The Forgetting Curve: What Ebbinghaus Actually Found, and What It Does Not Mean
The forgetting curve is real and was replicated in 2015, but it does not say you forget 70% in a day. What Ebbinghaus measured, and how to slow forgetting.
The forgetting curve shows that memory fades fastest just after learning and then more slowly. Hermann Ebbinghaus measured it in 1885 by timing how quickly he could relearn lists of nonsense syllables, and a 2015 replication found almost the same curve. It measures time saved when relearning, not the share of facts remembered, so popular claims like "you forget 70% in a day" misread it. Recalling material at spaced intervals is what slows forgetting.
The forgetting curve is one of the most shared images in education. A line drops steeply and then levels off, usually with a caption like "you forget 70% of what you learn within 24 hours".
The curve is real. It is one of the oldest results in psychology, and it was replicated as recently as 2015. The caption is wrong, and the reason it is wrong explains what the curve is actually useful for.
What Ebbinghaus did
In the early 1880s the German psychologist Hermann Ebbinghaus set out to measure memory itself. He published the results in 1885. He was his own and only subject.
To keep meaning out of the experiment, he learned lists of nonsense syllables. He had tried lines of poetry first, but found that they produced far more variable results. For each list he recorded how long it took to learn it perfectly. After a delay, from 20 minutes to 31 days, he learned the same list again and timed that too.
His measure was savings: the share of the original learning time he saved the second time round. If a list took 25 repetitions to learn and 20 to relearn a day later, the saving was 5 out of 25, or 20%.
The numbers
These are Ebbinghaus's own savings figures, as tabulated in the 2015 replication:
| Delay before relearning | Savings |
|---|---|
| 20 minutes | 58% |
| 1 hour | 44% |
| 9 hours | 36% |
| 1 day | 34% |
| 2 days | 28% |
| 6 days | 25% |
| 31 days | 21% |
The shape is the point. Most of the loss happens in the first hour. After a day the curve flattens, and even a month later relearning was a fifth quicker than learning from scratch.
Why "you forget 70% in a day" is wrong
That claim takes the one-day figure, 34%, and reads it as "you remember 34%". The figures do not support that, for three reasons:
- Savings is not recall. It measures how much faster relearning was, not what share of the list he could still produce. The two are related, but they are different measurements.
- The material was meaningless on purpose. Nonsense syllables are about the hardest thing there is to remember. Material that connects to what you already know is not lost the same way.
- It was one person learning to a fixed standard under carefully controlled conditions. It describes the shape of forgetting, not a rate that applies to every learner and every subject.
The honest version is less dramatic and more useful. Forgetting is fastest straight after learning and slows down sharply after that.
Does it hold up?
It does. In 2015 Jaap Murre and Joeri Dros published a full replication in PLOS ONE. One of the authors spent 70 hours over 2011 and 2012 learning and relearning lists of nonsense syllables at the same delays Ebbinghaus used. The curve came out very similar. The paper also compared it with two earlier replications, which showed the same pattern.
They found one detail worth knowing. The curve is not perfectly smooth. Savings go slightly up at the one-day point compared with what the trend predicts. Earlier researchers had noticed the same bump in Ebbinghaus's data and linked it to sleep.
How to slow the curve
The forgetting curve is the reason two well-established study techniques work.
Retrieve instead of rereading. In a 2006 experiment, Henry Roediger and Jeffrey Karpicke had students either reread a passage or try to recall it. Five minutes later the rereaders did better. Two days and one week later, the students who had practised recall remembered substantially more. The rereaders were also more confident, despite remembering less.
Space it out. Each time you successfully recall something, the next curve falls more slowly. Reviewing a topic soon after learning it, then again after longer and longer gaps, catches the memory before it has faded and makes it last longer each time. We cover the evidence in why rereading notes feels like studying.
This is exactly what spaced-repetition apps do. They estimate when a memory is about to fade and schedule the review just before.
What it means in practice
- Review on the same day. The steepest part of the curve is the first hour, so a quick recall session after a lesson is worth far more than one the night before an exam.
- Then lengthen the gaps: the next day, a few days later, a week later.
- Test, don't reread. Close the book and write down what you remember, then check.
- Sleep between sessions. The bump at the one-day point has been linked to sleep, and a tired brain remembers less.
The honest summary
Ebbinghaus's forgetting curve is real, and it replicated well more than a century later. It shows that forgetting is fastest immediately after learning and slows afterwards. It does not show that people lose 70% of everything within a day. That number comes from misreading a measure of relearning time, taken from one man learning meaningless syllables. The useful lesson is about timing: recall soon, then recall again at growing intervals.
Sources
- Murre, J. M. J., & Dros, J. (2015). Replication and analysis of Ebbinghaus' forgetting curve. PLOS ONE, 10(7), e0120644. doi.org/10.1371/journal.pone.0120644
- Ebbinghaus, H. (1885). Über das Gedächtnis. Leipzig: Duncker & Humblot. Translated (1913) as Memory: A Contribution to Experimental Psychology.
- Roediger, H. L., & Karpicke, J. D. (2006). Test-enhanced learning: Taking memory tests improves long-term retention. Psychological Science, 17(3), 249–255. doi.org/10.1111/j.1467-9280.2006.01693.x