This sounds like a good use for using a Markov Model to represent time steps and state changes. Use the randomness guarantee to modify the state change probability matrix.
1 Like
Thereās a true RNG on board, youād be better using that. The onboard randomness thing is, sorry, actually a unique source of static data derived from fabrication variation, making cloning of the flash storage onboard impossible (and accessible by other sources - so you could use it to generate a random animal thing).
Thatās actually neat to hear, that each will be uniquely random. My suggestion is more for tracking state changes overall though - if a given state change is completely random, you might end up with cases where the animal would change states from starving and stressed to full and content. A Markov model would allow you to evolve a probability matrix dynamically (using that onboard random generator to modify the state change probability matrix over the course of an evolution) to keep the game random but allowing for some control over state changes that wouldnāt make sense.
I did a research project in uni on Markov models, so I just recognized a potential use case and thought Iād share. Definitely not the only solution and almost assuredly not the best one - I just get excited when I get to actually use some of the research I did since I donāt use it in my current job 
1 Like
I agree your thoughts are gonna be a good idea. Iām not great at those kind of maths tbh, but itās something we may explore down the road just as a random project.