If this turns out to be a false positive from contamination, how many man-month do you estimate will be wasted on making models that don't describe reality before the error is caught?