The basic problem with lie detectors is that if you know what you're doing you can trick the machine. Also, people tend to react to the sensitive questions, pr

The basic problem with lie detectors is that if you know what you're doing you can trick the machine. Also, people tend to react to the sensitive questions, producing false positives. While they aren't used for it they can in certain circumstances produce results with an accuracy quite sufficient for the courtroom, and there isn't even any Fifth Amendment problems with the approach: You need things which the criminal will know but which others will not--you have to keep a bunch of details out of the press. The suspect is not asked any questions, but rather asked to read a bunch of statements about the crime--"she was wearing a red blouse", "she was wearing a green blouse", "she was wearing a white blouse", "she was wearing a blue blouse", each statement is innocent and thus will not trigger a reaction in the person who doesn't know the details. There is a 1 in 4 chance that someone will show the strongest reaction to the true statement--but what happens when you do the same thing with a bunch of details? Show the strongest reaction to the true one on 20 such 4-choice questions and now what are the odds it's a random? An innocent can't incriminate themselves because they don't know which are the sensitive answers. Note that you can't exonerate someone this way--someone who doesn't remember the details being asked about will look the same as someone who is innocent. However, if it fingers you to a sufficiently high degree of certainty guilt is assured.