A new method was proposed to solve the global minimization problems of the H¨older functions on compact sets obeying continuous functions. The method relies on the Monte Carlo batch processing intended for constructing the sequences of values of the “quasi-global” minima and their decrements. A numerical procedure was proposed to generate a probabilistic stopping rule whose operability was corroborated by numerous tests and benchmarks with algorithmically defined functions.