Is libertarian free will a necessary assumption for any decision theory because “one has to suppose that one has a choice to make”? It seems to me that decision theories needn't rely on the formulation that agent1 can x or ~x regardless of preceding states-of-affairs, but that it can equally well rely on agent1 x'ing because state-of-affairs1 determines that agent1 x's or agent1 ~x's because state-of-affairs2 determines that agent1 ~x's. The point is not whether the agent can make that decision with exactly the same preceding state-of-affairs, but whether the agent could make both decisions, however the decisions are brought about.
My reason for this position is that if libertarian free will is a necessary assumption for any decision theory is correct, determinists should not make normative suggestions ever. One would suggest this perhaps, because they'd claim determinists cannot ever make normative suggestions coherently. I believe this is wrong because normative suggestions a determinist makes could potentially be the difference that makes a difference between agent1 x'ing and agent1 ~x'ing – hence, a determinist can prescribe that an agent makes one decision instead of the other because it may be that with the additional argument for the former decision, a person would make said decision where they otherwise would not have. This strikes me as fairly clear: determinists are not automatically barred from making normative prescriptions on decision making.
My question, then, is whether there is something more to what a defender of libertarian free will's necessity in decision terms that I am missing or whether my approach to decision making from a determinist's perspective is severely flawed.