Sure, I think it makes intuitive sense to me that you should play riskier when you're behind. The surprising part to me is that when you're ahead, even if you know that your opponent will play "sub-optimally", that doesn't change your own optimal move.