Spaced Repetition and the Forgetting Curve

Memory does not fade at a constant rate. Hermann Ebbinghaus, a German psychologist, documented this in the 1880s through self-experiments: within 20 minutes of learning new material, a significant portion is already forgotten. Within a week, most of it is gone unless something reinforces it.

The pattern follows an exponential decay curve — fast drop-off immediately after learning, then a slower decline.

Time since learning	Approximate retention (no review)
20 minutes	~58%
1 hour	~44%
1 day	~33%
1 week	~25%
1 month	~21%

These are approximate values based on Ebbinghaus's original research on arbitrary syllable memorisation. Retention for meaningful, connected material decays more slowly — but the curve still applies.

The core finding: without review, most new material is forgotten within days, regardless of how attentive you were during the lesson.

A single review does not defeat the forgetting curve — it resets it. Each time you successfully retrieve a piece of information, the decay curve starts again from a higher baseline. The material takes longer to forget after each successful review.

Spaced repetition is the practice of reviewing material at increasing intervals timed to catch it just before it would be forgotten. The intervals grow longer after each successful review because each retrieval strengthens the memory trace.

How the retention curve changes with spaced reviews:

Review	Approximate interval	Effect
1st review	1 day after learning	Retention resets; decay now slower
2nd review	3 days after 1st	Retention resets again; decay slower still
3rd review	7 days after 2nd	Material entering long-term memory
4th review	14 days after 3rd	Decay very slow; minimal effort to recall
5th review	30 days after 4th	Material effectively consolidated

Without spacing, a student re-reading the same notes for 3 hours straight gets far less retention than a student who spends 20 minutes on day 1, 10 minutes on day 2, and 10 minutes on day 4.

A concrete example: you learn a set of 20 flashcards on binary numbers on Monday.

Day	Action
Monday	Learn the 20 cards for the first time
Tuesday	Review all 20 cards (first spaced review)
Friday	Review cards you got wrong on Tuesday + spot-check the rest
Following Friday	Full review of all 20 cards
Two weeks later	Quick review — most cards retrieved easily

The total review time over five weeks is roughly 60–80 minutes. A student who instead re-reads the same notes for 60 minutes on Monday will retain far less two weeks later.

The key principle: time your reviews to coincide with the point when you're about to forget. Reviewing material you already know perfectly wastes time; reviewing material you've partially forgotten produces the strongest memory trace.

Spaced repetition only works if each review involves active retrieval — attempting to recall the answer before seeing it. Re-reading notes or highlighting text does not trigger the same memory reinforcement.

The testing effect (also called retrieval practice) is one of the most replicated findings in learning research: the act of attempting to retrieve information strengthens the memory far more than re-exposure to the information passively.

What counts as active retrieval:

• Turning over a flashcard and answering before seeing the back
• Closing your notes and writing down everything you remember about a topic
• Attempting a practice question before checking the mark scheme

What does not count:

• Re-reading a chapter
• Watching a video again
• Reading through your own notes without covering them

If you look at the answer before attempting to recall it, the review does not count. The struggle to retrieve is what produces the learning benefit.

The Leitner system is a physical implementation of spaced repetition using a set of boxes (or dividers) and a card deck.

Setup: Label five boxes or sections as Box 1, Box 2, Box 3, Box 4, Box 5. Start with all new cards in Box 1.

Rules:

Box	Review frequency	What happens on correct answer	What happens on wrong answer
1	Every day	Move to Box 2	Stay in Box 1
2	Every 2 days	Move to Box 3	Back to Box 1
3	Every 4 days	Move to Box 4	Back to Box 1
4	Every 8 days	Move to Box 5	Back to Box 1
5	Every 16 days	Retired (mastered)	Back to Box 1

Cards you know well move through the boxes quickly and are reviewed less often. Cards you struggle with return to Box 1 and get daily attention.

The digital equivalent is Anki, which automates this process: it tracks your performance on each card and schedules the next review at an optimal interval. Historically Anki used a modified SM-2-style scheduler; newer versions also support FSRS, a more recent spaced repetition algorithm.

Cramming — reviewing a large amount of material in a single extended session immediately before an exam — produces short-term retention at the cost of long-term recall.

Property	Cramming	Spaced repetition
Retention after 1 day	High	High
Retention after 1 week	Low	High
Retention after 1 month	Very low	Moderate to high
Stress level	High (time pressure)	Low (built into routine)
Suitable for	One-off recall	Exams, long-term mastery

Cramming works for a test the next morning. It does not build the durable memory that A-level or GCSE exams require, where material from September may be tested the following June.

Cramming fills working memory temporarily. Spaced repetition encodes material into long-term memory. Only one of these survives sleep.

1. Reviewing too soon

If you review a card hours after learning it rather than the next day, the memory trace is still strong and the review produces little benefit. The spacing — the gap — is the mechanism. Reviewing too early wastes sessions.

2. Passive review

Looking at a card and thinking "yes, I know that" without actively generating the answer is not a retrieval attempt. Cover the answer and speak or write it before flipping the card.

3. Too much on each card

A card that asks "explain everything about binary addition" requires many separate memory traces. Split complex topics into one fact or one concept per card. A card should require a single, specific retrieval.

4. Starting too late

Spaced repetition requires time — the intervals need to play out. Starting two days before an exam gives you one review cycle, not five. Build the system at the start of the term, not the end.

5. Abandoning the system after a difficult session

On days when you get many cards wrong, the temptation is to feel the system isn't working. Those hard sessions are the most productive ones — the retrieval difficulty is exactly what drives memory consolidation.