A neighborhood installed smart lights in a riverside park, expecting that brighter paths would make every visitor feel safer. Reports of minor accidents did fall near the main entrance. Yet visitors using the eastern trail still avoided the park after sunset because trees blocked several lights and the emergency signs were difficult to see. The city did not conclude that smart lighting was useless. Instead, it treated the early improvement as incomplete evidence. Workers trimmed branches, moved two lights, and tested whether people could locate the emergency signs from different points on the trail. The final report argued that a safety measure should be judged not by its intended benefit or by one improved location, but by whether the remaining obstacles had been identified and corrected across the whole route. The passage therefore moves from an expected benefit, through a limitation and its evidence, to a revised standard for evaluation.