The paper written here regards a method which invokes a certain principle to reduce the content of a collection of data. The original content can be replicated in the exact same way. Meaning, there will be no data loss after the procedure. In this method, no data is needed to be stored as guidance/requirement for the reproduction of original data. This means that this method can be repeated to the resultant of the previous procedure done using this method and so on. For multiple processes, the number of times this method have been applied has to be stored (and obviously, the final data set in reduced form) to get the original data set. This method first converts a given data set into another data set which is termed “special” (how is it special?). This special data set can be reduced considerably to another data set of lesser content than the original and so on. The main idea here is to convert to that special form. Note has to be taken here that the reduction is reasonable which fits one practical criteria (that is a reasonable reduction). A particular pattern in these special numbers will be shown.