I'm working on a list of question types which come up frequently in day-to-day life but which I haven't yet found a reliable, rational way to answer. Here are some examples, including summaries of any progress made in the comments.
The third request in the Serenity Prayer is for "the wisdom to know the difference" between things we should accept with our serenity and things we should change with our courage. Pending an official response to the prayer, what are some rational criteria for deciding between those two responses to an unfavorable situation?
Practice the ability to judge how important something is to change, making sure to examine your criteria of importance. Identify the reasons you want to change it, and try to normalize your emotional response to the facts. Learn about the difficulty of changing a thing by investigating other peoples' attempts to. Be aware that, the less one knows about a field, the less one is able to judge how difficult a task in that field is. Ask an expert if you need to. Another heuristic for difficulty of changing something is that the closer it is to one's own mind, the more control one has over it. When you know as much as you can, do a cost-benefit analysis.
I know that asking for what I want is often the best way to get it ("Will you take your hat off so I can see the screen, please?"), but it's sometimes clearly not appropriate ("Will you please give me all your money with no expectation of benefit nor repayment?"). Those are two ends of a spectrum. When the thing I want is somewhere in between ("Can I have a little of your time to vent about something that's bothering me?"), how should I decide whether to ask?
Unreasonable requests are those which would only be fulfilled if the asker had power over the askee which they do not, which represent an unequal exchange between equals, or which are not actually possible. We don't want to make unreasonable requests because they are at best unfair social impositions and at worst rude and damaging to relationships. This is complicated because requests between friends aren't about direct exchange; it's expected that sometimes one person will need help, and sometimes the other will, and in the long run it'll even out. In a strong friendship where both people have treated each other well, it's more appropriate to ask for a large favor than it would be to ask a relative stranger the same thing.
The difficult requests to judge are those where the power balance or strength of the relationship, or the values of what's being exchanged, are unclear. That is, they require a more accurate judgment of either the relationship itself or of the other person's needs and abilities than one is confident of making. As in the previous question, specialized knowledge can help predict how much trouble a given request might be. In the specific example above, one would need to understand the cost of lending an ear, in terms that could be compared to the benefit of venting. Prior communication is the best way to achieve this; basing an estimate on similar past situations is also good. Knowing nothing else, use an assumption of equality and the basic responsibility for oneself as heuristics. Finally, how close you can get to the edge of what's acceptable may depend on how much you trust the other person to tell you if your request is not reasonable (rather than acquiescing resentfully).
How do I balance the need for comfort and short-term happiness (making it mentally easier to be useful) with that for productivity and long-term happiness (setting me up to be happier and more useful later)? Again, some examples are clear-cut: it's almost certainly a better idea to do my homework at some point than to spend the entire week playing video games. A more difficult case is sleep. Get more of it and feel more rested and alert, or get less and have more time for fruitful tasks?
It's possible to minimize the necessity of choosing between these two things by doing work which is enjoyable and taking breaks after earning them. When the two types of activity do conflict, one way to get around it is to use time that wasn't available for work anyway to do unproductive things. Another is to have (and frequently review) clearly defined medium- and long-term goals, and weigh short-term choices against them. Doing this regularly may make it easier to judge activity choices on the fly.
Go players describe plateaus that last for months, but which they eventually climb out of again. This creates a sort of human halting problem: How do you tell the difference between a long plateau without improvement, and actually having reached your peak in a skill?
No comments yet, but here are some questions this raises for me: What does it mean to have reached your peak in a skill--is there actually a maximum amount you can usefully learn and practice, or just a (potentially variable) point of diminishing returns? Is it possible to know there's more to learn but not be able to learn it?
The William James zone is the positive feedback loop of mental and physical anger responses which keep you a person even after the conflict has been addressed or resolved. I find myself in the WJZ sometimes when I remember or anticipate something which made/would make me angry, even when no conflict is presently occurring. This happens primarily when I don't have a ready distraction from the upsetting thought, e.g. when I'm in the shower or waiting to fall asleep. Other than simply waiting for it to pass, how can I get out of the WJZ or avoid entering it?
So far, the only mitigating factor I've found is my overall physical and mental state. Being hungry, tired, or stressed makes it easier to fall into the anger cycle and harder to get out. Therefore, taking care of myself in general helps to prevent it, but it's not always possible to remedy those problems after the cycle has already started. When circumstances permit, physical activity may provide an outlet for the energy that keeps this cycle going.
Where is the line between acceptable nonverbal communication and unacceptable manipulation? Is it in the thing being sought, the manner of seeking, the intent of the communicator, or something else?
A conversation offsite led to the following: Manipulation involves both deliberate instigation of emotion and trying to persuade someone to do what you want, but isn't defined by either. (The first one describes gestures of affection, and the second includes ordinary debate.) The definition we settled on was "using emotion to bypass someone's normal decision-making process." Trying to get someone to do what you want is not inherently manipulative; trying to make them feel something so that they will do what you want is.
Naturally I'm looking for ideas about how to answer these questions, including links to earlier thoughts about them, but you get bonus points for supplying actually usable heuristics, rather than just opining on my examples. But I'd also like to hear it if you've got any questions of your own that fit this form. Consider it a sort of lowbrow subset of open problems--difficulties you're aware of having on a regular basis but haven't yet been able to solve.
(Tag suggestions are appreciated. I'm unaccustomed to using content tags, so I made some guesses based on the site's tagcloud and what's on the similar Open Problems post.)
 I actually quite like the Serenity Prayer, despite being entirely nontheistic, because it presents a set of traits to aspire to for specific purposes I can get behind.
 Until I've read the entire LW archive, I'm constantly paranoid that anything I post will be a second-rate rehash.