“…To various orders, fully explicit results have been computed in [28][29][30]. Both M (1,0) and M (1,3) come with interesting properties which are simple to see in our formalism: We would expected M (1,0) to be function of all variables {ŝ, t, š, ť, p 1 , p 2 , p 3 , p 4 } but it happens to depend only on the specific combinations s = ŝ + š and t = t + ť, u = −s − t − 4. Moreover, it does not depend explicitly on the charges p i at all.…”