If, after all of the normal interviews, you randomly selected candidates who passed and do not make them offers . . . you generally produce false negatives (assuming the interviews mostly worked).

But it's entirely pointless.