There is usually at least a tiny grain of truth hidden somewhere behind even the most absurd belief. For example, if you lend a book to someone, you might as well kiss it good-bye. You won't get it back without asking for it back, and maybe not even then. If you're a library, you'll probably get it back; but for anyone else, lending books is about the same as giving them away.

What is the belief? People don't return borrowed books when they borrow them from other people.

There is a grain of truth here. Some people aren't good about returning borrowed books and perhaps that's most people, depending on who you lend your books to.

Is this true on average? I don't know and doubt that you know either. Even so, I believe it. How about you? Do people return books other people lend to them?

Now suppose that I ask you to lend me a book. If you think people don't return borrowed books, you might give the book to me, not expecting it to be returned. However, if you think people are good and usually return borrowed books, you still may not lend it to me, depending on the value of the book to you. But why?

When we make choices involving other people, we are always playing the odds, using our own idiosyncratic computational system. We attach odds to their doing or not doing something, the likelihood of their behaving one way or another, the probability of their reacting as we expect or surprising us instead. Our action, behavior or involvement with them then depends on those calculations. First comes the calculation and then we act, based on our nearly instantaneous decision process.

When dealing with other people, we may occasionally stop and think, do some research, carefully consider why we will or will not make a particular choice, and otherwise be more deliberate than we typically are. Even then, we mostly focus on our side of the equation. Will we be better off or less well off, what is or is not in our interest, what will be the consequences for us if we do or do not proceed? Most of the time though, we just go with our first calculation. We make a flash judgment about the person and the situation and then do or not do whatever the calculation calls for.

Do we often get burned or disappointed? If not, we are definitely toward the safe end, running the risk of being too skeptical, too mistrusting. If we often are disappointed or get burned, we are too far toward the other end where naive and gullible come to mind.

As we see, whether borrowing a book or having any other interaction with people, we infrequently give much if any consideration to the criteria we use with respect to the other person, especially if we haven't spent much time and effort in getting to know them. -- And we have actually invested that level of time and energy in only a very small minority of people we know or come into contact with. – For most people, most of the time and in most situations, we put them into our sorting algorithm, assigning them to "average," on whatever sorting criteria we are used to using.

How do we know what is average? Well, we usually don't. It's like people who borrow books from other people. On average, do they return the borrowed books without needing to be reminded? We don't actually know; but nevertheless, we likely have a quick criterion we instantly apply whenever anyone asks to borrow a book, particularly if they want to borrow our first edition of a rare book or perhaps our checkbook.

Here's the rub. It's that idiosyncratic computational system and its algorithm. The automatic criteria we use to judge people and to make decisions and choices is based on averages, as we understand them. The problem is twofold. First, "average" is a tricky concept. If on average, men are more violent than women, knowing that tells us nothing about most men or most women, because most men are not more violent than most women. It's only at the extreme that men are more violent than women. If we drop the extremes from our algorithm,