Radial microspine grippers have been proposed for many purposes by NASA, such as low gravity mobility and asteroid operations. Specifically, the asteroid redirect mission concept consisted of a microspine gripper to carry a boulder from an asteroid to Lunar orbit. For manned missions to low gravity bodies, radial microspine grippers have been proposed to anchor hand-operated percussive drills to the drill target. Microspine anchors or grippers consist of many small hooks that grab onto surface imperfections, also known as asperities. Microspines can anchor to many surfaces, and have found use in climbing robots for the scaling of vertical concrete walls and rock cliffs on Earth. While the micro-scale physics of microspines are fairly well known, modeling gripping spine performance on the macro scale and in aggregate is still an active area of research. We propose the use of Bayesian networks to model uncertainties in the anchor surface and create a probabilistic model of where the radially arranged anchor spines will catch on a surface. A Bayesian network is a graphical probabilistic tool used to understand the structure of complex Bayesian processes with many random variables. This provides first-order estimates of the "grippability" of a surface patch, and can be integrated into future motion planning for evaluating or scoring potential grips.