This is statistics, and it deals with trends and outcome probabilities, not with individual predictions. As Murray and Herrnstein wrote, they could NOT predict any individual student's fate, but they could perfectly predict the trends and tendencies: in a class with 30 imbeciles about 10 will end up in the criminal justice system - this is predictable; which ones will end up there is not. A similar picture obtains with higher IQs: on the average, they tend to do well [still, not all of them do]. But which ones will fall through the cracks is not easily predictable. And on the same lines, IQ [statistically] tends to beat environment [again, not always, but in most known cases].