It indicates that people tend to overweight low probabilities and underweight high probabilities. The diagonal straight line in the figure indicates the linear weighting of probability. However, other works in this domain propose and experiment with parametric models which capture both inverse S-shaped as well as S-shaped probability curves