I've scoured the internet for a clear answer on entropy and I see contradictory answers everywhere so I'm going to refine my question here. Please read carefully as the question is not as it seems at first.
The formation of spherical planets from dust clouds in the universe seems to be an example of particles coming from a random more distributed state to a more spherical state. This seems to look like entropy decreasing.
There are more microstates for total dust particle positions and velocities spread out over the volume of a gas cloud then there are possible dust particle positions and velocities of particles confined to a spherical planetary shape.
Thus if the formation of planets from dust clouds involve particles moving from a higher amount of microstates to a lower amount of microstates then entropy is seemingly decreasing.
The common response to this statement is that it is wrong. When particles clump under gravity, heat radiation is generated and spread out to the universe increasing the overall entropy of the universe while decreasing local entropy.
Now watch this video: https://vimeo.com/47349336
In this computerized simulation heat radiation is removed from the equation. Also there is no concept of local entropy vs. overall entropy as you are observing the entire simulated universe in the video. That's right you are looking at the entire simulated universe and observing all the entropy within it, not a localized part. No heat or radiation or anything else. The only rules here are newtonian mechanics. That's it.
Yet in this video we see the amount of microstates spontaneously decrease. We see particles organize themselves into spheres. Why? What is wrong with my understanding of entropy? Can someone explain what's happening to entropy in this video from a statistical mechanics point of view?
I understand this simulation may be inaccurate to the universe. Energy is seemingly be removed the system as the velocities of these particles are artificially lowered during collisions. My goal however is to understand this system in terms of microstates and macrostates from a statistical mechanics point of view. If someone can play devils advocate and explain it from this perspective it will be helpful. Thank you.