I mock your claims further:
I can burn chicken bones to detect lies. To claim that the ash patterns could never detect a lie is just flat out wrong.
Sure. That means they are imperfect. It does NOT mean their results have "no connection" to lying. Burned chicken bones are not perfect. Their accuracy is far below the "reasonable doubt" threshold needed for evidence in court. But to extrapolate from that and claim that they don't work at all is nonsense. They are "good enough" for preliminary screening.
Hey, flipping coins will get you 50% accuracy, so a polygraph is at least that good, right? Can you at least find a study that that proves that?