people come to me all the time complaining about the number of joins in the data vault. i thought i might take a crack at answering this question/complaint, and provide some solid mathematical proof behind the fundamental design of the data vault model. i welcome your feedback (i hope you take the time to comment).