online card games)

The commit-reveal pattern is used in:

• Blockchain smart contracts (preventing front-running)
• Secure multiparty computation
• Online poker and gambling
• Sealed-bid auctions
• Voting systems

References

• Commitment Schemes (Wikipedia)
• Mental Poker (Wikipedia)