Hi Uly,

Thank you for your comments. The Monte-Carlo module you described is more like an evaluation function, rather than a tree searching algorithm. Actually the "tree searching" part of the MCTS paradigm can involve zero randomness, such as the UCT algorithm, which is a totally deterministic ...

those resources are better applied increasing the engine's horizon, which is the most effective thing that can be done in the most common positions.

Is there any experiments/trials to support this conclusion? I would appreciate it if you could give some pointers to related reports. Actually I am ...

Hello,

Monte-Carlo Tree Search (MCTS) has been successfully used in several games, but seems less popular in computer chess so far. Is there some news about MCTS in chess? Anyway, even if it failed to generate strong engines, there should have been reports about this failure, right?