While justifying that independence is a canonic coupling, the authors show the existence of a second equilibrium to reduce the information conveyed from the margins to the joined distribution: the so-called indetermination. They use this null information property to apply indetermination to graph clustering. Furthermore, they break down a drawing under indetermination to emphasis it is the best construction to reduce couple matchings, meaning, the expected number of equal couples drawn in a row. Using this property, they notice that indetermination appears in two problems (Guessing and Task Partitioning) where couple matchings reduction is a key objective.