How agents update beliefs by observing neighbors — wisdom and cascades
Round: 0 | Mean belief: 0.50 | Correct: 0%
N=50 agents on a random network each hold a belief P(H=1) updated by private signals (noisy observations of the true state) and social signals (neighbors' beliefs). Bayesian weight balances private vs. social evidence. Watch for wisdom of crowds (beliefs converge to truth) vs. information cascades (agents herd on wrong answer when social weight is too high).