Something I learned very, very rapidly when I started building things that other people actually used: one of the absolute worst things you can do is assume that the people using what you make either 1) know what they are doing 2) skimmed the manual/docs 3) are capable of doing anything that isn't explicitly stated in their list of things to do.
There are of course underlying reasons for these issues - lack of sleep, the marathon of raising children, the struggle of staying on top of bills, too many things to do in too few hours - but that does NOT invalidate them and you cannot design a system that ignores those issues and depends in part on human interaction and expect it to actually work consistently and properly.
A couple things I have read this past year suggest that mainstream economists are finally waking up to what people in tech support have known for decades. Never assume someone is acting rationally/has all the information/doesn't have extenuating circumstance, because you will probably be wrong.
There are of course underlying reasons for these issues - lack of sleep, the marathon of raising children, the struggle of staying on top of bills, too many things to do in too few hours - but that does NOT invalidate them and you cannot design a system that ignores those issues and depends in part on human interaction and expect it to actually work consistently and properly.
A couple things I have read this past year suggest that mainstream economists are finally waking up to what people in tech support have known for decades. Never assume someone is acting rationally/has all the information/doesn't have extenuating circumstance, because you will probably be wrong.