Bloodstain pattern analysis (BPA) is an integral part of crime scene investigation. For violent crimes involving gunshots, standard practice in police departments worldwide have some physical limitations. For instance, the effect of gravity and air drag on trajectories of blood droplets are neglected using current reconstruction methods, which results in a well-known overestimation of the height of the source of blood. As a consequence, more sophisticated models for blood spatter trajectory reconstruction are being developed, two of which are highlighted in the present work. They allow the prediction of bloodstain patterns produced from backward spattered blood droplets from blunt and sharp bullets. Our recent models attribute the splashing of blood to the Rayleigh-Taylor instability which arises when blood is accelerated towards lighter air. This physically-based description comes with the powerful predictive capability to correlate features of bloodstain patterns with the specific bullet and gun that produced them, as well as with the body position. The results of the numerical models were compared with four experiments simulating blood spatter deposition on a vertical wall through the number of stains produced, average stain area, and average impact angle at the surface, and the agreement found is fairly good. Moreover, further insight is obtained by probing and explaining the influence of observable parameters on the resulting spatter pattern, with the goal of aiding BPA experts evaluating a crime scene.