User blog comment:Koteitan/Categorizing of the rule sets for all sub versions of bashicu matrix/@comment-11227630-20191122094441/@comment-30846880-20191124033618

p(j) only depends on the first row

Exactly. That's strange point of BM2. So I suggested BM2.3 to fix it.

the whole bad part ascending rule is different from this one.

I think the definition of that page is not equal to the BM2. I think it is same as BM2.3. Please ask to Ecl1psed276 or Alemagno12 about that.