Excellent point BodMaster.
My first reaction was "that's silly" and "we've had different kinds of split-ups to counter that assumption", but...
Now that I think about it, the first time I noticed this was when I was the only one that split out and yesterday it occurred at least twice with about the same guys, but this time markv split out with me on both occasions,
but they weren't the
same guys, markv had made 31 over the weekend.
So indeed, in my experience, it's the 31s that sometimes are forked to their own branch of the continuum. And on all those instances that I've seen it happen it's the 31s that hosted the games (me on Friday and Saturday and markv on Sunday).
This may very well be a coincidence, but there is a pattern at least in these few occurrences that I've stumbled onto.
((Considering that a chance for a particular split-up of five is one in sixteen and three confirmed split-ups in the same certain way is one in 4096. And since you initiated the idea we don't have to fully counter that with human patter recognition abilities. Two things are for sure, we'd need more data to know and they'd need to fix this.
))
Bookmarks