Models are useful in business as in many other applications, such as science. Let’s say that two groups of analysts have created models to predict interest rates, and both models use the same input data. One model predicts higher interest rates for next year, while the other predicts that rates will drop. You are unable to examine their inner workings, but you must evaluate them today to tell which, if either, is right (both might be wrong!). How would you do this? What is the practical significance of this evaluation, if any?

Models are useful in business as in many other applications, such as science. Let’s say that two groups of analysts have created models to predict interest rates, and both models use the same input data. One model predicts higher interest rates for next year, while the other predicts that rates will drop. You are unable to examine their inner workings, but you must evaluate them today to tell which, if either, is right (both might be wrong!). How would you do this? What is the practical significance of this evaluation, if any?