Sure but my point is that the Game Theory approach does not helps us in building better Chess playing agents. E.g. how color is used in photography might make a very interesting paper, but from perspective of building an image recognition model its irrelevant.