Procedural Generation: Fun = Feasability + Fairness + Interestingness
summary
Procedural generation is built as a design tool: Party Golf’s system (prototype in 12–48 hours) uses Unity/C#, cellular automata smoothed with Perlin-like noise, and marching squares to emit meshes and colliders; a custom pre-fill step preserves tee-and-hole flats. Feasibility is checked by simulated AI shots but impossible-but-interesting levels are intentionally kept; fairness uses proximate tee placement plus rubber-banding; interestingness comes from 300 editable parameters, cave generation, and trials — the system ships and scales, with many fun edge cases left in.
AI-written summary · about
chapters (23)
- 0:00Introduction
- 3:12Party Golf overview
- 5:59Procedural generation history
- 7:24Desert Golfing generation
- 10:50Early Party Golf generation
- 15:28Balancing fairness
- 17:17Addressing interestingness and variety in levels
- 18:01Cave generation system
- 20:09Cave grid smoothing
- 25:05Connecting cave regions
- 26:35Cave grids to meshes
- 29:44Custom mesh generation
- 31:53Platform placement
- 34:53Edge cases handling
- 37:26Marching Squares terrain
- 40:15Tees and holes placement
- 42:41Distance fields for AI
- 45:13Ensuring feasibility
- 47:00Challenging yet possible levels
- 48:46Multiplayer fairness
- 49:28Player customization
- 50:54Trials mode caves
- 51:35Q&A
related talks
38mProcedural Generation at Runtime: From Basic to Complex by Example
03mPlayer-Centric Procedural Content Generation: Enhancing Runtime Customization by Integrating...
50mBuild Your Workforce
45mPanel: Can Generative AI be infused in Gaming?
08mCreative Console: A Player-driven Game Based on A Modular Fast-Evolving-And-Verifying Framework
33mProcedural Content in JS
34mGenerative AI + Spatial Computing = ∞
15mGaming the Future of Architecture