
How Memory Works
The Forgetting Curve (and How to Beat It)
We forget new material fastest in the first hours and days, then more slowly. What Ebbinghaus actually found, what it does and does not mean, and five habits that bend the curve in your favour.
Say you are an actor with three auditions this week, each with around twenty lines. Monday's lines go onto one memory palace, a familiar route in your mind with a picture for each line at its own stop, Wednesday's onto a second, Friday's onto a third. After each audition you leave that palace alone. By the weekend the Monday palace is close to empty. The pictures have faded, and in a few more days it can take a new set of lines.
Now change one thing. Monday goes well and you get a recall, a second audition, on Friday. You walk the Monday palace on Tuesday evening and again on Thursday. Same pictures, same palace, same brain. On Friday they are all still there.
That is the forgetting curve inside one week. Leave something alone and it fades, quickly at first and then more slowly. Come back to it at the right moments, in the right way, and it stays. Actors can use the first half on purpose, letting palaces they no longer need go blank so they can be reused. For almost everything else, you want the second half.
Here is that curve: where it comes from, what the original experiment actually showed, and the handful of habits that bend it in your favour.
Where the curve comes from
The research: Ebbinghaus, 1885, and Murre and Dros, 2015
- In short: In the 1880s Hermann Ebbinghaus measured his own forgetting and found it is fastest soon after learning and slows over time. A careful replication in 2015 found the same shape.
- What they did: Ebbinghaus learned lists of nonsense syllables, letter combinations with no meaning, then relearned them after different delays and timed how much quicker the second learning was. The time he saved showed how much was still there. In 2015 Jaap Murre and Joeri Dros repeated the method with delays of 20 minutes, 1 hour, 9 hours, 1 day, 2 days and 31 days.
- Who was involved: One person each time. Ebbinghaus was his own subject. The 2015 study also had a single participant, who spent about 70 hours learning and relearning lists.
- What they found: The loss was steepest early and then levelled off. The 2015 curve closely matched the original data. It also showed an upward jump at about the 24-hour point, which the authors discuss in relation to sleep and say needs more study.
- What they concluded: The shape of the forgetting curve is reproducible.
Worth saying what kind of evidence this is. One person, twice, learning material with no meaning at all. It is a classic and it has held up, but it describes the worst case. The exact numbers do not carry over to things that mean something to you, so treat any percentage for how much you lose in a day with suspicion.
What I take from it
Two things.
The first is timing. If the steepest loss comes early, your first review has to come early too. Leave it until next week and you are not reviewing, you are relearning. Aim for the first review about an hour after learning.
The second is the reason Ebbinghaus chose nonsense. He wanted to measure memory with nothing helping it. So the curve he drew is the curve for information you have done nothing with. Understanding it, turning it into a picture, placing it on a location: each one starts you higher than that line. You still forget. You forget less, and more slowly.
That is what happened to the audition palaces at the start. The one left alone for two or three days faded enough to reuse; the one walked on Tuesday and Thursday held. Two or three days comes from practice, not from a measured study, but it is the curve doing what Ebbinghaus said it would.
What the curve does not tell you
The curve is often quoted with confident numbers: you lose 50% in an hour, 70% in a day, 90% in a week. Those figures are not a law of memory. At best they are loose readings of one man learning meaningless syllables, and the real numbers change as soon as the material means something, as soon as it is pictured, and as soon as it is recalled.
Three things it does not say:
- That everyone forgets at the same rate. How fast you lose something depends on how well it went in, how much it matters to you and what else you learn afterwards.
- That forgetting is failure. A brain that kept everything would be swamped. An actor reusing an audition palace is relying on it.
- That there is one perfect review schedule. Ebbinghaus mapped the loss. Working out the best gaps between reviews came much later, and the answer depends on when you need the material.
Five habits that bend the curve
Each of these has its own article and its own research behind it. Here is the short version.
1. Start higher. Understand it first, then turn the key ideas into vivid pictures and put them somewhere, on a memory palace or a chain, a story that links each picture to the next. Ebbinghaus measured material with nothing to hold on to. Give it a picture and a place and you begin well above his line.
2. Come back early. The first review belongs about an hour after learning, while there is still plenty to save. Five minutes is enough.
3. Space the reviews. There are a few ways to do it. A widening schedule that has been around for decades runs one hour, one day, one week, one month, three months, six months, then as needed. A flashcard box system moves each card you get right into a box you review less often, and sends it back to the first box when you miss it. Apps can schedule each card for you. The research on spacing adds one rule whichever you use: fit the gaps to when you need the material, days apart for a test next week, weeks apart for something you need next year. The Spacing Effect (Cepeda 2006) Explained goes through what that research found.
