Poor information means that characteristic information presented in the subject investigated is incomplete and insufficient, with a small sample and the lack in prior knowledge of probability distributions and trends. For a long time, experiments on the rolling bearing performance concern primarily with statistical evaluation of the fatigue life. With the development of aeronautic and astronautic undertakings, new requirements for the variability of the bearing performance, such as friction, wear, vibration, and temperature rise, are driven. Because the number of characteristic data is small and prior information on probability distributions and trends is poor, statistics has difficulty in assessing the bearing performance. How can the problem be solved There is a phenomenon in Nature, i.e., genetic variations of a species are likely to result in some strange change. Inspired by this, the paper advanced a hypothesis, viz., maybe some genetic variation drive the nonlinearly dynamical evolvement of the information poor process of the bearing performance. To demonstrate this hypothesis, the paper found the variant gene, presented the concept of the poor information equivalence relation, and proposed the method for construction of the information poor space to reveal the new properties of the nonlinearly dynamical evolvement of the modern bearing performance.