The forgetting curve: what to do the day after a lecture
What the forgetting curve shows, what it doesn't, and a 20-minute routine for the day after a lecture that keeps much more of it in your memory.
The forgetting curve shows that most forgetting happens soon after you learn something, and then slows down. Hermann Ebbinghaus first measured it in the 1880s, and a modern replication found the same shape. For a learner, the lesson is about timing: the first review should come within a day, and it should be a test of what you remember, not a reread. This article explains what the forgetting curve does and doesn't show, and gives you a short routine for the day after a lecture, a class or a chapter.
The answer in brief
- The forgetting curve is steep at first and flattens over days and weeks.
- Ebbinghaus measured it with nonsense syllables, so the exact numbers don't transfer to your course; the shape does.
- The day after a lecture is the cheapest time to save what you learned.
- Recall first (write what you remember without notes), then compare with your notes.
- Each successful recall flattens the curve, so later reviews can come further apart.
What is the forgetting curve?
The forgetting curve is a graph of how much of something you have learned is still available to you as time passes. It drops fast in the first hours, then levels off. Most people know it as a downward curve with a steep start and a long, slow tail.
The idea comes from Hermann Ebbinghaus, a German psychologist who published Memory: A Contribution to Experimental Psychology in 1885. He was his own subject. He learned lists of nonsense syllables, such as made-up three-letter combinations, until he could recite them twice without error. Then, after a delay, he learned the same lists again and timed how long it took.
What did Ebbinghaus actually measure?
This is where most summaries go wrong. Ebbinghaus didn't measure the percentage of a list he could remember. He measured savings: how much less work it took to relearn a list than to learn it the first time.
In the chapter on retention over time, based on 163 tests he ran in 1879 and 1880, the savings were roughly:
| Time since learning | Approximate savings when relearning |
|---|---|
| About 20 minutes | A little over half |
| 1 hour | About half |
| About 9 hours | About a third |
| 1 day | About a third |
| 6 days | About a quarter |
| 31 days | About a fifth |
Two things stand out. First, the biggest drop happens within the first hour or so. Second, even after a month, something is left: relearning is still faster than learning from scratch.
A modern replication
In 2015, Murre and Dros repeated the experiment with one subject who spent about 70 hours learning and relearning lists after delays from 20 minutes to 31 days. The results were similar to Ebbinghaus's original data. They also found the curve isn't perfectly smooth: it seems to jump up slightly at the one-day point.
What the forgetting curve doesn't show
- It isn't a law for every kind of material. Ebbinghaus chose nonsense syllables to strip out meaning. A lecture you understand, connected to things you already know, is a different kind of memory. The numbers above won't match your experience; the steep-then-shallow shape is the part that carries over.
- It doesn't say you lose 70% by tomorrow. You will see that figure, and others like it, in study-tip graphics. They usually come from misreading savings as recall, or from no source at all.
- It isn't fixed. Each time you successfully recall something, the next stretch of forgetting is slower. That is the whole basis of spaced review.
Why the day after a lecture matters most
Put the curve and the research on recall together, and the day after a lecture becomes the best value in your study week.
- Less is lost yet. On the steep part of the curve, a little effort saves a lot. A week later, you are relearning instead of reviewing.
- Recall strengthens memory more than rereading. In Roediger and Karpicke's 2006 study, students who practised recall remembered much more a week later than students who restudied the same passage many times.
- You still know what confused you. The day after, you can still reconstruct which parts made sense in the room and which didn't. A week later, it all blurs together.
- The next lecture builds on this one. In most courses, week 4 assumes week 3. Gaps compound.
University learning centres say the same. Memorial University's guide to learning from recorded lectures, for example, advises reviewing your notes within a day or two.
A 20-minute routine for the day after a lecture
You don't need hours. This routine takes about 20 minutes for an hour-long lecture. The order matters: recall comes before you open your notes.
Write what you remember, before you look. Take a blank page and write the main ideas of the lecture from memory, in your own words. Don't worry about order or completeness. Five minutes is enough.
Open your notes and compare. Mark what you remembered, what you forgot, and anything you got wrong. The forgotten parts are what this review is for.
Answer your open questions. Look through any questions you wrote during the lecture. Answer the ones you now can, in a sentence or two, from your notes or the course material.
Turn the forgotten parts into questions. For each idea you missed in step 1, write a question it answers. These are the questions to test yourself on next time.
Write a three-sentence summary. What was the lecture about, what was the most important idea, and what is still unclear?
Test yourself once on the new questions. Cover the answers and recall them. Anything you miss now comes back first next time.
Decide when the next review is. About 3 days later, then about a week. Put it in your calendar or let your tool schedule it.
Why recall comes before notes
If you open your notes first, everything looks familiar and you will believe you remember it. Writing from memory first shows you the real gaps. It is also the step that does the most for memory, because recalling is practice, while reading is not.
