It's possible to be irrational without realizing it. Such a person can then be expected to misjudge what they rationally ought to do. So it's unsurprising that in some situations (e.g. if one is sufficiently muddle-headed), one has little hope of acting rationally. On the contrary, we should object to any theory that implied the opposite. If a theory implies that people can always work out what they (rationally) should do, no matter how muddle-headed they might be, something has gone wrong. It can't be a true theory of rationality, because rationality is sufficiently objective to allow the possibility of a 'no-hoper' -- a person so incompetent as to be beyond the reach of rational guidance.
I bring this up because people are sometimes tempted to treat rationality in an excessively subjective fashion. (See, e.g., here and here.) People say things like, "How could Bob be rationally required to believe that P, if he has misinterpreted the evidence in such a way that he's led to think it supports not-P instead?" (Imagine Bob is a counter-inductivist. He notes that the sun has risen every day in the past, and misinterprets this evidence as supporting the proposition that it will not rise tomorrow.) The answer, naturally, is that Bob's mangling of the evidence has led him astray. It's true that Bob doesn't realize this. There are no alarm bells ringing in his head. But that doesn't mean Bob is somehow rational after all. It merely means he's unaware of his own incompetence. He really should believe that P - that's what the evidence supports - no matter his failure to recognize this epistemic requirement. Ignorance is no excuse.
All this is to say that we should expect the possibility of a situation in which we're beyond the reach of rational guidance. Again, I hear people object to some proposed rational rule R, "but in some situations I will lack the competence to reliably follow rule R; I confuse it with P and Q; my best attempts will not be good enough." This is true, but it is not an objection. Trying hard is not enough to make one rational. This is entirely to be expected.
Now, there may be multiple standards of rationality, corresponding to different degrees of objectivity or idealization. In addition to the ideal norms followed by a perfectly rational agent, we may also need non-ideal norms to guide imperfect agents like ourselves. But my point is that even these must have some minimal degree of objectivity to them. Any norm for which effort entails success is not any kind of rational norm at all. Any rule that can infallibly be followed (no matter how muddle-headed the agent) is too empty to qualify as a genuine rule of rationality. If there's any real substance there, then we must be prepared to admit the possibility of no-hopers: people who violate the norms without realizing it, and so predictably fail to understand rationality's recommendations - or follow its guidance - despite their best efforts.