The basic insight of the literature concerning the Median Voter Theorem and its applications, dating back to Black (1948), is that the policies implemented by di¤erent parties once in o¢ce should approach the median voter's preferred policy if they run in a single-dimensional and democratic electoral space. This strong prediction has been challenged in recent years using arguments related to the observation that usually the political spaces concern much more than one single dimension and that, once we consider such a space, the Median Voter Theorem cannot be applied. Our idea is that one can challenge the median voter predictions even if we keep considering just one single dimension. Infect also in electoral competitions characterized by a very important issue seen by voters as "salient" is almost impossible to observe in reality the convergence predicted by the Black's theorem. In the present model we introduce a simple assumption over the process of opinions' formation of the voters and we show as in equilibrium strategic considerations lead the parties to choose polarized platforms.