The article is a short review of interpretation of probability and its significance for risky decisions. It especially deals with a concept of subjective probability and decision weights introduced by the prospect theory. While dealing with both subjective deviation from objective probabilities as well as subjective probability transformation as probability weighting function it was observed that people treat low probabilities differently. In both situations people tend to overestimate rare events. Recently it has been said, that in the decision based on direct observations of frequency of events, overestimation is much weaker.