Is there any way you can preset a search path in your algorithm? Since these are heuristics based, it wouldn't be a bad idea to walk a finite subset of the game

Is there any way you can preset a search path in your algorithm? Since these are heuristics based, it wouldn't be a bad idea to walk a finite subset of the game tree for each side and ensure the "best decisions" are made. You could also test subsets of the algorithm's nodes to determine they are pruned properly. My background is mainly in NLP, so I only have the basic knowledge of game playing, but usually bootstrapping manually can and is the best route.. Usually tests are deterministic, but when you start throwing probabilities in the mix, a max likelihood/score for anything usually comes down to a hand crafted situation..