User blog comment:LittlePeng9/Random Turing machines/@comment-5150073-20130406103030/@comment-1605058-20130406110234

Proving that BB(100,2)>BB(55,6) would be very hard, if not impossible. I know way to find x such that, given a,b and c, BB(a,b)>=BB(c,d), but for 2,55,6 it'll give something around 600. I think it'd be easier to redesign my machine to work with 2 colors. But I don't think it'll be easy to fit within 100 states