This article is devoted to the issues of the efficiency increasing for buckwheat seed production by reducing mechanical damage during mechanized processing. It is indicated that one of the possible ways to reduce grain damage by machines’ details on a mechanized line is to assess the degree of damage adequately. Due to the fact that the samples analysis for mechanical damages taken from a grain batch that has passed according to any technological processing scheme is a labor-intensive and time-consuming procedure, the reliability of the results of which is largely dependent on the human factor, it is proposed to increase accuracy measurements on the basis of reliable experimental data, reveal the patterns of injury depending on the state of the grain material and its physical and mechanical properties, and in accordance with this, in the future, predict the quantitative increase in mechanical damage in general and by the types of injuries inflicted on the grain by a particular machine by using mathematical relationships. There are the theoretical dependence for the mechanical damages indicators on the number of passes through the machine and experimental dependence on humidity in this article established by using Markov chains. It was revealed that with mechanized processing there is a process of successive transition from less dangerous injuries to more dangerous injuries. The indicators of mechanical damage prediction will permit to compare the efficiency of individual machines of the production line in terms of their damaging ability and purposely design the technological process with minimal damage of the grain material.