Key points are not available for this paper at this time.
The effects of a lucky event and irrational beliefs about luck were examined. In two experiments, some subjects experienced a lucky event, whereas others did not. All subjects then completed an unrelated decision task rated their confidence, and placed a bet. The effects of a lucky experience depended substantially on individual beliefs concerning the causal properties of luck. After the lucky event, those who believed in luck (i.e., thought of luck as a stable, personal attribute) were more confident and bet more, whereas those who did not believe in luck (i.e., thought luck was random) were less confident and bet less. A third experiment identified analogous effects using multiple-choice test questions that included a monetary penalty for errors. Increased expectations following initial luck were interpreted in terms of a lucky streak effect, whereas the paradoxical decrease in expectancy was viewed as an instance of the gamblers' fallacy.
Darke et al. (Tue,) studied this question.