Forget a Little, On Purpose
Spacing, sleep, and the ten-second pause that lets your brain replay
Cramming works. That is the annoying part. Read something ten times tonight and you will know it tomorrow morning. The trouble is what happens by next week, and next week is when the exam, the client, or the real decision usually turns up. More than a century of memory research agrees on one fix: spread the same work out, and let yourself forget a little in between.
The spacing effect
Hermann Ebbinghaus documented it in 1885 by memorizing lists of nonsense syllables and plotting how fast they faded. The modern version is bigger. In 2008 Nicholas Cepeda, Harold Pashler and colleagues taught facts to more than 1,350 people, brought them back after gaps of up to three and a half months for one review, and tested them up to a year later. For every test date there was a best gap, and the best gap grew with the distance to the test: a day or two for a test next week, a few weeks for a test a year away. Review too soon and the second pass is wasted, because nothing has been forgotten yet. Review too late and you are relearning from scratch. The sweet spot is when recall takes effort but is still possible.
Robert Bjork calls these desirable difficulties. A memory you have to work to retrieve gets rebuilt stronger than one that was simply sitting there. The struggle is the mechanism. It is also why the second read tonight feels so productive and does so little: there is no struggle in it.
How the review deck uses this
Every lesson you pass adds a handful of flashcards to your review deck, and each card sits in one of five boxes. A card you get right moves up a box and comes back later: tomorrow, then three days, then a week, then two weeks, and every two weeks after that. A card you miss drops back to the first box, comes round again before the session ends, and is due again today. You never schedule anything. Open the deck when the home screen says cards are due, and the cards you see are the ones about to slip. The streak counts days, not hours, for the same reason: three short sessions across three days beat one long one, and the app is built to reward the pattern that lasts.
Sleep is when it gets written down
Learning carries on after you stop studying. The connections you made during the day are replayed and stabilised during sleep, and the night right after learning does most of that work. An all-nighter before an exam trades the consolidation of everything you learned that day for a few extra hours of exposure, which is a bad trade. Huberman's practical advice is simple: protect your sleep on the night you learn something, and if you are short on it, a ten to twenty minute non-sleep deep rest or a short nap after a study session helps the material settle.
The ten-second pause
Here is a finding that sounds too small to matter. In 2021 a team at the National Institutes of Health, writing in Cell Reports (Buch, Claudino, Quentin, Bönstrup and Cohen), had people learn a typing sequence in ten-second bursts with ten-second rests between them. Almost all of the improvement happened during the rests, while people were doing nothing, and brain recordings showed the hippocampus replaying the just-practiced sequence during those pauses, compressed about twenty times faster than it had been performed. Your brain rehearses for you, for free, if you give it a gap. Huberman's version of the protocol: while learning anything, pause every so often for ten to thirty seconds and do nothing. Do not reach for your phone. Let the replay run.
Focus is trainable, and it is the bottleneck
None of this works if attention is elsewhere, which is why the top medical students in the last lesson went to such lengths to remove interruptions. Two things help. First, set the room up before you start: phone in another room (or, if you are studying on your phone, notifications off), one tab open, and a blank sheet beside you to write down any distraction the moment it arrives, so it stops circling and you can see what to design out next time. Second, attention itself responds to practice. Huberman points to studies in which five to ten minutes a day of simple mindfulness practice, or the exercise of holding your gaze on a single point and bringing it back every time it drifts, improved focus and recall.
Study in short sessions across days. Sleep on it. Pause. Let the app decide when you see a card again. Your job is to show up and try to remember; the scheduling is done for you.
Next: the part nobody teaches, which is what to do with your attention while you read, and how to test yourself the way the test will.