User blog comment:PsiCubed2/My own version of BMS/@comment-30846880-20181121133049/@comment-30754445-20181121140437

"Regarding bad root selection, I think Buchholz Hydras (BH) works as same as BM2 and BM4"

Let's hope so.

Because it is quite clear to everybody how pair sequences are supposed  to work. This was already clear when BM1 was proposed. The only question is whether, for a given version BMS, the rules actually work as intended.

So either BM2 and BM4 work as they are supposed to, or they don't worsk at all. If they work as they are supposed to, then they should reduce to simple hydras (and as a collary - they should give the exact same results for PS).