User blog comment:PsiCubed2/My own version of BMS/@comment-32697988-20181120222401/@comment-30846880-20181123132736

To @PsiCubed2, I strongly suspect that it is equivalent to the PSS portion of BM4 even without this correction, as long as we limit ourselves to standard sequences.

This matrix (0,0,0)(1,1,1)(2,2,1)(3,3,1)(4,2,0)(5,1,1)(6,2,1)(7,3,1) is standard and it is expanded differently between BM4 and yours. So I think BM4 and yours are different in the 3 rows BMS or over.

I think the issue came from your rule (3) which see only first entry to decide Y whichever row you analyze. BM4 sees previous entry.