why is insert best done by splaying the would-be parent, splitting the tree, then attaching the two resultant trees as children of the node to be added? how is this different (better) than just doing a normal BST insert and splaying that to the top?
thanks 🙂