Finding fast solutions to the game of Mastermind

There are two main research issues when playing the game of Mastermind: one of them is finding solutions that are able to minimize the number of turns needed to find the solution, and another is finding methods that scale well when the size of the search space is increased. There is a trade-off between these two paths, and in this paper we will present a method that uses evolutionary algorithms to find fast and scalable solutions to the game of Mastermind. The advantages of this method are the few parameters that are needed to tune it and its robustness even with different parameter settings. We will show the results for different sizes and compare them with published results, showing that there is not a big difference with the best solutions found so far but the number of evaluations needed to find them is considerably reduced and it even scales much better than solutions tuned for minimization of the number of turns needed to win.