I had previously put out some videos in a similar style on the channel but wanted to see how far I can get the visualizations and how much I can automate it with python. The biggest time sink when creating these videos is adjusting the individual parts of the video using matplotlib, which I can mostly reuse across videos. Here is what I show in the video beside the chessboard:

The bottom visualization shows the engine evaluation across different depths, which allows you to (1) follow the engine evaluation of the game and (2) compare evaluations across depths to find surprising moves (which I originally implemented for this repo: https://github.com/CYHSM/chess-surprise-analysis).

The right lineplot indicates the number of games which have previously reached the highlighted position. This allows you to see when a completely new game starts (Thanks Agadmator) with steep drops indicating more exotic lines. The title in the plot also indicates the opening name, for as long as they are identified.

Please let me know what you think and if I should upload more videos in the same style.