Efficient Game Solving Through Transfer Learning