Abstract

We intend here to substantiate the claim that the intensive use of trees to tackle Bayesian problems may lead us to “miss the Bayesian wood”, particularly if we just focus on the static trees and ignore germane random walks on them. Our main point is that random walks on networks or grids instead, provide a more fruitful and insightful metaphor to address Bayesian problems and fathom the underlying “Bayesian flows”. Besides recalling the main tenets of our theoretical background, we discuss below the relation of our claims with related research in this field and give some illustrative classroom examples, arising mainly from our teaching stochastics to non-mathematically inclined first year university students and prospective mathematics teachers.