Making the choice

This chapter is a method. It's deliberately plain. I've put it together from the research in the earlier chapters and from common sense, and nobody has tested it as a whole. It takes about an hour for a large decision and a few minutes for a small one.

Write the musts. On paper, list what the choice has to do for you. Keep it to four items at most. The point is to separate the things you need from the things that would be nice. For the sofa, it has to fit the alcove, survive the dog, cost under nine hundred pounds, and be comfortable to lie on. For the course question the list is harder to write, and an honest one might read like this. My work should interest me on most days. We can still pay the mortgage. I'm home at least three evenings a week. If a fifth item wants to come in, one of the four goes out. The weighted spreadsheet fails here, because with twelve rows every option is good at something and the totals come out level.

Cap the options at three. Chapter 5 gave the reason. Include the default as one of them, written out as an option with its costs, as in chapter 3.

Check each option against the musts. Give each must a yes or a no. Don't give scores out of ten. An option that fails a must is out, however attractive it is otherwise. This often settles the matter. If Stay fails "my work should interest me" and you've known that for two years, the contest is between the other two.

Ask for information only with a question attached. Use chapter 4's test. If a fact would change the outcome and someone can supply it, go and get it, and let the date allow for it. Otherwise carry on.

Set the date. Pick it from the time allowance in the last chapter. Write it down and tell someone. On that date you choose with whatever you know.

Pick the first option that passes. If only one option meets all the musts, you've finished. If two do, they're both good enough by your own standard, and chapter 6's argument applies. The difference between them is small compared with the cost of another year. Take the one you'd be more pleased to tell a friend about, or the one that's easier to undo, or use the next step.

If it's level, use a coin. You won't necessarily do what the coin says. The toss is there to show you something. Assign the options, toss it, and notice how you feel in the second after it lands. People often find they're disappointed or relieved, and that tells them what they wanted. If you feel nothing either way, the options really are level for you, and you may as well do what the coin says.

There's some evidence for this. The economist Steven Levitt ran a large experiment in which volunteers who were stuck on a decision let a coin toss advise them. Some of the decisions were major, such as quitting a job or ending a relationship. People whose coin told them to make the change were more likely to make it. Six months later, those told to change were more satisfied with their decision and reported being happier than those told to stay as they were. Levitt's conclusion was that people may be too cautious when they face choices that would change their lives.

The volunteers chose to take part and rated their own happiness, so it isn't proof. It does fit the rest of this book, though. People who'd been stuck and then made a change were, on average, glad afterwards.

Sleep on it once. For a large decision, make the choice on the date and do nothing about it until the next morning. A night's gap is a fair check against a choice made in a rush of feeling. If it still stands at breakfast, it goes ahead. Give it one night only, because a second night of thinking it over would be another lap. And the morning test is whether a new fact has turned up. You're bound to feel some doubt, and that doesn't count against the choice.

Act within a day. A choice that stays in your head isn't finished, as the eleven o'clock feeling showed. Within twenty-four hours of that morning, do something that makes it real. Send the application, pay the deposit, or book the meeting with your manager. For the course, that means sending the form.