Spaced Repetition Explained: The Science of Long-Term Memory

Hermann Ebbinghaus's forgetting curve is one of psychology's most replicated findings. He memorized nonsense syllables and tested himself at intervals, discovering that memory decays exponentially without review: roughly 56% forgotten in one hour, 66% in one day, and 75% in six days.

How Reviews Strengthen Memory

When you review material before completely forgetting it, the next forgetting curve starts higher and decays slower. Each subsequent review extends the time before the next forgetting event. This is the spacing effect: distributing study time across multiple sessions produces dramatically better retention than concentrating it into one session.

Over 300 studies across different age groups and subjects confirm the spacing effect, making it one of the most robust findings in cognitive science.

Key Concepts

• Ebbinghaus Forgetting Curve: Memory decays exponentially without review. Each review session flattens the curve and slows future decay. Published in 1885.

• Spacing Effect: Distributed study produces significantly better long-term retention than massed practice (cramming). Effect size is typically 0.5-1.0 standard deviations, a large practical impact.

• Optimal Interval: The ideal review time is just before you would forget. Not too soon (wasted effort), not too late (starting from scratch). Algorithms calculate this for each card based on your history.

• Memory Consolidation: The neurological process converting short-term memories to long-term memories, primarily during sleep. Spaced repetition leverages consolidation by allowing time between reviews for neural connections to strengthen.

The history of spaced repetition mirrors computing evolution. Sebastian Leitner's cardboard box system (1972) was the first practical implementation: flashcards sorted into compartments reviewed at different frequencies.

Correct answers advance cards to less-frequent review. Incorrect answers send cards back to daily review.

Computer-Based Systems

Piotr Wozniak created SM-2 in 1987, the first computerized implementation. It calculates optimal intervals using a mathematical formula based on your performance history. SM-2 powered SuperMemo and Anki for over 30 years.

In 2022, Jarrett Ye published FSRS (Free Spaced Repetition Scheduler), which uses a three-parameter memory model (Difficulty, Stability, Retrievability) to calculate intervals with 20-30% greater accuracy than SM-2. Anki adopted FSRS in version 23.10, and FluentFlash uses it as the default algorithm.

Algorithm Comparison

• Leitner System (1972): Physical flashcard box with 3-5 compartments. Simple and effective analog spaced repetition without technology.

• SM-2 Algorithm (1987): Uses an 'ease factor' per card that adjusts based on performance. Simple but tends to over-schedule easy cards and under-schedule difficult ones.

• FSRS Algorithm (2022): Modern algorithm using a DSR memory model. Tracks how fast you forget each card and adjusts individually. 20-30% more accurate than SM-2 in peer-reviewed studies.

• Desired Retention Rate: Your target probability of remembering a card when reviewed. Default is typically 90% (remember 9 of 10 cards). Higher retention (95%) means shorter intervals. Lower retention (85%) means longer intervals.

Using a spaced repetition app is straightforward: you create flashcards, then review them when the app indicates they are due. Each day, you open the app and see cards at the optimal point for review, just before you would forget them.

You look at the question, attempt to recall the answer, then rate your performance on a 4-point scale: Again, Hard, Good, or Easy. The algorithm recalculates the next review interval based on your rating.

Your Daily Workflow

1. Create flashcards for your material. In FluentFlash, generate cards from a topic, notes, URL, or PDF using AI.

2. Review new cards and rate your recall honestly: Again (didn't know it), Hard (struggled), Good (knew it with some effort), Easy (knew it instantly).

3. Return daily to review cards that are due. The algorithm selects only cards needing review today based on their individual schedules.

4. Trust the algorithm. Resist reviewing cards not yet due, as this wastes time without improving retention.

5. Over weeks and months, your intervals lengthen. Cards you consistently rate Good or Easy appear less frequently, freeing time for new material.

The Efficiency Advantage

Most users spend 15-30 minutes daily because the algorithm constantly filters out material you already know. A card rated 'Easy' might not appear again for 30 days. A card rated 'Again' might reappear in 10 minutes. Your study time focuses only on what needs reinforcing.

Spaced repetition adoption is strongest in two areas: medical education and language learning. Medical students face thousands of diseases, drugs, pathways, and clinical associations across four years, plus board exams and clinical practice.

The AnKing deck for Anki, covering USMLE Step 1 and Step 2, contains over 40,000 cards and is used by a majority of U.S. medical students.

Language Learning

Language learners use spaced repetition to acquire vocabulary efficiently. Apps like Anki, FluentFlash, and Memrise schedule vocabulary review at optimal intervals, allowing users to learn 20-30 new words daily while maintaining previously learned vocabulary.

Beyond Medical and Language Studies

Spaced repetition is also used by law students (bar exam prep), pilots (aviation regulations), programmers (API documentation), musicians (sight-reading), and professionals preparing for certifications in finance, IT, and healthcare.