A score produced by any predictive model must be taken with a very particular grain of salt. Scores speak to trends and probabilities across a large group; one individual probability by its nature oversimplifies the real-world thing it describes. If I were to miss a single credit card payment, the probability that I'd miss another in the same year may quadruple, based on that factor alone.
