Your deadline is lying to you.

You already know the real date. You typed it in. So why would an earlier one you invented on purpose change anything?

Because the problem was never that you forgot when it was due. The problem is that you planned as though nothing would go wrong. Kahneman and Tversky named this the planning fallacy in 1979: people systematically underestimate how long their own tasks will take, even when they have done the same task before and know it ran long. Buehler, Griffin and Ross put numbers on it in 1994 by asking students to predict when they would finish their theses — most missed their own predictions, and a striking share finished later than the worst case they had been asked to imagine. The estimate is not a lie. It is a best case, and a best case with no buffer in it is a plan that breaks the first time anything is slightly harder than expected.

A fake deadline does one specific thing about that: it moves the moment of urgency forward to a point where urgency is still useful. Panic three days out produces a finished draft. Panic three hours out produces whatever you can type in three hours.

Yes — but less well than deadlines somebody else sets, and that gap is the entire reason this tool exists.

The clearest evidence comes from Ariely and Wertenbroch, writing in Psychological Science in 2002. They gave students the option to set their own deadlines for a series of papers, with real penalties for missing them. Two things happened. First, most students voluntarily imposed costly deadlines on themselves — people know perfectly well that they procrastinate, and will pay to be protected from it. Second, those self-imposed deadlines helped: students who set them outperformed students who only faced the end-of-term date.

But the students who did best were the ones given evenly spaced deadlines by the instructor. Self-imposed structure beat no structure. Imposed structure beat both.

Which leaves an obvious question: if external deadlines work better, and nobody external is going to set you one, what do you do? This tool is one answer. It takes a date you chose and hands it back to you wearing somebody else's clothes.

A date you set yourself is a date you can move yourself, and you know it. That is why "I'll aim to finish early" almost never survives contact with a bad Tuesday — there is nothing behind it except your own intention, and your own intention is exactly what is under negotiation.

So this tool attaches a reason to the date. Some of the reasons are plausible and bureaucratic; some are openly ridiculous. That mix is deliberate. You are not trying to genuinely deceive yourself — that is not possible and it is not the point. You are giving the earlier date a story, because a date with a story attached is harder to quietly delete from a calendar than a bare one.

There is a related finding worth knowing here. Gollwitzer's work on implementation intentions, summarised in American Psychologist in 1999, found that people who specify in advance when and where they will act on a goal follow through substantially more often than people who merely hold the goal. The specificity does the work. A deadline sitting in a calendar with a stated reason is a far more specific object than a vague intention to start early, which is why this tool offers to put it in your calendar rather than just showing you a date and wishing you luck.

Not a flat percentage, which is the mistake this tool made at first. Both extremes fail, but they fail at different ends of the scale:

• Too small a gap and it is not a deadline, it is a rounding error. Eighteen percent of a week is one day, and nobody has ever changed their behaviour because something moved to Thursday.
• Too large a gap and you stop believing it. If a task due in a year is "due" in three months, your brain correctly files that as fiction and ignores it — along with the next one.

So the share cut grows as the time shrinks. A month-long project loses about a week. A two-week job comes due in nine days or so. A week becomes three days, and three days becomes tomorrow. On long horizons it settles to roughly a quarter, because a quarter of a year is already months.

There is one hard floor: the issued date is never today. However little time you have, you are left at least one whole day to work in.

It will not help on a task that is genuinely too large for the time available — a buffer cannot manufacture hours that were never there. It also does not work if you keep re-rolling until you get a date you already liked, which is just the original deadline wearing a costume.

And it works less well the more people are involved. A deadline shared with a team has its own gravity; the useful move there is to agree a real internal date with the other humans rather than a private fiction with yourself.

One more limit, and it is the one people run into: this is a scheduling trick, not a treatment. If you are not starting things because of burnout, depression, or ADHD that is going unmanaged, an earlier date is a heavier stick to beat yourself with rather than a solution. The technique assumes the work is doable and the problem is when you begin. When that assumption is false, the tool is the wrong tool.

Related, but it is worth being precise, because the two get conflated constantly. "Work expands so as to fill the time available for its completion" comes from a satirical essay C. Northcote Parkinson published in The Economist in 1955. It is a sharp observation and an adage — not a research finding, and it is usually quoted as though it were one.

The adage suggests shortening the time available. The planning-fallacy research suggests something subtly different: that your estimate of the time required was too optimistic to begin with. This tool sits with the second reading. It is not trying to compress the work into less time. It is trying to put the buffer back that your estimate quietly left out.

Entirely in your browser, from the date you type. It takes the days remaining until your real deadline and cuts a share of them. The share is 18% + 2.6/days, capped at 70% — which is a long way of saying "about a quarter for long projects, well over half for short ones" — plus a little random jitter so repeat visits do not produce an obviously mechanical number. The result is then clamped so it can never be today and never be the real date.

The excuse is drawn at random from 61 hand-written options, and re-rolling changes only the excuse, never the date.

Nowhere. Everything on this page is computed in your browser. What you type never reaches a server, there is no account, and this site has no database of deadlines because it has no database.

Your saved deadlines are kept in this browser's own local storage, on this device. That means two things worth being clear about. They are private — we cannot see them, and nobody else can either. And they are fragile: clear your browser data and they are gone, and they will not appear in a different browser, on a different device, or in a private window. There is no copy anywhere for us to restore.

The calendar options hand the event to Google, Outlook or your own calendar app directly, from your browser. This site is not in the middle of that and never sees it.

• Kahneman, D., & Tversky, A. (1979). Intuitive prediction: Biases and corrective procedures. TIMS Studies in Management Science, 12, 313–327. — the paper that named the planning fallacy.
• Buehler, R., Griffin, D., & Ross, M. (1994). Exploring the "planning fallacy": Why people underestimate their task completion times. Journal of Personality and Social Psychology, 67(3), 366–381.
• Ariely, D., & Wertenbroch, K. (2002). Procrastination, deadlines, and performance: Self-control by precommitment. Psychological Science, 13(3), 219–224.
• Gollwitzer, P. M. (1999). Implementation intentions: Strong effects of simple plans. American Psychologist, 54(7), 493–503.
• Parkinson, C. N. (1955). Parkinson's Law. The Economist. — an essay, not a study.