I have a minimax tree as follows:
The number of branches can get extremely large (C(k,n)*C(n,j) to be specific), even though the depth is constant and very low. Is there any way to tackle this problem intelligently rather than going over each possible node? I can generate the value of each possible node but even generating that value might consume some time. Hence I am looking for a intelligent strategy that allows me to find the optimal solution very quickly.
I checked alpha-beta pruning, but it seems to be useful when the depth is higher and seems to have no impact when depth is very low yet the number of branches is very high.
Additional question 1: Is there an appropriate heuristic to this problem?
Additional question 2: Can applying alpha-beta pruning useful when we have to generate each node value as we go on? In my case, the computational problem is to generate all node values. Hence, I generate node values as I go on. But in alpha-beta pruning, it seems that all the node values are known a priori. Do you think is there a way to apply it to my case as well?
P.S: Sorry for the terrible paint representation, it's the best I can do in short time.