If you think this works, apply a standard deviation to it and Z-score it
If what works? StdDev of what? I wasn't presenting a magic method of finding a random value produced by a uniform distribution. I'm not gonna debate another 50 pages on this subject, anyone can do whatever they want according to whatever they believe, even though it's wrong as shit after all.
It's tough to accept reality, so better to overcomplicate as hell something that already has the most efficient solution: check all keys as fast as it can possibly be done, knowing that no shortcuts exist, and any illusion of a shortcut inevitably ends up slowing things down, just by its pure existence of being put in practice.