In the paper: Learning to Solve Game Trees , it isn’t about you learning how to solve game trees but rather if the person playing the game will be able to reach the final goal in your game.  This paper is pretty much readable by someone with an AP in math, computer science or getting Computer science degree easily and if you are brave enough to ignore the equations it will make sense to most everyone with a desire to learn more about machine learning.

The game researched is Go, a game with simple rules but complex play.  Similar to most games generated by hobbyists and even more advance game designers.