Examples of MCTS visualizations for the game Go, based on a naive built-in MCTS using KataGo's neural network. This feature will be available in LizGoban 0.8.1-pre2.
(This video is at 4x real-time speed.)
The result can be exported to a self-contained single HTML file.
The node colors represent estimated winrates (blue = high, red = low for Black). Use the i/o keys to zoom. Hover over a node for more details.
GitHub