User blog comment:Koteitan/Categorizing of the rule sets for all sub versions of bashicu matrix/@comment-11227630-20191119133850/@comment-11227630-20191121141123

Well, the strange point in the BM2.2 shown here is that

(0,0)(1,1)(2,2)(3,1)(2,2)[3] = (0,0)(1,1)(2,2)(3,1)(2,1)(1,1)(0,1)(-1,1)[4], with Δ = (-1,0).