Comment by Ono-Sendai

Anyone else notice how the entropy in the 1000 bouncing balls simulation goes down at some point, thereby violating the second law of thermodynamics? :)