One of the oft encountered problems while applying TRIZ to real world problems is to identify from a large set of promising solution directions, the ones that are most promising, elegant or closest to ideality. This is especially true when the problem can be defined at multiple system levels, each level having multiple paths flowing towards the ideal final result. This paper describes such a situation encountered while analyzing a real world problem (rebar tying) with TRIZ. Steel reinforcement bars are tied together to form grids and reinforce concrete. Rebar tying machines have evolved over the years from pneumatic and mechanical systems to electro-mechanical systems with embedded electronics. TRIZ-based analysis and ideation resulted in a large number of interesting evolution paths for the rebar tying system. However, the question “Which paths would evolution favor” was difficult to answer. TRIZ does offer clues to pre-select the most promising paths in advance, but needs an integrated approach for the same. The paper describes such an integrated approach with the rebar tying case study as reference.