Abstract: In perceiving the random sequence, people incline to consider two actually independent outcomes as positively related or negatively related, which are the hot hand fallacy or gambler’s fallacy, i.e. the positive or negative recency effect in the random sequence. In the exploration of the psychological mechanism of those two fallacies, Kahneman and Tversky adopt the local representativeness heuristic. Since their explanation has several shortcomings, several new models (i.e. casual model, Urn model, memory model, adaptive model and two stages Gestalt model) are presented to explain the two fallacies in different ways. In the conclusion the authors point out that the research methods, neurotic mechanism, individual differences and the application in daily life would be the possible development in this field

Key words: random sequence, recency effect, hot hand fallacy, gambler’s fallacy