Abstract. For models of binary deterministic statistical experiments (SE) defined recursively by solutions of difference deterministic equations for the probabilities of binary states alternatives, we analyze the classification depending on the values of directing parameters, which define the increments of regression functions of alternative probabilities. This classification is justified by the limit properties of the solutions of the deterministic difference equations, which generate the alternatives probabilities.

Keywords: statistical experiment, deterministic difference equation for probabilities of alternatives of binary states, persistent regression, equilibrium, attracting state, repulsive state, dominance of alternatives.