This is constantly overlooked in debates over IQ, SAT, etc. In most cases where these tests are used, it is far more mportant to separate the above median from

This is constantly overlooked in debates over IQ, SAT, etc. In most cases where these tests are used, it is far more mportant to separate the above median from below median, or +2 STDDEV from median, etc, and not to pick winners in the Krelboyne Super Bowl. Even assuming there is a total ordering of intelligence, how would a single 300-question test rank 300 million people? Why fill half the test with questions that 99% of takers will get wrong, just to differentiate among the top 1% of scorers?