Spaced Repetition

Also known as SRS, spaced repetition system, Anki method, Leitner system

Review each item just as you're about to forget it, so a fixed amount of study buys far more long-term retention.

Spaced repetition is the memory technique behind Anki and every app that schedules your reviews for you. It rests on one of psychology's oldest and most replicated findings: Hermann Ebbinghaus, testing himself on nonsense syllables in the 1880s, showed that memory decays sharply and predictably after learning, and that re-studying an item resets and flattens that curve. Massed study — twenty repetitions today — is far less durable than the same twenty spread out over months. This 'spacing effect' is one of the more robust results in learning research.

The software turns that into scheduling. Each time you review a card you rate how easily you recalled it, and the algorithm pushes the next review out to the point where you are on the verge of forgetting — a day, then three, then a week, then a month, then a year. Because intervals grow, a mature deck of thousands of cards can be maintained in twenty minutes a day. Piotr Woźniak built the first such program, SuperMemo, in 1987 and published his SM-2 algorithm openly in 1990; Damien Elmes's Anki, released free and open source, implemented a variant of SM-2 and became the default tool for language learners, with newer schedulers such as FSRS now fitting intervals to your own review history.

For languages, the technique arrived through two doors. Japanese learners adopted it early for kanji and vocabulary, where the sheer item count makes any unscheduled review hopeless. Gabriel Wyner's Fluent Forever then popularised a particular way of using it for languages generally: pronunciation first, pictures and personal associations instead of translations, and cards built by you rather than downloaded.

A week of it is undramatic. You open the app, you clear the queue, you add a small number of new cards, you close it. The discipline is not intensity but continuity — spaced repetition punishes gaps harshly, because a fortnight away turns a maintained deck into a backlog of hundreds of overdue reviews.

The important limit is what a card can hold. Spaced repetition is extremely good at making a specific cue reliably produce a specific response. It is not a way of acquiring a language: it fixes items that exposure has already made meaningful, and it works best as a supplement to reading and listening rather than as the main event.

How it works

Pick one system and let it schedule you — the whole benefit is in not choosing review dates yourself.
Make cards small: one fact per card, one clear cue on the front, one answer on the back.
Prefer cards you make from material you have met over large downloaded decks, which are mostly unfamiliar and mostly forgotten.
Limit new cards per day; every card added is a review obligation for years.
Review every day, briefly, rather than in long sessions — the schedule assumes daily contact.
Delete ruthlessly. A card you keep failing is usually a bad card, not a weak memory.

Origin Hermann Ebbinghaus, then Sebastian Leitner and Piotr Woźniak, 1885 → 1970s → 1987 — Ebbinghaus documented the forgetting curve and the spacing effect; Leitner turned it into a paper box-and-card system; Woźniak's SuperMemo made it algorithmic, and Anki made it free.

Good for

Large item counts with arbitrary form-meaning links — vocabulary, kanji, genders, irregular forms
Maintaining a language you can't use daily
Learners who can sustain a small daily habit

Trade-offs

It teaches what is on the card and nothing else. Knowing a word's gloss under a familiar cue is a long way from recognising it at speed in speech, and decks routinely overstate how much you 'know'.
The maintenance is permanent and unforgiving. Miss two weeks and you return to a wall of overdue reviews; many learners quit at exactly that point rather than resetting the deck.
Adding is cheap and reviewing is not. Deck bloat — thousands of cards nobody chose to maintain — is the standard way SRS study collapses.
It is easy to mistake for the whole job. Time in Anki is time not spent reading or listening, and cards only consolidate what input has already introduced.
The spacing effect is well evidenced; the precise scheduling algorithms are far less so, and arguments about SM-2 versus FSRS versus custom settings absorb energy out of all proportion to their effect.

Who advocates this

Piotr Woźniak Creator

Built SuperMemo from 1987 and published the SM-2 spacing algorithm that later systems adapted.
Damien Elmes Creator

Wrote Anki, the free open-source SRS, originally to support his own Japanese study.
Gabriel Wyner Creator

Wrote Fluent Forever, which popularised SRS for language learners with self-made, image-based cards.

Learners who used this

Field notes whose stack included this approach. They are an account of what one person did, not evidence that it works — and note how few run a single method on its own.

A 341-day streak, six tools, and five hours of silence at a family dinner · 12 months · blended with 1 other approach
Three years stuck on pure input, then a textbook, a teacher and an exam date · 60 months · blended with 2 other approaches
A 1,480-day Anki streak ended, and the relief was the lesson · 48 months · blended with 2 other approaches
Twenty-seven apps over eight years, reviewed one by one · 96 months · blended with 1 other approach
Ten years of classes, textbooks and apps — and still barely past beginner · 120 months · blended with 1 other approach

All field notes →

Related methods

Resources that use this

Anki

Multi-language

The gold-standard spaced-repetition flashcard app with community decks for every language.