Much of the philosophical discussion about the interaction between theoretical and practical reasoning, or, as I will put it, between prediction and decision, tends to focus on what I will call the unhappy cases: cases of apparent irrationality, irresponsibility, or bad faith. One familiar character is Professor Procrastinate, who predicts that, due to his inveterate procrastination, he will not write the review he otherwise has strong reason to write. I would like to consider, instead, the happy cases: cases in which you predict you will do what you have strong reason to do. I hope thereby to better locate what has gone wrong in the unhappy ones. My ultimate quarry is the idea that prediction and decision, or theoretical and practical reasoning, are somehow isolated from one another, that they take place in different “standpoints,” such that moving between them involves a shift in “point of view,” and any attempt to combine them is somehow illicit or a sign of bad faith. The idea is inherited from Kant, who sought to render science and human freedom compatible by arguing that the activity of the rational will cannot be known by the methods of science. Kant was addressing the problem of free will by appeal to transcendental idealism and in-principle mystery. But the contemporary legacy of Kant’s thought, in the metaphor of standpoints, leaves the idealism behind and extends to other contexts, where it brings more confusion than clarity. Or so I hope to suggest.