Numerical experiments demonstrate that the accuracy of stretching force constants k(e) provided by Badger's rule is unlikely to be substantially improved either by modification of the functional dependence on the equilibrium bond length R-e or the inclusion of bond parameters related to electron density. These results, based upon both the experimental and QCISD/6-311++G(3d2f, 3p2d) values of R-e and k(e), imply that most of the universal characteristics of the bond strength vs. bond length dependence are accounted for by Badger's rule, the more detailed features being unexplainable by first-order response properties such as electron density.