For the master theorem. The information we are truly concerned with is the rate of subproblem proliferation (RSP). Master theorem has three cases.
1. Same number of work with each level. Same as merge sort so O(n^dlogn)
2. Less work as we continue on, away from the top level so O(n^d) meaning the top level has the most effect on the amount of work and thus O.
3. As we continue down the tree, the leaves produce more and more work resulting in the amount of leaves the algorithm has contributing to the O so it is O(number of leaves)
