So if the question, "Did you smoke marijuana while on parole?" received a "No" answer from Harmon Leon, isn't that a lie because he was never actually on parole?  So technically the "lie detector" caught nothing.  Am I missing something?

In each of the four polygraph "tests" I have had to endure the examiner solemnly told me that they, and not the machine, were the true "lie detector."  Each examiner assured me, even before they hooked me up to the machine, that they would easily detect any lie I told them, because they were highly trained at reading body language, pupil dilation and movement, and other physical cues to detect deception.

If that is true, I would think that detecting deception in someone who fabricated their entire background story would be ridiculously easy for an experienced polygraph examiner.  Apparently it is not.

I think it is even more ridiculous that they now claim that the exam was proven to be valid because it showed no deception on the question of smoking marijuana, which the subject did not in fact smoke.  Bravo!  The examiner detected no deception on any part of the entirely fabricated story, including the fabricated part about not smoking marijuana but being accused of doing so.  How does that validate the accuracy of the polygraph?