Isn't there a chance of some contrived results with 2 going through from groups of 3?