The aim of this paper is to understand why the molecule has its own structure (optimized geometry) which differs from the standard one (norm). Basic strategy to obtain the optimized geometry is to start from the norm. The emphases are 1) a theory predicting polarization exemplified by norbornene HOMO distortion and 2) when we apply any theory of 1), what framework should we use? The latter is very important due to the fact that the optimized geometries of cations and anions differ greatly from the norms. For polarization, we have elaborated "in-bond orbital method." The method is based on the well-known correlation between separated atoms and united atom and able to represent polarization within minimum basis set only by the first-order perturbation theory. For cationic and anionic hydrocarbons, a new method consisting of an assumption and a few recipes is developed. Although the assumption that the framework distortion surpasses the orbital distortion (polarization) is based solely on organic chemist's intuition, this method correctly suggests the structure (framework) to which the in-bond orbital method should be applied. All of the optimized geometries used in this study are obtained by HF/6-31 + G*.