I think games like NASA's Moonbase Alpha demonstrate why attempts like this aren't all that valuable: https://en.wikipedia.org/wiki/Moonbase_Alpha_(video_game)

I think games like NASA's Moonbase Alpha demonstrate why attempts like this aren't all that valuable: https://en.wikipedia.org/wiki/Moonbase_Alpha_(video_game) https://en.wikipedia.org/wiki/Moonbase_Alpha_(video_game) Have a look at the trailer: https://www.youtube.com/watch?v=nis2t9EubBs https://www.youtube.com/watch?v=nis2t9EubBs Then have a look at how seriously people took it: https://www.youtube.com/watch?v=Hv6RbEOlqRo https://www.youtube.com/watch?v=Hv6RbEOlqRo I mean, i guess people can approach games like EVE Online seriously, in part due to the financial investment that they may have into them, but otherwise they're likely to just mess around for entertainment value. On a different note, if you have such complete knowledge of all the constraints to be able to make a simulation the results of which would be useful, why not just run a neural network to simulate more outcomes than players for this game of yours would? Why depend on the game being interesting and successful when you can just automate the players away?