Abstaining from prediction was appropriate because the observation set was insufficient and matched a known bias pattern. Trying to make predictions based on this would likely lead to errors.