Judging by Resemblance
When we assess the probability that something belongs to a category, we often judge it by how much it resembles our mental prototype of that category — rather than by statistical base rates. This is the representativeness heuristic, another of the foundational heuristics identified by Tversky and Kahneman in their 1974 Science paper.
The heuristic is efficient: if something looks like a duck and acts like a duck, it probably is a duck. But it produces systematic errors when the resemblance to a prototype is misleading, or when it overrides more statistically relevant information about how common each category actually is.