The 100 win 0 loss automated binary robot
Only a quantitative analysis, such as that offered by QT est, reveals that DM1's single violation of modal choice is more serious than DM13's four violations combined. This teaches us that superficial descriptive indices are not even monotonically related to quantitative goodness-of-fit, hence can be very misleading. Note that QT est is also designed to avoid the proliferation of Type-I errors that a series of separate Binomial tests creates, since it tests all constraints of a given probability model jointly in one test.
The modal choice test is also useful for the quantitative decision scientist: Since the modal choice prediction is rejected for DM1, we can conclude that constant error models and a very general class of econometric models of.