It’s almost as though by allowing for this natural behavior, randomness gets more extreme (aims towards extremities). Wondering if there are ways this model collapses - for example, if any jump from one bar to the next gets too big (over a threshold), does that cause the whole model to collapse?