Consider the following adversarial search tree expressed in minimax form.

(a) Draw the tree filling out the nodes with their appropriate minimax values.

(b) Now draw the search tree showing how alpha-beta pruning would work.

(c) Explain the resulting differences between the two search trees. Are any of the interior nodes different? If so, what would justify using alpha-beta pruning if the result is a different search tree? Based on the search trees, what is the best sequence of moves for the agent and his/her opponent?

(d) Suggest a simple modification to the ordering of the leftmost subtree that would reduce the number of computations in the alpha-beta case.