August 2018
Beginner
334 pages
10h 19m
English
We developed our recipe using a pseudo-exploratory algorithm and stored the payoff from taking a certain action. The recipe first tests every action there is, and then keeps making moves using the formula from the UCB1 algorithm. It is good enough to be challenging, and at the same time it is defeatable because the algorithm explores different options—Thus, creating better changes for the player to improve their skills, or just to be lucky.
Read now
Unlock full access