In Nucl. Phys. B391 (1993) 127, the value of the oblique correction parameter S for walking technicolor theories was estimated using a technique called Analytic Continuation by Duality (ACD). We apply the ACD technique to the perturbative vacuum polarization function and find that it fails to reproduce the well known result S=1/6\pi. This brings into question the reliability of the ACD technique and the ACD estimate of S.