Two equally careless drivers can be judged differently because a pedestrian happens to cross in front of only one. Resultant moral luck concerns this dependence of assessment on outcomes outside the agent’s control. It separates the quality of a decision from the harm that follows it.
The comparison reveals several forms of assessment that ordinary language tends to merge. We can evaluate the drivers’ decisions, assign blame, recognize responsibility for an injury, and decide what repair is owed. Those judgments need not be identical. An unlucky driver may have exactly the same degree of carelessness as a lucky driver while acquiring a relationship to an injured person that the lucky driver never acquires. Control over the original choice and responsibility afterward come apart.