Abstract: Combined with practices, the theories and methods ofMarkovian randomprocess, the creation of transition probabilitymatrix and the calculation of stable state probability are concisely described. Consider- ing that in Markovian decision process, a great number of random factors lead to the distortion of transition probability, and hence producing random errors, the author proposes a newviewpoint, and creates a scien- tific, rational, workable and predicative decision-making technique, which combines Markovian decision- making planningwith sensibility analysis. Besides, by making full use of an example, the author discusses the decision-making process and steps in detail.