def total_reward(state):
    return (
        win_loss_score(state)
        + piece_advantage(state)
        + format_compliance(state)
        - invalid_move_penalty(state)
    )