The state transition algorithm creates an output with an array of numbers whose order is chosen by a series of probabilistic calculations of elements. Each number is chosen by the probability of moving from one state to another. The extent of the matrix is limited by the 6x6 dimensions; however, a smaller matrix can be used by leaving the last rows and columns blank. Each active square must be filled by a number or left blank for a zero value. The squares of the first column or row must have a number between 0 and 88. These elements, or entries, represent current or new states. The rest of the matrix entries represent percentages of probability. each row must add up to 100.

This particular matrix has current and new states that are mirrored by the same integers. In this sense the matrix lends itself well to creating permutations.

The state transition matrix is a first order tool for a Markov chain, which is one of the earliest methods for algorithmic composition from the late 1950's. Markov chains use probabilistic information for analyzing trends and outcomes which are useful for business, advertising and sociology. The chains represent a series of observations and probabilities.