Entropy Doom

Entropy Doom is a procedurally driven maze game that turns randomness into the product: every run drops you into a one of a kind challenge seeded by verifiable entropy. We integrate Pyth Entropy to source cryptographic randomness for each map seed not just to make levels unpredictable, but to make them provably fair and uniquely reproducible. That means each maze has an auditable seed so players can confirm a map’s origin, replay exact runs, or settle disputes with on chain evidence.

Gameplay centers on fast, tense exploration and player choice. The Character Select lets you pick from distinct avatars, each with unique visuals and subtle playstyle differences, so your strategy and expression matter from the first step. The map generator uses the seed to deterministically assemble rooms, corridors, and key/object placements, ensuring enormous variety while adhering to design constraints that make the game fun and solvable.

Why it matters: Pyth sourced entropy makes every map both rare and trustworthy. Players get endless fresh content, streamers get shareable, reproducible runs, and competitive modes can rely on provable randomness. Entropy Doom is about emergent stories born from fair, verifiable chance.

The game runs on a Node.js backend with an Express API that handles level generation, tile-stitching, and optional blockchain calls. It uses ethers.js to interact with Pyth Entropy for cryptographic randomness, with a local fallback so development doesn’t require on-chain fees. User data and scores can be stored in MongoDB, with JWT and bcrypt providing lightweight authentication. On the client side, React + Vite manage UI and routing, while Phaser renders the real-time gameplay scene. A set of small utility modules drives deterministic generation, custom RNG, and server-side image assembly.

Every level is created from one numerical seed. That seed initializes a deterministic PRNG so every map is fully reproducible: the same seed will always yield the same rooms, corridors, enemies, keys, and exit placement. The generator ensures solvability by enforcing connected paths before selecting valid object locations based on simple spatial constraints. Tilesheets are composed into a single PNG asset, so the client receives a compact payload: layout string, seed, tilesheet URL, tile size, and coordinates for objects and player start.

This design gives the game its core benefits: single-seed determinism for perfect replayability and verifiable entropy for fairness. Players can share seeds, replay iconic runs, and trust that levels are unpredictable yet auditable, ideal for competitions, streaming events, and community driven challenges.