Ahh, turns out a line of code related to reverse indices & SP21 somehow got commented out. I've restored it and sent you an update. If it's OK, I'll update the site later today.

Incidentally, this is related to inverse split indices. Braun's indices show that 88v9 is a reverse index in Hi-Opt II. This really isn't true, at least in any practical sense. Hi-Opt II says split<39. This is the same as always split, since in the extraordinarily rare circumstance that the count is that high, you may as well flip a coin as the remaining cards are unpredictable.