For a long time, political science has been influenced by philosophy and ethics and, politics mainly limited to advise political rulers. After great and deep changes in the field of natural sciences, orthodox positivism dominated scientific methodology and very soon it was followed by social sciences including politics. Hereby, experimental and quantitative methods were used by social and political sciences.The article, beyond the extremes in both sides, seeks to study those areas in political facts and events which are preditable by applying statistics and mathematics while it also believes that there are areas which are not predictable. Therefore, “conditional prediction” is the concept which the article prefers to use in this regard.The linkage between political science and mathematics makes easy to formulate modles such as SWOT. Here, on the basis of time serials analytical modle, the author examines the nature and extent of predictability in politic’s.The main goal of this research is to study the interdisciplinary relation between political science and statistics in the framework of future finding. The time serials pattern as an analytical model is regarded as a link between past and future. This will help us to foresee different changes during time fluctuation. In this regard, conditional prediction is measured by:yt=T+C+S+Rand therefore, it seems more compatible with political facts and events.