Forgetting curve

What is the forgetting curve?

The forgetting curve is a graph describing how the retention of newly learned information declines over time when the material is not reviewed. Plotted with time on the horizontal axis and the proportion of information still retained on the vertical axis, it falls steeply at first and then levels off, producing a characteristic curved shape that approximates exponential decay.

The concept was introduced by the German psychologist Hermann Ebbinghaus in his 1885 monograph Über das Gedächtnis (On Memory). His central observation was that the rate of forgetting is not constant: most loss happens quickly, within hours of learning, after which the remaining memory decays much more slowly. The forgetting curve is therefore less a single fixed line than a family of curves whose steepness depends on the material, the learner, and how the memory was formed and reviewed.

• Vertical axis: amount of information retained (or its inverse, the amount forgotten).
• Horizontal axis: time elapsed since learning.
• Shape: rapid early decline followed by a long, shallow tail.

Ebbinghaus's 1885 experiment and retention numbers

Working largely as his own sole subject between 1880 and 1885, Ebbinghaus memorized lists of nonsense syllables: consonant-vowel-consonant combinations such as WID or ZOF, chosen because they carried no meaning and so could not be aided by prior associations. To measure retention objectively he used the savings method: rather than asking how much he could freely recall, he measured how much less time it took to relearn a list after a delay compared with learning it the first time. A larger time saving indicated more of the original memory survived.

Across delays he recorded a steep initial drop followed by a slower decline. Commonly cited figures derived from his data show roughly 58 percent saved after about 20 minutes, around 44 percent after an hour, near 34 percent after a day, about 25 percent after six days, and approximately 21 percent after a month. These percentages should be read as illustrative of the curve's shape rather than universal constants, since they came from one trained subject memorizing meaningless syllables under specific conditions.

Ebbinghaus also fitted his results to an equation, b = 100k / ((log t)^c + k), with constants c = 1.25 and k = 1.84, where b is savings and t is time. Modern treatments often use a simpler exponential form, R = e^(-t/S), where R is retrievability, t is time, and S represents the stability of the memory.

What the curve's shape tells us about memory

The most important lesson of the forgetting curve is not the exact numbers but the geometry: forgetting is front-loaded. A large share of what is lost disappears soon after learning, while memories that survive the first day or two are comparatively durable. This implies that the timing of review matters as much as the amount, because a review delivered before the steepest part of the decline preserves far more than the same effort applied later.

Ebbinghaus's work also helped establish memory as something that could be studied quantitatively, an idea that shaped experimental psychology for the following century. His findings have held up well: a careful replication by Jaap Murre and Joeri Dros, published in PLOS ONE in 2015, had a single subject spend about 70 hours learning and relearning lists across intervals from 20 minutes to 31 days and reproduced results similar to the original. The replication also found that the curve is not perfectly smooth and most probably shows a slight upward jump starting around the 24-hour mark, consistent with the consolidating effect of sleep.

Factors that steepen or flatten the curve

The forgetting curve is not fixed. Several factors change how steeply a given memory decays, and understanding them is the basis for slowing the loss.

Material that is meaningful, well organized, or connected to existing knowledge decays more slowly than isolated or arbitrary facts, which is precisely why Ebbinghaus chose meaningless syllables to isolate raw memory from the help of meaning. Physiological state matters too: adequate sleep supports the consolidation that stabilizes new memories, while stress and fatigue tend to accelerate forgetting. Finally, how the memory was formed shapes its durability, with deeper processing and stronger initial encoding producing flatter curves.

• Meaningfulness: connected, sense-making material is retained longer than rote or arbitrary content.
• Prior knowledge: new facts that attach to an existing framework decay more slowly.
• Sleep and consolidation: rest after learning stabilizes memories.
• Stress and fatigue: tend to steepen the decline.
• Encoding depth: elaborate, effortful processing flattens the curve relative to passive exposure.
• Review: each successful, well-timed review resets the curve to a gentler slope.

Spacing and active recall as countermeasures

Two evidence-based techniques directly counter the forgetting curve, and they work best together. The first is spaced repetition, which distributes reviews across increasing intervals rather than concentrating them in a single session. This exploits the spacing effect, the robust finding that the same total study time produces better long-term retention when spread out. A large meta-analysis by Cepeda and colleagues in 2006, synthesizing 317 experiments, found that the optimal gap between reviews grows as the length of time you need to remember the material grows.

The second technique is active recall, also called retrieval practice or the testing effect: actively retrieving information from memory, for example by self-quizzing, strengthens it far more than rereading. In a well-known 2006 study by Henry Roediger and Jeffrey Karpicke, learners who repeatedly restudied a passage went on to forget about 56 percent of what they could originally recall, whereas those who repeatedly practiced retrieval forgot only about 13 percent over the same interval. Combining the two, retrieving information at expanding intervals, repeatedly resets the curve to a flatter slope and is the engine behind flashcard systems such as Anki and SuperMemo.

• Spaced repetition: review at growing intervals (for example a day, then several days, then weeks).
• Active recall: test yourself rather than rereading; the effort of retrieval is what strengthens memory.
• Combine them: retrieving at spaced intervals flattens the curve more than either method alone.

From flattening the curve to outsourcing it

Spacing and active recall fight the forgetting curve by re-strengthening memories inside the brain. A complementary strategy is to accept the curve and offload information to a reliable external store, the principle behind note-taking systems, reference libraries, and the second-brain approach to personal knowledge management. Here the goal is not to remember every detail but to be able to find it on demand.

AI memory tools and second-brain applications extend this idea by letting people save documents, photos, and notes and retrieve them later through semantic search rather than manual filing. Such tools do not flatten the biological forgetting curve; they make its effects matter less by ensuring information remains retrievable even after it has faded from memory. The two approaches are complementary: deliberate practice for the knowledge you want to internalize, and external memory for everything else.