Ruining of the technological culture in universities has affected quality of the high level education. In the postgraduate education which has become the third level of education it is shown especially brightly. Frequently applicants have no outlook that is researches and the Ph.D thesis preparing. Also choice of a research subject leaves on a background at all. More and more potential postgraduates have such troubles. In the paper, an algorithm of finding a research problem is described for the potential post-graduate students. The field doesn't matter but some attention is paid to open problems of automatics. Various scenarios and strategy in detection of the research problem are considered. Features of good problems are revealed. Do not forget about risk estimation because of the problem should be finalized within frameworks of postgraduate study. Issues of effective interaction with advisor are mentioned. Evaluation of quality of post-graduate work is carried out by published papers, therefore some advices concern publication activity are given. It is the useful information for those who has reflected for the first time on postgraduate applying, but probably overdue for those who has already enlisted.