Research simulation · Shooter

AI Shooter — Dynamic Difficulty

A third-person shooter testbed for dynamic difficulty research — bot profiles, rule-based and fuzzy DDA, and full metric export.

Confidential client project. Builds and source are under NDA, so there’s no public link — get in touch and I’ll walk you through it privately.

AI Shooter third-person shooter research simulation
Role
Unity Developer
Platforms
Windows
Genre
Research simulation · Shooter
Outcome
Dynamic difficulty research tool
Engine
Unity · C#
Availability
Under NDA · demo on request

The game

A research simulation that measures how dynamic difficulty adjustment (DDA) affects players. Researchers configure bot profiles, switch DDA on or off, choose between rule-based and fuzzy-logic algorithms and run repeatable episodes, while the game samples performance metrics and adjusts a difficulty multiplier in real time.

My contribution

I built the shooter, the bot behaviour profiles, both DDA algorithms, the episode runner and the metrics pipeline that exports every session to JSON and CSV for analysis.

Highlights

  • Configurable bot profiles and episode runner
  • Rule-based and fuzzy-logic DDA algorithms
  • Real-time difficulty multiplier
  • Metrics sampling with JSON and CSV export
  • Simulation settings UI for researchers

Tech stack

  • Unity
  • C#
  • Fuzzy logic
  • JSON / CSV