Parametric statistical distribution can be divided into groups of normal related distribution, exponential related distribution, bounded domain distribution, extreme value distribution, heavy tail distribution, quantile based distribution and systems of distribution. In each group there are a few distributions. The first scope for each distribution is to generate a set of pseudorandom numbers which follow that particular distribution. My questions are I cannot see and cannot tell the different among each group of distribution. What are the purposes for each group of distribution? For example, what is the different between exponential related distribution and extreme value distribution and among others? Is it based on type of data or based on type of application or based on type of industry or based on particular range of data or what? Where is the formula to look at? There isn't much explanation for each distribution. How to study and be familiarize with each distribution? under the scope, each distribution can generate a set of pseudorandom numbers. My question is what is the different between the set of pseudorandom number generated from each distribution? is it fall under a particular range of value or what. Secondly, if I want to generate a set of pseudorandom numbers, which distribution should I choose? what criteria or characteristic should I look at? what properties should I know? What are the guidelines should I follow in order to choose the correct and appropriate type of distribution?