Field notes
Sleep, researched properly
Everything here is written in the first person by one named human, checked against primary sources, and linked to those sources at the foot of the page. Where the evidence is thin, it says so, including when that is inconvenient for the calculator on the front page.
Start here
The three pieces the calculator is built on. Read them in this order if you are reading all three.
What a sleep cycle actually is, and why 90 minutes is a lie you can use
Four stages, wildly unequal shares of the night, deep sleep front-loaded and REM stacked at the end. Where the 90-minute figure came from, how wide the real spread is (70–120 minutes, with one large dataset putting the median at 110), and why I kept 90 in the calculator anyway.
Published 15 Aug 2026
How much sleep do I actually need, by age
The AASM and NSF numbers side by side, why they are ranges, where “eight hours” came from (a labour slogan, not a lab), the rare people who genuinely need less, and a two-week method for finding your own number.
Published 15 Aug 2026
Are sleep calculators accurate?
An audit of my own tool. What I will defend, what is educated guesswork, the claim the whole category makes that nobody has demonstrated, and the four inputs that matter which no calculator can see.
Published 15 Aug 2026
Sleeping better, and what actually helped
A first-hand record of five changes, each traced to whatever evidence exists behind it. Some of it is very strong. Some of it is almost nothing, and I say which is which.
Five things I changed to sleep better, and how good the evidence is for each
Fixed wake time, phone out of the bedroom, morning sunlight, exercise, seeing people. Ranked by how good the evidence actually is, which turns out to be wildly uneven. Four of the five were a 1972 clinical protocol I had never heard of.
Published 31 Aug 2026
Sleep regularity beats sleep duration, on the biggest study we have
Across 88,975 people wearing accelerometers, when you sleep predicted mortality more strongly than how long. The most regular fifth had around 30% lower risk. And the caveats it deserves, applied consistently.
Published 31 Aug 2026
The 20-minute rule: get out of bed, and where the number comes from
Not folk advice. It is stimulus control therapy, formalised in 1972, and the only nonpharmacological insomnia treatment with the highest AASM rating. The threshold across studies ranged from 10 to 25 minutes, and you should not be timing it.
Published 31 Aug 2026
Morning sunlight and sleep: a week of camping moved people 1.4 hours
Eight people camped for a week with no electric light and their melatonin onset shifted 1.4 hours earlier, in every single participant, night owls included. Plus the evening half of the light problem that nobody mentions.
Published 31 Aug 2026
Loneliness does not shorten your sleep, it breaks it up
Lonely people had poorer sleep efficiency and more time awake after falling asleep, while total duration was unchanged. The research I said did not exist, filed under a word I had not searched for.
Published 31 Aug 2026
Getting the phone out of the bedroom, and why blue light is the least of it
The study everyone cites used four hours of screen reading, not a two-minute check. The stronger argument is conditioning. Plus my own 28 nights: 16 minutes to fall asleep with a screen, 9 without.
Published 31 Aug 2026
Where the numbers come from
Every number people repeat about sleep, traced back to the research it actually came from. Mostly American sources, because that is where most of this guidance originates.
Where sleep numbers come from
Eight claims traced to their primary source: the 90-minute cycle, the eight-hour rule, the twenty-minute nap, the ten-hour caffeine cutoff and more. Three of the eight do not survive the trace intact. Updated as claims get checked.
Published 31 Aug 2026
Does napping wreck your sleep at night? It depends on two things
Short and early is generally fine. Long and late is the combination that costs you, because a nap spends the sleep pressure you need at bedtime. Plus why this matters much more if you already sleep badly.
Published 31 Aug 2026
How to actually fall asleep for a nap, with my own numbers
My median across 28 nights was 12 minutes, with one night in ten over 31. If you are awake at minute fifteen you are in an ordinary range. What moved my number, and the release valve nobody leads with.
Published 31 Aug 2026
Are long naps bad for you? What those studies can and cannot show
Long naps are associated with worse health outcomes. But ill people nap more, and the researchers say so themselves. The association is real; the causal claim is not established.
Published 31 Aug 2026
The coffee nap: the one nap trick with a real trial behind it
Drink coffee, then nap 15 to 20 minutes. In the trial that tested all three conditions, the combination cut driving incidents to about 9% of placebo against 34% for caffeine alone. Plus the bedtime cost nobody mentions.
Published 31 Aug 2026
The best time of day to nap, and why it has nothing to do with lunch
Early afternoon, and not because of your sandwich. The dip shows up in people who have eaten nothing and do not know what time it is, which makes it the body clock rather than digestion.
Published 31 Aug 2026
Why you feel worse after a nap, and why almost nothing fixes it
Sleep inertia is measurable impairment, not a mood. One study found a 41% performance drop on waking from deep sleep, and a 2019 review found no reactive countermeasure with clear evidence behind it. A problem to prevent rather than treat.
Published 31 Aug 2026
Where the 20-minute nap rule actually came from
The only study that put 5, 10, 20 and 30-minute naps side by side found ten minutes best, and that twenty did nothing measurable for the first thirty-five minutes after waking. Plus what the NASA study everyone cites actually measured.
Published 31 Aug 2026
The 10-3-2-1-0 sleep rule, traced back to whoever made it up
It is described everywhere as expert-backed. It came from a fitness and productivity author in a 2016 HuffPost article, and has never been tested as a protocol. Each of the five numbers checked against the study it is credited to.
Published 31 Aug 2026
How this started
Less a reference page than an explanation of why the rest of them exist.
What you will not find here
Supplement affiliate links, mattress rankings, or a claim that a free web tool can fix your sleep. If something on this site turns out to be wrong, it goes in the corrections log with the date and what changed, rather than being quietly edited. The method behind every article is set out in the editorial standards, and what this site is for is set out in full on the about page.
The tool these notes are about
Two modes, adjustable cycle length, adjustable time to fall asleep, and every assumption printed on the page rather than buried.
Open the sleep calculator