Future Chess Program

[Kid Chester Asuncion]

Dear LeelaChess Team,

I am writing to you today to propose an idea that I believe could greatly improve the functionality and user experience of your computer chess program. Specifically, I am suggesting that you consider implementing a feature that allows the program to adopt the playing style of a specific human chess player based on their moves.

As a chess player myself, I have always been fascinated by the different playing styles and strategies employed by various players throughout history. For example, the aggressive and tactical style of Mikhail Tal, the positional mastery of Tigran Petrosian, and the aggressive and uncompromising style of Bobby Fischer, just to name a few.

To achieve this feature, I propose that you use a neural network algorithm that is capable of learning and mimicking the playing style of a specific player. The algorithm should be trained on a set of PGN chess games from the selected player, and it should be designed to analyze and understand the player's moves, their style, and their strategies. By doing so, the algorithm can then adopt the player's playing style and replicate it in future games.

Furthermore, I suggest that you also use an explanation algorithm that can explain every move played in a game in natural language format. This algorithm should be designed to analyze the position on the chessboard and provide insights into the strategies and thought processes of the players involved. This feature would greatly enhance the educational value of the program and provide users with valuable insights into the strategies and thought processes of their opponents.

I understand that the implementation of these features may require significant resources and expertise, but I believe that they could greatly enhance the value and user experience of your program. Thank you for your time and consideration, and I look forward to hearing your thoughts on this proposal.

Sincerely,
Kid Chester Asuncion

[brian.p.r...@gmail.com]

Neither is practical today.

No human has played even close to enough games for current net training approaches to learn a particular style.

Obtaining human readable explanations of how large nets come up with answers remains a very challenging research topic.

[sme...@gmail.com]

I'm not sure this is true. The Toronto team has made a lot of progress in predicting individual player style since the original Maia concept. E.g. see this new paper: 

"We demonstrate that we can significantly improve prediction accuracy of a particular player’s moves by applying a series of fine-tuning methods. Furthermore, our personalized models can be used to perform stylometry—predicting who made a given set of moves—indicating that they capture human decision-making at an individual level." 

[brian.p.r...@gmail.com]

Matching moves against a set of players to pick the most likely player to have made a move is not the same as playing with a particular human player's style.

Training nets to adopt the style of an individual human player has been tried several times with Lc0.

The results are "fair" but not "good", IIRC.  Moreover, the nets are quite weak relative to the top strength nets, but that does not matter all that much in this case.

Accordingly, this area remains a topic of research.