In this paper, an original approach is proposed to compare modeling and relevant statistical experiments using -Ti21S Body Centered Cubic (BCC) metal as a challenging benchmark. Our procedure allows the evolution of microstructural defects to be tracked in situ with excellent spatial resolution, while observing a bulk sample region sufficiently large to be statistically representative of the material. We identify multiple mechanisms such as slip transfer, slip traces, pencil glide, etc. We demonstrate that for small plastic strains ( < 0.25 %) under uniaxial tensile loading, the Schmid law is satisfied statistically. Under these circumstances, changes in Critical Resolved Shear Stress (CRSS) are minimal and accommodation of incompatible deformation between grains has not yet become important. The majority of the observed slip plane traces at the mesoscale corresponds to the {123} family. Fully automated while precise, the reported approach compares this data with four crystal plasticity models, and provides a methodology for similar analyses in other materials. * Corresponding author. X-Ray Diffraction HR-XRD, Electron BackScattered Diffraction EBSD…) that provide valuable atomic scale information, such as activation energies and critical shear stresses. However, one may argue that most multiscale strategies usually address only very specific cases, and that they are not necessarily statistically representative of the myriad of mechanisms that can occur in real polycrystals. For instance, one can cite HR-TEM [4] studies of special grain boundaries (structure, energy, migration under stress and temperature…) and their interactions with dislocations (emission, absorption, transmission…). Indeed, such available studies are far from covering the full range of possible grain boundary configurations in a real polycrystal. In addition, preparing and testing of thin foils for electron transmission-based techniques [5-7] (TEM, Transmission Kikuchi Diffraction TKD) can unfortunately affect the defect microstructure and the mechanisms investigated because of inherent size, strain rate, and external surface effects. Besides, these techniques do not provide enough data for statistical analysis of physical mechanisms.Experimental findings at an intermediate, mesoscopic scale, can be used to improve our knowledge of polycrystal plasticity and to improve models at the grain scale. They provide validation data for advanced numerical simulations that predict the mechanical behavior of polycrystals (strain hardening, ductility etc.