Take-the-best heuristic
Decide using the single best clue, and ignore everything else.
What it means
A fast-and-frugal decision rule for choosing between two options: search cues in order of how valid they are, stop at the first cue that discriminates between the options, and decide on that cue alone — ignoring all remaining information. By relying on 'one good reason' rather than integrating and weighting every variable, it trades completeness for speed and robustness. Strikingly, under conditions of limited data, high uncertainty, or redundant cues, take-the-best can match or even beat complex regression models, because it does not overfit the noise in a sample. It is a flagship example of ecological rationality: the heuristic's accuracy depends on the structure of the environment, especially a skewed distribution of cue validities. It matters as a corrective to the assumption that good decisions must weigh all evidence, and as a model of how experts actually choose under time pressure.
Examples
Asked which of two cities is larger, a person recalls one has a major-league team and the other does not, decides on that single most-valid cue, and ignores population hints, climate, or anything else.
A nurse deciding which of two waiting patients to see first stops at the first cue that separates them — one is short of breath — and never weighs age, history or notes.
Choosing between two used cars, a buyer checks the one cue she trusts most, a full service history, and buys on that alone. Mileage, colour and trim never enter it.
First described in Gigerenzer & Goldstein (1996).