Duck & Cover

September 2017 - April 2018

This was a 2-person project with the goal of developing a toolset with which game programmers can efficiently generate realistic positions for AI to hide and shoot at a target. After analyzing current methods of sampling, we determined a phyllotaxic pattern could be used to eliminate drawbacks that some patterned and random methods run in to, most of which stem from sample density limitations. The rest of our efforts went towards refining our culling methods to generate 'useful' outputs that can be tuned to varying character/map sizes and layouts.