Mazocarta: A Seeded Procedural Deckbuilder for Instrumented Game Development

Mazocarta is an open-source procedural deckbuilder game built in Rust and WebAssembly that serves as a reference architecture for instrumented game development. The project demonstrates how a single rules engine can support interactive play, automated testing, balance simulation, and local multiplayer, with evaluation showing 36.1% single-player and 34.9% two-player win rates across 1,000 deterministic seeds.

Mazocarta represents a shift in how game developers approach technical infrastructure by prioritizing reproducibility and automated validation alongside player experience. Rather than treating gameplay simulation as separate from production code, the architecture unifies interactive play, command-line testing, and multiplayer into one deterministic engine, enabling developers to detect balance regressions and validate mechanics changes systematically.

The project emerges from broader industry recognition that game balance and stability require rigorous engineering practices similar to financial systems or safety-critical software. Procedural deckbuilders have grown increasingly complex, with card interactions creating exponential balance spaces that manual playtesting cannot adequately cover. Mazocarta addresses this by providing reproducible "balance probes"—automated simulations that generate consistent signals about game state across thousands of scenarios.

For game developers and studios, this architecture reduces the friction between rapid iteration and stability. QR-mediated WebRTC pairing for local multiplayer demonstrates how browser-native technologies can replace traditional networking infrastructure, lowering deployment barriers for indie and mid-size teams. The open-source release positions Mazocarta as a template for similar projects rather than a commercial product.

The significance lies not in the game itself but in the engineering pattern it validates. As AI-assisted game design tools become more prevalent, having deterministic, reproducible evaluation frameworks becomes essential for distinguishing between legitimate balance improvements and statistical noise. Future adoption will likely focus on whether this architecture scales to larger, more complex games and whether the reproducibility model extends to narrative-driven or physics-heavy titles.

• →Mazocarta unifies interactive play, automated testing, and balance simulation in a single Rust-based rules engine, enabling reproducible game development validation.
• →The system achieved measurable win rates (36.1% single-player, 34.9% two-player) across 1,000 deterministic seeds, demonstrating feasibility of automated balance probing.
• →Browser-native WebAssembly compilation combined with native simulation support allows developers to test balance changes across platforms without recompiling.
• →QR-mediated WebRTC pairing simplifies local multiplayer setup, reducing networking complexity for browser-based games.
• →The open-source reference artifact provides a template for game studios seeking to implement deterministic regression testing and automated playtesting workflows.