A worked example: one lecture, one week
Say you watch week 3 of an introductory statistics course on Monday evening: an hour on sampling and bias. During the lecture you write 6 questions and mark 2 moments that confused you.
On Tuesday, before opening your notes, you spend five minutes writing what you remember. You get the difference between a population and a sample, and the idea of a biased sample. You have forgotten the names of two sampling methods and can't say why a convenience sample is a problem. You open your notes, mark those gaps, and answer 4 of your 6 questions. The 2 confusing moments turn out to be the same idea, so you rewatch that 3-minute stretch and write one clear question about it. Then you write three sentences of summary and test yourself once on everything. Twenty minutes, and you know exactly what you don't know.
On Friday, you answer all 8 questions from memory. You miss one. On the following Monday, before lecture 4, you spend three minutes writing what you remember of lecture 3, then start the new lecture. That is about 30 minutes of review in a week, spread over three short sittings, and almost none of it is rereading.
How the routine changes from one course to another
- A video course you take alone: do the routine before you start the next video, not after you have watched three in a row. Watching several lectures back to back leaves no time to recall between them.
- A live class: do step 1 on the way home or the same evening if you can. In a live class you can't rewind, so your notes have more gaps, and the earlier you fill them, the more you can reconstruct.
- A textbook chapter: close the book and write what the chapter said, then compare. Turn headings you couldn't explain into questions.
- Exam preparation: the routine still applies, but add a step: file each question under the exam's topics, so you can see which ones are thin. See how to study for a certification exam from its outline.
What to do after the first day
The day-after review isn't the end. It resets the curve so that the next drop is slower. Then you review again before too much is lost.
| When | What to do | How long |
|---|---|---|
| End of the lecture | Answer what you can; mark what was confusing | 5 minutes |
| Next day | Recall from memory, compare, write questions | About 20 minutes |
| About 3 days later | Answer the lecture's questions from memory | 5 to 10 minutes |
| About 1 week later | Answer them again; rewrite any you keep missing | 5 minutes |
| About 1 month later | One more pass, or before the exam | A few minutes |
The intervals are a sensible starting point, not a precise rule. Research on spacing by Cepeda and colleagues (2008) found that the best gap between reviews grows with how long you need to remember. For a schedule you can keep, read spaced repetition: a simple review schedule.
Recall the last session before the next one
There is a second moment that works as well as the day after: the minutes before the next session of the same course. Before you start lecture 4, spend three minutes writing what you remember of lecture 3. It is spaced recall, and it warms up the ideas lecture 4 is about to build on.
This is close to the Cornell method's summary step, done from memory instead of from the page. If you use Cornell notes, see Cornell notes on paper and on screen.
Common mistakes after a lecture
- Rewatching or rereading the whole lecture. It takes as long as the lecture did and mostly builds familiarity. Recall first, then go back only to what you missed.
- Rewriting notes neatly. Copying notes into a tidy version feels productive but is mostly copying. Turn them into questions instead.
- Waiting for the weekend. By then you are on the flatter part of the curve, but there is less left to save, and the next lecture has already happened.
- Reviewing without checking. "I think I remember that" is not a test. Write or say the answer, then compare.
Doing it in Soundings
When you open a new session of a course, Soundings can ask you to recall the last one first. If the session before has no summary yet, it asks, for example, Before you start: what do you remember of Day 2? You write from memory, select Save and compare, and see what you wrote then beside what you wrote in that session, so the gaps are clear. What you wrote becomes that session's summary (Use the Cornell layout and write summaries).
Each session also gets look-backs 1, 3, 7 and 30 days after it: a reminder to skim it, answer its questions from memory and mark it reviewed (Look back at a session). Answered questions become recall cards that come back at growing gaps in Review (How recall cards work). Moments you marked Confusing during the lecture wait in a list to revisit.
Common questions
How fast do we forget what we learn?
It depends on the material and how well you learned it. Ebbinghaus's experiments with nonsense syllables showed the steepest drop in the first hour and a much slower decline over the following weeks. Meaningful material behaves differently in detail, but the fast-then-slow shape is the useful lesson.
Is it true that we forget 70% of a lecture within 24 hours?
There is no good source for that figure. It usually comes from misreading Ebbinghaus's savings measure, which records how much faster relearning is, not how much is remembered. Treat any single percentage for "how much you forget" with suspicion.
When is the best time to review after a lecture?
Within a day is a good rule, because little has been lost and you still remember what confused you. Then review again a few days later and about a week later. Each review should start with recall, not rereading.
Can you flatten the forgetting curve?
Yes. Each time you successfully recall something after a gap, you forget it more slowly afterwards. That is why reviews can come further and further apart: 1 day, 3 days, a week, a month.
What should I do if I missed the day-after review?
Do it now. Write what you remember, compare with your notes, and turn the gaps into questions. You will recall less than you would have the day after, but a late review is far better than none.