[bengarrr]

Entropy IS always increasing. Water is always gonna flow downhill. Its not possible to go from maximally ordered to even more ordered, except in some kind of highly localized phenomena. But it is possible that it can return to that maximally ordered state after it reaches a maximally unordered state. This is the idea Roger Penrose has for the big bang is that it is the state of the universe immediately succeeding the heat death of the universe, at a certain point maximally unordered and maximally ordered can look the same, at least mathematically.

[pizza]

I'm not really talking about things that violate energy conservation or statistical mechanics. I'm talking about how the uniformity of the state during the big bang lead to something clumpier. Structure formation. And I still find it suspect to apply the concept of thermodynamical entropy, something defined for isolated systems at thermal equilibrium, primarily, to even make sense at universe-scale.

[bengarrr]

Structure is not entropy though unless I am missing what you are saying. The arrangement of macro level objects like a supercluster has less to do with entropy and more to do with dark matter/energy or what ever that stuff is. Also I dont follow the reasoning that this concept of entropy is incorrectly being applied to the universe... energy is energy I have never heard that the thermodynamics only applies to isolated systems in equilibrium.