You are mixing up “Devils advocate” with “prove the negative for me.” The point of Devils advocate is to test assumptions, not to accept the first correlation as gospel.

If pedestrian and cyclist deaths rise 80% and 50% while vehicle size, road design, lighting, speeding, and impairment trends also shift, then asking whether those factors matter is not “sowing doubt.” It is literally how causal analysis works. If your position is that questioning causality is illegitimate unless I hand you a fully formed alternative theory, then you are not defending evidence. You are defending certainty.