Mazocarta: A Seeded Procedural Deckbuilder for Instrumented Game Development
A new academic paper introduces Mazocarta, a seeded procedural tactical deckbuilder designed primarily as an instrumented reference artifact for game development rather than a commercial product. Implemented in Rust and compiled to WebAssembly, the system features a unified rules engine that supports interactive browser play, native command-line simulation, automated end-to-end testing, and local-area multiplayer via QR-mediated WebRTC pairing. The core technical contribution is its deterministic run model, which enables reproducible balance probes and regression checks. An evaluation involving 1,000 deterministic seeds demonstrated that the simulation pipeline produces consistent development signals, with single-player and two-player autoplay win rates of 36.1% and 34.9%, respectively. These metrics serve as repeatable baselines for detecting balance shifts rather than final player-facing statistics. By integrating diverse testing workflows into a single shared production core, Mazocarta offers an open-source solution for automated playtesting and mechanical balance analysis, highlighting advancements in software engineering practices for interactive entertainment systems.
Wire timeline
Mazocarta: A Seeded Procedural Deckbuilder for Instrumented Game Development
A new academic paper introduces Mazocarta, a seeded procedural tactical deckbuilder designed primarily as an instrumented reference artifact for game development rather than a commercial product. Implemented in Rust and compiled to WebAssembly, the system features a unified rules engine that supports interactive browser play, native command-line simulation, automated end-to-end testing, and local-area multiplayer via QR-mediated WebRTC pairing. The core technical contribution is its deterministic run model, which enables reproducible balance probes and regression checks. An evaluation involving 1,000 deterministic seeds demonstrated that the simulation pipeline produces consistent development signals, with single-player and two-player autoplay win rates of 36.1% and 34.9%, respectively. These metrics serve as repeatable baselines for detecting balance shifts rather than final player-facing statistics. By integrating diverse testing workflows into a single shared production core, Mazocarta offers an open-source solution for automated playtesting and mechanical balance analysis, highlighting advancements in software engineering practices for interactive entertainment systems.
cs.AI updates on arXiv.org