we should clarify to ourself how the g2 incidence geometry axioms work... not quite correct to say that given a generic pair of configurations there's a unique way of getting from one to the other in 3 rolls...