The data does not support that a student already with a 31 has a good chance of improving. When it does happen the improvement is a point. You have to weigh the risk of it going down compared to benefit of a 1 point improvement. I don’t think the decision in many situations hinges on a 31 vs a 32.