The Parliament House: Battle of 272
Unmute
An Indian Parliament 3D fighting & political strategy game featuring authentic politician voice lines, dynamic 3D combat, and a real-time Explainable AI (XAI) decision engine.
What tools did you use to create your project?
How much experience does your group have? Does the project use anything (art, music, starter kits) you didn't create?
Experience: Beginner to Intermediate web and game developers.
What we created:
- Built the entire game engine in React 19 + TypeScript + Vite.
- Engineered the real-time Explainable AI (XAI) decision tree system and Expected Value (EV) combat evaluation from scratch.
- Created custom Three.js procedural animation rigs, combat hitboxes, podium capture logic, and virtual gamepad controls.
- Hand-coded the cartographic SVG Map of India with accurate geographical coordinate projection for constituencies.
External Assets Used:
- Open-source Three.js 3D humanoid base model for character skeletal rigging.
- Authentic public domain / transformative parody voice audio clips and sound meme hooks for Indian politicians.
- Lucide React vector iconography.
What we created:
- Built the entire game engine in React 19 + TypeScript + Vite.
- Engineered the real-time Explainable AI (XAI) decision tree system and Expected Value (EV) combat evaluation from scratch.
- Created custom Three.js procedural animation rigs, combat hitboxes, podium capture logic, and virtual gamepad controls.
- Hand-coded the cartographic SVG Map of India with accurate geographical coordinate projection for constituencies.
External Assets Used:
- Open-source Three.js 3D humanoid base model for character skeletal rigging.
- Authentic public domain / transformative parody voice audio clips and sound meme hooks for Indian politicians.
- Lucide React vector iconography.
What challenges did you encounter?
1. Real-time 3D Performance in the Browser: Integrating Three.js 3D character rigs alongside a 60 FPS 2D canvas debate stage and camera shake without dropping frames required careful texture optimization and memory management.
2. Designing an Explainable AI (XAI) System: Instead of a simple black-box bot, we built a transparent decision tree that calculates mathematical Expected Values (EV) across 8 combat actions every frame and provides plain-English counterfactual explanations ("Why did you do that?").
3. Dual Win Condition Balance: Balancing a traditional fighting game (0 HP Knockout) with a political strategy game (capturing 272 Majority Seats on the podium) required fine-tuning knockback velocities and seat accumulation rates.
4. Full Cross-Platform Controls: Ensuring responsive playability across both desktop keyboards and mobile touchscreens with custom multi-touch virtual joysticks.
2. Designing an Explainable AI (XAI) System: Instead of a simple black-box bot, we built a transparent decision tree that calculates mathematical Expected Values (EV) across 8 combat actions every frame and provides plain-English counterfactual explanations ("Why did you do that?").
3. Dual Win Condition Balance: Balancing a traditional fighting game (0 HP Knockout) with a political strategy game (capturing 272 Majority Seats on the podium) required fine-tuning knockback velocities and seat accumulation rates.
4. Full Cross-Platform Controls: Ensuring responsive playability across both desktop keyboards and mobile touchscreens with custom multi-touch virtual joysticks.
Made with 💖 by robots running our open-source software and GraphQL APIs.