The Ebbinghaus forgetting curve: when should you review?
Understand the forgetting curve without universal retention percentages. See a practical review routine and the limits of fixed review dates.
The forgetting curve describes how retention can decline as time passes after learning. It is not a personal countdown that reveals exactly when you will forget. Use it as a reason to plan later reviews, then check what you can actually recall.
knowloop · Updated
What the curve can tell you
Ebbinghaus studied learning and relearning under controlled conditions. Murre and Dros later replicated and analysed this work. A curve from such an experiment describes results under those conditions; it is not a forecast for every learner.
Your familiarity with a topic, the material and the way you practise matter. Do not interpret a stock chart as proof that you personally retain a particular percentage after one hour or one day.
Is there a best review time?
There is no single clock time or universal interval that this website can prescribe. A manageable starting approach is to review once soon after learning, then return on later days with concrete questions.
Knowloop’s default is 20 minutes and 1, 2, 4, 7, 15 and 30 days after learning. Treat those dates as a planning choice. The generator below turns your start time into actual dates; it does not measure memory.
Turn a vague review into an observable task
Suppose you learned the difference between a primary key and a foreign key. Instead of writing ‘review databases’, ask: ‘Which key identifies a row, and which key refers to another table?’
At the scheduled time, answer without the note, then check the definitions and a concrete table example. If the answer is incomplete, practise the missing distinction. The result tells you more about this topic than a generic retention percentage.
How to read the chart on Knowloop
The home-page diagram is illustrative. Its curves are not measurements of your progress, its time labels are not equally spaced on the horizontal axis, and the upward turns do not promise a specific memory gain.
Use your completed reviews and card feedback to track what you actually did. A useful learning record says which question you could answer, rather than estimating an unsupported memory score.
Use the plan alongside your deadline
For a short deadline, first list the topics you need and the dates you have available. Check that important material gets a real retrieval attempt before the exam; a reminder arriving after the exam cannot help preparation.
For long-term learning, keep items small enough that you can review them during an ordinary day. Start with a few topics and adjust your workload based on completed sessions.
Research references
Common questions
Will I forget exactly half of everything after an hour?
A universal percentage cannot describe your individual learning. Knowloop does not calculate or display a personal retention prediction.
Are the seven review times Ebbinghaus’ proven best intervals?
No. They are Knowloop’s fixed planning defaults. Research on forgetting motivates review, but does not establish these exact offsets as optimal for everyone.