Spaced repetition, explained properly

Most studying is a fight against a curve measured in 1885. Hermann Ebbinghaus taught himself nonsense syllables, tested his own recall at intervals, and drew the result: memory falls off a cliff. Within a day, roughly two-thirds of new material is gone; within a week, most of the rest follows. Every technique in this article is a way of bending that curve — and the two that bend it hardest, testing and spacing, are exactly what flashcards combine.

The forgetting curve

The curve's shape matters more than its exact numbers: loss is steepest immediately after learning and flattens with time. Whatever survives the first day has a much better chance of surviving the week. That has a practical inversion — the highest-value moment to review something is just before you'd otherwise lose it, and early reviews are worth far more than late ones.

Each successful review doesn't just restore the memory; it flattens the next stretch of curve. Recall something one day after learning it and it might last a week; recall it again at the week and it can last a month. Forgetting slows down every time you catch it — which is the entire mechanism spaced repetition exploits.

The testing effect

The second finding is less intuitive: retrieving a memory strengthens it far more than re-encountering it. In study after study, students who read a passage once and then tested themselves remembered more a week later than students who re-read it four times. Re-reading feels productive because the material becomes familiar; but familiarity is recognition, and exams, conversations and life demand recall. The struggle of dragging an answer out of memory — even a failed attempt, followed by checking — is the event that builds the trace.

This is why a flashcard beats a highlighter: the card's front forces an attempt before the back gives the answer. Highlighting files information under “seen”; retrieval files it under “mine”.

Why spacing multiplies it

Ten retrievals in one evening and ten spread over a month are not the same ten. Massed practice lets each retrieval happen while the memory is still warm — easy, and therefore weak. Spaced practice forces each retrieval to happen after partial forgetting, when it is effortful — and effortful retrieval is what resets the forgetting curve to a flatter slope. The counterintuitive core: a bit of forgetting between reviews is not the enemy of learning, it is the ingredient.

What intervals to use

The classic expanding schedule, good enough for almost everyone:

• First review: the same day, a few hours after learning.
• Then: 1 day later → 3 days → 1 week → 2 weeks → 1 month → 2 months.
• On a miss: the card drops back to short intervals and climbs again.

The exact numbers matter far less than the shape: expanding gaps, with misses recycled sooner. Our flashcard tool applies the within-session half of this automatically — any card you miss returns to the end of the run until you get it — and the between-session half is you reopening the deck tomorrow, then later in the week. The deck is saved on your device, so the reopening is the only effort left.

Writing cards that work

• One fact per card. “The three causes of X” is three cards. Compound cards get half-known and stay that way.
• Both directions. Term→definition and definition→term are separate memories. For vocabulary especially, make both.
• Write your own. Phrasing the question forces you to understand the answer — a borrowed deck skips the most instructive step.
• Make the front unambiguous. If two answers fit, you'll practise guessing which one the card wants instead of knowing the material.
• Shuffle. A fixed order teaches sequence cues that will not exist in the exam. Shuffle before every run.

The mistakes that waste it

Three habits quietly cancel the method. Flipping the card before honestly attempting the answer turns retrieval back into re-reading. Marking a card “known” when the answer took ten seconds and three hints defers the forgetting to exam day. And cramming the whole deck the night before — the classic — produces real short-term recall and near-total loss a week later, which is fine for one exam and useless for a language, a syllabus or a profession.

The method in one sentence: test yourself, space it out, be honest about misses — and let short, regular sessions carry it rather than heroic ones.