The sense of touch is represented by neural activity patterns evoked by sensory input forces. The rodent whisker system is exceptional for studying the neurophysiology of touch in part because the input forces can be precisely computed from high-speed videography of whisker deformation. The standard model of rodent whiskers assumes quasi-static dynamics and a conical profile with linear taper. Here we evaluate the validity of these assumptions by comparing whisker shapes computed using the standard model with experimentally-determined shapes during controlled deflections. We find significant discrepancies between model and experiment. Real whiskers bend more than predicted upon contact at locations in the middle of the whisker and less for distal locations, which affects the amplitudes of forces during tactile sensation. On the other hand, real whiskers maintain contact for greater push angles during interactions near the whisker tip, which allows rodents to collect greater information from objects and surfaces prior to whisker slip-off. Our data imply that precision of computational models relating sensory input forces to neural activity patterns can be quantitatively enhanced by taking non-linear taper into account with a provided corrective function.