The ChessAIThon project (2025-1-ES01-KA220-VET-000354329) is co-funded by the European Union. The views and opinions expressed in this publication are those of the author(s) only and do not necessarily reflect those of the European Union or the Spanish Service for the Internationalisation of Education (SEPIE). Neither the European Union nor the National Agency SEPIE can be held responsible for them.
Table of Contents
Pedagogical Strategy (Comparison with Brute Force)
Teachers can compare brute force with the task of memorizing every possible variation of openings and defenses: it is immensely costly and often ineffective beyond the first few moves due to the complexity of the game. The arrival of AlphaZero marked a paradigm shift by using a generic Reinforcement Learning (RL) algorithm to learn tabula rasa (from scratch), without pre-existing human knowledge, only the rules of the game. AlphaZero trains a deep neural network that guides its search. Interestingly, its search is much more selective and slow, analyzing only 80 thousand positions per second, but compensates with a much more accurate evaluation of promising variations.