Would the most likely conclusion on US arms contributing to killings not just be based on the % of arms sold? i.e. 69% of those killed will most likely have bee

Would the most likely conclusion on US arms contributing to killings not just be based on the % of arms sold? i.e. 69% of those killed will most likely have been as a result of the US providing the arms? I'm sure there are other ways of performing that analysis, such as by looking for identifiable fragments, but that seems unnecessary. If you add the financial funding in to the analysis the number would surely go up significantly to a point is another consideration.