The take-the-best heuristic was first identified by behavioral scientist Gerd Gigerenzer and psychologist Daniel Goldstein. Gigerenzer and Goldstein bravely contradicted what the fathers of behavioral science Daniel Kahneman and Amos Tversky had said, and claimed that heuristics were not a sign of irrational decision-making, but an adaptive tool that helps humans deal with the vast amount of information they are presented with in the modern world.7
In 1996, Gigerenzer and Goldstein first introduced the take-the-best heuristic in their paper, Reasoning the fast and frugal way: Models of bounded rationality. Up until this point, economists and psychologists had put forward models of decision-making that assumed our minds (and the minds of animals) had unlimited time, knowledge, and cognitive ability, which allowed us to make rational decisions. Alternatively, Gigerenzer and Goldstein suggested that we can make rational decisions thanks to fast and frugal algorithms which help us to quickly sort through data and make optimal decisions.8
In their study, Gigerenzer and Goldstein used a computer simulation to compare the outcomes of two different decision-making processes: the take-the-best heuristic, or classic ‘rational inference procedures’ like multiple regression. They found that the take-the-best algorithm outperformed the classic rational algorithms, in speed and accuracy. They concluded heuristics do not impede rationally, but are accurate and dependable tools we can use to make optimal decisions.8
In 1999, Gigerenzer continued his research on the take-the-best heuristic and published a book with cognitive scientist Peter Todd, Simple Heuristics That Make Us Smart.9 In this book, the two introduced another take-the-best approach. In this alternative approach named the Take The Last approach, individuals don’t just pick the first cue that differentiates between alternatives — they pick a cue that has proven to work in the past. If in the past, a criterion helped you make an optimal decision, you’ll pick that cue instead of searching through possible cues.9 This approach suggests that individuals retain information on how valid a cue is, which we might sense as intuition.