Overview
Mapped by Hermann Ebbinghaus in the 1880s through self-experimentation, the forgetting curve describes the rate at which newly learned information decays from memory over time. The core finding: without any review, approximately 80% of new information is lost within 48 hours. The curve is steep initially, then flattens — but without intervention, even well-encoded material degrades toward zero.
It is the scientific foundation for Spaced Repetition.
The Curve
- Day 0 (immediately after learning): ~100% retention
- Day 1: retention drops sharply — roughly 60% remains without review
- Day 2: ~40% remains
- Day 7: ~20–25% without any reinforcement
- Month 1+: minimal retention of a single-exposure memory
The exact figures vary by material type, meaningfulness, and individual — but the shape (exponential decay) is robust across studies.
The Spacing Effect — Flattening the Curve
Each time you successfully retrieve a memory before it fully decays, two things happen:
- The memory is reconsolidated and strengthened
- The next forgetting curve for that item is shallower — it decays more slowly
Repeated spaced retrieval eventually produces a very flat curve: the item is effectively permanent and needs only occasional maintenance reviews. This is the mechanism behind spaced repetition systems like Anki.
The optimal moment to review is just before forgetting — which is exactly when it feels hardest to recall. That difficulty is productive; it creates the strongest reconsolidation signal.
Key Implications
- One session is never enough — even an excellent study session produces a forgetting curve. Without spaced follow-up, it decays like any other exposure.
- Cramming works short-term, fails long-term — a massed study block before an exam delays the curve briefly but doesn’t flatten it. The material disappears quickly post-exam.
- “I learned this once” is not durable — frequency of spaced retrieval, not initial depth of learning, determines long-term retention.
- Time between sessions matters — the same amount of study time distributed across days outperforms the same time in a single block, dramatically.
Historical Note
Ebbinghaus used himself as the sole subject of his experiments — memorizing thousands of nonsense syllables over years. The rigor was remarkable for the 1880s. His work established memory as a measurable, scientific phenomenon rather than a philosophical abstraction. → Hermann Ebbinghaus
Contradictions / Open Questions
- The forgetting curve was derived from nonsense syllables — meaningful, emotional, or personally relevant material has a different (flatter) decay rate. How much does this change the quantitative claims?
- Individual variation is large and not well-accounted for in the original research.
Related
- Topics: Spaced Repetition · Active Recall · Fluency Illusion
- People: Hermann Ebbinghaus
- Resources: How I Actually Retain What I Study (The System That Replaced All My Notes)