With respect though, you are talking intentional violations.

I'm talking all violations. Given that a certain number of the travelling public will fail security even if it was 100% efficient (i.e. at times we all screw up) the number of unintentional violations (and in that I'll include those who know they are violating but are doing so for reasons of convenience (e.g. my smuggled lipsalve ) rather than for ill intent) is likely to be vastly greater than intentional violations.

But, security can't actually tell an intentional violation from an unintentional violation at the moment of detection (or moment of failure). So (from my point of view) you have to include both types of violation in any failure rate.

Of course, having the same people going through security more than once messes up your independence a bit, since if they are caught on the first pass through, it's highly probable they will be clean on future passes. So repeat people passing by will slightly depress the failure rate.

Right, my head hurts now I think we're getting pretty esoteric now, and I think we aren't too far away - it's just a difference in interpretation of how you assess failure.