Originally Posted by
Busbuoy
I don't think the inclusion of "average" here implies any such thing about the failure "rate". While the acceptable rate of failure may well, and definitely should, be specified elsewhere in the legislation, in this case surely this is merely specifying that procedures to cope with the failure of any component must not depend upon techniques that the "average" pilot might not be able to master within the course of their career development and specific training to be approved to fly the aircraft in question.
To your point about the vagueness of the "average pilot" specification, attempting to more tightly define this at the legislative level would be a herculean task I would have thought. To date we have relied on the industry ensuring, as much as possible, that "below-average" pilots do not progress into flying positions that require "average" skills. While this sounds subjective, in my experience (as a military instructor pilot and senior international airline captain) I am very comfortable with the way "averageness" was tested non-subjectively. No doubt this is imperfect and boundaries will always be tested by the unscrupulous, but I wish the best of luck in the world to anyone who would try and codify in a central document exactly what the skills of an average pilot should look like. We should never let up on the unscrupulous 'tho.
Yep, that's why I say, to move away from these vague certification criteria. The moment you can use a ruler to check on certification compliance, you give management far fewer chances to cheat on these things.
"Average" has little to do with predictability, more with a large sample result calculation/evaluation afterwards. But, hey, certification is not about registering history, though setting rules, how history should develop. Average has no place in that.