Writing computer programs to play games is a pleasant avocation. To capitalize on this opportunity it makes sense to use the computer's abilities for computation and storage that human competitors lack. Algorithms that are usable by people in a reasonable amount of time are often poor choices for computerization. I believe that Roa's algorithm to play the game of mastermind as published in the October 1982 SIGART Newsletter serves as an example. The algorithm exhibits several fatal flaws:□It is predictable.□It is reproducible.□it performs at the novice level.
/lp/association-for-computing-machinery/writing-a-computer-program-to-play-mastermind-that-appears-to-be-gL8e0XI49Y