I’d say the question is already potentially hard to answer.

We have deep learning systems that perceive and strategise. Who’s to say that AlphaGo doesn’t have some experience of playing out millions of games?