Notes 225

Notes 2/25 turboSquid asset store achieve 60fps / vr requires 120fps or 180fps networked games: aim for 16ms response time

fps: 70-110 degree field of view

cameras      require smooth motion           spherical linear interpolation (slerp)           easing+tweens       jerk, low framerate, or wide field of view can cause motion sickness           artificial nose!      camera clipping           physics: implement camera collision and bump the object to safe zones           avoid clipping      camera effects           motion blur, shake, film grain           smoothly fade in/out as objects reach visual horizon input      for precision games, you need to develop a sense of mastery over your input device      for mobile: swipe, tilt, touch      funny tutorial: “Look down” and bind whatever they do to down      glow, other affordances to show responsiveness      Control design           Fitts’s law: Bigger targets -> Faster movement to target           Hick’s law: More choices -> logarithmically longer decision time AI      Unity packages implement AI      Decision trees: finite state machines. Can be a performance issue with n^2 transitions between states      Behavior trees           hierarchical state machine. strategy / tactics / actions           can make decisions as a group (like flocking)           Maslow’s hierarchy of needs           -> but requires careful tuning of heuristics. And indecision. 

Navigation      a* gridmap       navigation mesh: navigable nodes and polygons      watershed partitioning or monotone partitioning

Idea: make a game of social interaction with heuristics.  Emergent score

1. How will students interact in a space? 
   1. Film with Kinect; collect kinematic data.
   2. Use this to weight edges in a neural network.
   3. Use the neural network as a heuristic to animate animations of the same students in new spaces.