Odysseus knew exactly what he was. He knew that when the Sirens sang, he would want nothing in the world as much as to steer toward them, and that he would be completely sincere about it. So he did not rely on character. He had his crew plug their ears with wax and tie him to the mast, with instructions to ignore every order he gave until the island was behind them.

The part that gets lost in the retelling is that he was not weak. He was the cleverest man in the story. His cleverness was in admitting that the person who would be standing on that deck later was not the person making the plan, and that the later person could not be trusted with the wheel.

Most of us have heard the Sirens in some form this week. We have also made the plan: phone in the other room, no email until noon, finish the draft tonight. And we have watched the person we become at 2:40 in the afternoon quietly dismantle it.

Why good intentions keep losing

Behavioral economists have a name for the problem: time-inconsistent preferences. Ask yourself on Sunday night what you want from Monday, and you will choose the long-term option almost every time. Ask at 10:15 on Monday morning, with a hard task open and a notification glowing, and the short-term option wins. Same person, different answer, because the pull of an immediate reward grows disproportionately as it gets close. This is usually called present bias, and it is why the plan feels sincere when you make it and flimsy when you need it.

The interesting question is what you do once you know this about yourself. Economists Ted O'Donoghue and Matthew Rabin drew a useful distinction between two kinds of people. The naive one assumes their future self will simply follow the plan. The sophisticated one expects their future self to be tempted and plans around it. The naive person makes resolutions. The sophisticated person makes arrangements.

A resolution asks your future self to be better than your present self. An arrangement changes the situation your future self walks into. Odysseus made an arrangement.

What a commitment device actually is

A commitment device is any choice you make now that makes it harder, costlier, or impossible to do the tempting thing later. It does not depend on willpower in the moment, because the decision was already made, at a time when you were calm and thinking clearly.

They come in a few flavors:

  • Hard commitments remove the option entirely. A time-locked container for your phone. A website blocker you cannot disable until a set hour.
  • Soft commitments leave the option open but raise its price. Money on the line, a public promise, a friend who will ask how it went.
  • Friction commitments do not forbid anything. They add steps: logging out, deleting an app from your home screen, leaving the charger in another room.

None of these is exotic. The evidence for the underlying idea is a little more nuanced than the productivity internet usually admits, which makes it more interesting.

What the research shows (and what it doesn't)

In a well-known 2002 experiment, Dan Ariely and Klaus Wertenbroch gave students in a class three papers to write and let some of them choose their own deadlines. Left free, students could have set all three for the last day, which is the optimal choice if you are perfectly rational and your only goal is flexibility. Many instead chose earlier, spaced-out deadlines that carried real penalties for lateness. They were binding themselves. And those who did so performed better than students who gave themselves no constraints.

The detail worth remembering is the second half of that finding: self-chosen deadlines helped, but they did not help as much as deadlines imposed from outside, which on average led to better performance still. People bind themselves, but not quite enough. We tend to set the restraints a little looser than we need them, because the person drawing up the contract is still slightly optimistic about the person who will live under it.

Outside the lab, a field study by Xavier Giné, Dean Karlan and Jonathan Zinman offered smokers in the Philippines a savings account in which their own deposits would be forfeited if they failed a nicotine test months later. Only a minority of those offered the product took it up. But those given the offer were modestly more likely to pass the later test than those who were not, and the effect persisted at a follow-up. The takeaway is balanced: commitment devices can work, many people decline them, and the ones who decline are not always the ones who need them least.

That last point deserves a pause. Part of what keeps us from binding ourselves is the same overconfidence that makes binding necessary. "I won't need it" is the naive self talking.

The design rules that make them work

A commitment that is too strict gets dismantled, often at the worst moment, and teaches you that your own rules are negotiable. A commitment that is too soft does nothing. A few principles help.

Bind the moment, not the whole life. Odysseus was tied up for the length of the Sirens' range, not for the rest of the voyage. A commitment with a clear, short window (this 90-minute block, this evening) is easier to accept and easier to keep than a permanent rule.

Make it at a cool moment. Set the arrangement when you are calm and the temptation is not in front of you. If you are drafting your rules while the urge is in the room, you are negotiating with the person you are trying to restrain.

Make the cheat expensive, not impossible. Total lockdowns invite a dramatic override. A modest speed bump, such as a delay, a typed phrase, or a walk to another room, filters out the impulsive version of you and still lets the genuinely urgent version through.

Pick the point of failure. Look at where you actually break. For most people it is not the decision to start working but the first moment of discomfort ten minutes in. Put the device there.

Tell a witness. Reputation is a cheap and surprisingly strong stake. Saying "I'm offline until 1" to one specific person changes what ignoring it would mean.

Your next moves

These take under half an hour in total, and you can do them today.

  • Find your Siren. Write down the single thing that most often pulls you off your most important work. Not a category, the specific app, site, or habit.
  • Pick one 60–90 minute window tomorrow and write down its start and end time. Put it in your calendar like a meeting.
  • Add friction at the exact point of failure. Log out of the app, move the icon off your home screen, or put your phone in another room before the window starts. Do it tonight, while you are calm.
  • Attach a small stake. Tell one person what you will finish and when, or promise a specific amount to a cause you dislike if you do not. Choose something that stings a little, not a lot.
  • Schedule a five-minute review. After the window, note whether you tried to override the arrangement and when. If you did, tighten it next time. If you did not, you set it right.

Binding yourself is not a character flaw

It is tempting to hear all this as an admission that we are not in control. The old story says the opposite. Odysseus is the one who heard the song, the only one, and lived to describe it. He did not win by being immune. He won by arranging, ahead of time, for his judgment to outlast his desire.

There is something quietly kind in that. When you set up the arrangement, you are caring for the person who will be tired, pulled, and tempted later, and you are doing it from the one position where you can still think straight.

That is the idea behind Reclaim: a way to make the arrangement once, while you are clear-headed, so that your future self meets a little friction instead of a fresh argument. If you want to try it, you can find it at https://reclaim.lumenlabs.works. And if you would rather not, tie yourself to the mast some other way. The Sirens are not going anywhere.