Ray,

I would suspect (assuming the facts have you have presented are correct) that the odds on most probable explanation is that the investigating officer(s) had correctly identified (had strong suspicions relative to other suspects) Westerfield as the guilty subject before the administration of the various exams.  The influence of investigative bias is very strong and can lead to either a large number of correct results or a large number of wrong results depending on the nature of that bias.  It is for this reason and coupled with the fact that case investigators are generally pretty skilled at identifying guilty parties (leading to a higher percentage of guilty subject polygraph exams than chance would predict) that you and other polygraphers are largely blind to your actual/unassisted error rates in real field cases.  You should thank investigators every time you get a yearly job performance rating...