The field of Monte Carlo Tree Search (MCTS) is moving towards improving sample efficiency and adapting to different game environments. Researchers are exploring new strategies for choosing exploration constants, developing novel abstraction algorithms, and enhancing existing frameworks to detect state equivalences and group nodes with known value differences. These advancements aim to increase the accuracy and robustness of MCTS in various domains. Noteworthy papers include: Investigating Scale Independent UCT Exploration Factor Strategies, which introduces a new strategy for choosing the UCT exploration constant that outperforms existing methods. AUPO - Abstracted Until Proven Otherwise: A Reward Distribution Based Abstraction Algorithm, which presents a drop-in modification to MCTS that improves its performance. Grouping Nodes With Known Value Differences: A Lossless UCT-based Abstraction Algorithm, which proposes a new abstraction framework that detects significantly more abstractions than existing methods. Discovering State Equivalences in UCT Search Trees By Action Pruning, which provides a weaker state abstraction condition that trades a minor loss in accuracy for finding many more abstractions.