4. Make every review a recall. Close the book and pull it back out before you look. Studies on the testing effect found that recalling beats rereading over days and weeks, even though rereading feels better at the time (Roediger and Karpicke, 2006). The Testing Effect (Roediger and Karpicke 2006) Explained covers it properly.
5. Take breaks, and sleep on it. Short breaks inside a study session help, and sleep between reviews does part of the work for you. Spaced Learning: The One-Hour Lesson That Matched Four Months tells the story of one school that built a whole lesson around the breaks, and Sleep and Memory: Why the Night Does Half Your Learning explains what happens overnight.
If you want this done for you, LOCI turns your own material, a book or a YouTube video into key ideas and images, builds them into a memory palace, and has spaced retrieval practice built in, with flashcards that carry the real information on the other side of each image.
A one-week test
Try this on something you need. Twenty terms, a short speech, the key points of a chapter.
- Monday: learn half of it with pictures on a palace. Leave the other half as notes.
- An hour later, Tuesday and Thursday: recall the pictured half from memory, then check. Leave the notes half alone.
- The following Monday: test yourself on both halves.
You will see the curve for yourself. The half you did nothing with will have slid down Ebbinghaus's slope. The half you pictured and recalled will mostly still be there.
The takeaway
The forgetting curve is not a verdict on your memory. It is a description of what happens to information you give your brain no reason to keep: it fades fast at first, then more slowly. You cannot switch it off, and you would not want to. But you can choose what survives. Start high with a good picture, come back early, space your reviews to the date, and make every review a recall.
For the bigger picture of how memory is built, How Memory Works (A Practical Guide) is the place to start. For the everyday reasons things slip, read Why Do I Forget Things?
To beat the curve on material you have to use, How to Memorise a Sales Pitch takes you from a first draft to a pitch you can deliver without the script, and How to Memorise Dates and Historical Events shows you how to keep a whole timeline from sliding down the slope.
Memory map

- Leave it alone and it fades, fast at first: an unused audition palace goes blank in days. Image: an empty stage with faint outlines where the props used to stand.
- Ebbinghaus measured his own forgetting in the 1880s by relearning nonsense syllables. Image: a Victorian gentleman at a desk timing himself with a pocket watch.
- In 2015 Murre and Dros repeated it with one person and got the same shape. Image: a modern student at the same desk, holding a sloping paper ribbon that matches the gentleman's.
- It is the worst case, material with no meaning, so pictures and meaning start you higher. Image: blank jigsaw pieces tumbling down a slope while one brightly painted piece stays near the top.
- Confident percentages are not a law of memory. Image: a weather forecaster pointing at a chart that keeps changing shape.
- Forgetting is useful: it clears space, and actors reuse palaces because of it. Image: a stagehand sweeping a stage clean for the next show.
- Come back early, about an hour after learning. Image: a kitchen timer ringing beside an open notebook.
- There are a few ways to space reviews; whichever you use, fit the gaps to when you need it. Image: a wall calendar with flashcards pinned at wider and wider gaps across it.
- Make every review a recall, not a reread. Image: a student looking up from a closed book, eyes searching the ceiling.
- Take short breaks and sleep between reviews. Image: a basketball resting on an open textbook beside a bed.
Sources
- Mark Channon, How to Remember Anything (Hodder Education)
- Mark Channon, Learning Your Lines (Nick Hern Books)
- Ebbinghaus, H. (1885). Über das Gedächtnis (Memory: A Contribution to Experimental Psychology).
- Murre, J. M. J., and Dros, J. (2015). Replication and analysis of Ebbinghaus' forgetting curve. PLOS ONE, 10(7), e0120644. https://doi.org/10.1371/journal.pone.0120644
- Cepeda, N. J., Pashler, H., Vul, E., Wixted, J. T., and Rohrer, D. (2006). Distributed practice in verbal recall tasks: a review and quantitative synthesis. Psychological Bulletin, 132(3), 354 to 380. https://doi.org/10.1037/0033-2909.132.3.354
- Roediger, H. L., III, and Karpicke, J. D. (2006). Test-enhanced learning: taking memory tests improves long-term retention. Psychological Science, 17(3), 249 to 255. https://doi.org/10.1111/j.1467-9280.2006.01693.x
Loci
Build the memory palace in this article inside Loci — store your locations, link images, and rehearse with spaced repetition.
Try it in LociMemory School community
Learn directly with Mark inside the Memory School community on Skool — monthly lessons, practice, and Q&A.
Join Memory SchoolAbout the author
Mark Channon
Grand Master of Memory (1995). Author of three books on memory published by Hodder and Nick Hern. Creator of BBC1's Monkhouse's Memory Masters. Memory coach on Channel 4's Can I Improve My Memory?
Read full bio →