This paper presents an approach to improve pass prediction in Soccer Simulation 2D (SS2D) games using Deep Neural Networks (DNN) and Random Forest (RF). The authors propose an embedded data extraction module that records agent decision-making in real-time and generates features for training models. They apply four data sorting techniques to prepare training data and evaluate the performance of their models against six top RoboCup 2019 teams. The results show that their proposed methodology improves pass prediction accuracy by up to 10% compared to baseline methods. The study also examines the importance of different feature groups on pass prediction, finding that features related to the ball holder's position have higher significance. The authors also test the robustness of their models against changes in opponent team strategies and find that sorting methods like X sorting and Field Evaluator sorting enhance model performance. The study highlights the importance of feature engineering in improving pass prediction in SS2D games and suggests future work in exploring other models and scenarios.This paper presents an approach to improve pass prediction in Soccer Simulation 2D (SS2D) games using Deep Neural Networks (DNN) and Random Forest (RF). The authors propose an embedded data extraction module that records agent decision-making in real-time and generates features for training models. They apply four data sorting techniques to prepare training data and evaluate the performance of their models against six top RoboCup 2019 teams. The results show that their proposed methodology improves pass prediction accuracy by up to 10% compared to baseline methods. The study also examines the importance of different feature groups on pass prediction, finding that features related to the ball holder's position have higher significance. The authors also test the robustness of their models against changes in opponent team strategies and find that sorting methods like X sorting and Field Evaluator sorting enhance model performance. The study highlights the importance of feature engineering in improving pass prediction in SS2D games and suggests future work in exploring other models and scenarios.