Clearing Lines with Evolution: A Genetic Algorithm Approach to Tetris

Tetris is a challenging puzzle game requiring real-time decision-making and game strategy optimization. This research explores genetic algorithms (GA) to develop an optimal strategy for playing Tetris. The algorithm maximizes the number of cleared lines while minimizing undesirable grid characteristics, such as aggregate height, holes, and bumpiness. Heuristics evaluate each move, with the genetic algorithm evolving strategies across generations to optimize performance. This research experiment analyzes the influence of parameters such as the number of generations, mutation rate, and candidate replacement percentage on the algorithm's performance. The experimental results show that the best combination of parameters achieved the highest fitness value of 239. The findings demonstrate that GA-based approaches can optimize Tetris gameplay strategies with consistent heuristic weights across experiments.