> The "many more" and "random" ideas are made precise in measure theory (and elsewhere).

Well... one of the consequences of that precision is the theorem that there is no such concept as choosing a real number "at random".