Merilaita and Lind (2005) used artificial prey and an artificial background, but real avian predators, to test the hypothesis that background matching is maximized when coloration visually matches a random sample of the background. They used two types of prey patterns as experimental stimuli, one type judged difficult to detect and another type judged easy to detect. Prior to the actual experiment, "ease of detection" for the two prey types was judged by