Do all the possible computations on very large matrices look alike? Nobody knows, but let's play a game and see how far we get!