Overview
Feedback loops fail when they are treated as polish rather than architecture. Most developers add feedback (screen shake, hit sparks, damage numbers) after the mechanics are designed, as decoration. But feedback is the mechanism by which the player learns the rules of the system — if the feedback doesn't clearly communicate what happened and why, the player cannot form accurate mental models, and the mechanics might as well not exist.
The Feedback Loop Engineering prompt builds feedback systems with three properties: (1) stimulus-response mapping — every player input produces at least one feedback response within 200ms, and every game state change that affects the player produces at least one feedback signal, (2) reinforcement scheduling — desired behaviors receive stronger and more varied feedback than undesired behaviors, using variable-ratio reinforcement that resists habituation, and (3) channel separation — different types of information are communicated through different sensory channels (visual for spatial, audio for temporal, haptic for impact), preventing channel overload where multiple signals compete for the same perceptual bandwidth.
What you get: - Stimulus-response map with latency budgets per channel - Reinforcement schedule for desired vs. undesired behaviors - Channel allocation matrix (visual, audio, haptic, numerical) - Habituation resistance model (when feedback stops producing response) - Feedback hierarchy (critical vs. secondary vs. ambient) - Edge case handling (feedback conflicts, simultaneous events, accessibility)
Built for: gameplay engineers, UX designers, and game designers who need feedback that teaches mechanics — not just looks pretty.