An adaptive pursuit strategy for allocating operator probabilities