[closewith]

The models do do that, just at the next iteration of the model. And everyone gains from everyone's mistakes.