Understand the method
Spaced repetition: review just before you forget.
Why we forget, why reviewing at the right moment changes everything, and how to apply the method, with or without an app. Sources are cited at every step.
Ebbinghaus, 1885 Testing effect Your rhythm What is spaced repetition?
Say you have thirty minutes to study. You can spend them in one go the night before the test, or split them into three ten-minute sessions across a week. Same total time, but the second option leaves a far more durable memory. That's the whole idea: you don't study more, you study at better-chosen moments.
The method comes down to two habits. The first is spacing: letting time pass between two reviews of the same idea, and stretching that time as it settles in. The second is testing yourself: each time, trying to pull the answer from memory instead of rereading your notes. The two reinforce each other, and it's the combination that research supports best.
The ideal moment to revisit an idea is just before it fades. Too early, and the review teaches you almost nothing: the answer is still there, no effort needed. Too late, and you have to relearn it from scratch. In between, recall takes real effort but succeeds, and that effort is what strengthens the memory.
Ebbinghaus's forgetting curve, and what we know now
In 1885, the German psychologist Hermann Ebbinghaus published Über das Gedächtnis ("On Memory"). He used himself as the subject: he learned lists of meaningless syllables, then measured how long it took to relearn them after a few minutes, hours or days.
His finding has held up for well over a century. Forgetting is fast right after learning, then it slows down. And each round of relearning makes the next drop gentler: you forget less, and more slowly.
The experiment had limits, and they're worth stating. One subject, made-up syllables: a course that makes sense fades more slowly than a list like "DAX, BUP, ZOL". But the shape of the curve, a steep fall followed by a plateau, is still where all research on spacing begins.
What the research says
Why spacing and self-testing work: five studies worth knowing.
Experiments run in labs and classrooms, published in peer-reviewed journals. They are the foundation of the method, and what Mindlet builds on.
- Testing effect
Testing beats rereading
Students read a passage, then either reread it or tested themselves on it. A week later, those who tested themselves remembered more, even though the rereaders felt more confident.
Roediger, H. L. & Karpicke, J. D. (2006). Test-enhanced learning. Psychological Science, 17(3), 249-255.
- Repeated retrieval
Retrieval is what makes it stick
Once a word has been recalled, studying it again adds almost nothing to long-term memory. Continuing to retrieve it from memory does.
Karpicke, J. D. & Roediger, H. L. (2008). The critical importance of retrieval for learning. Science, 319, 966-968.
- Spacing
Spacing beats cramming
This meta-analysis pools hundreds of comparisons. Spaced study outperforms massed study, and the ideal gap between sessions grows with how long you need to remember.
Cepeda, N. J., Pashler, H., Vul, E., Wixted, J. T. & Rohrer, D. (2006). Distributed practice in verbal recall tasks. Psychological Bulletin, 132(3), 354-380.
- Personalised review
+16.5% retention
Middle school students learned Spanish vocabulary over a semester. Reviews scheduled by a model of each student's memory gave 16.5% more retention than massed study.
Lindsey, R. V. et al. (2014). Improving students' long-term knowledge retention through personalized review. Psychological Science, 25(3), 639-647.
- Scheduling
An algorithm that learns from your answers
The authors describe a scheduler that estimates, for each memory, how strong it is and how fast it fades, based on millions of real reviews. It's the family of algorithms Mindlet uses.
Ye, J., Su, J. & Cao, Y. (2022). A stochastic shortest path algorithm for optimizing spaced repetition scheduling. KDD '22, 4381-4390.
Fixed schedules, Leitner boxes, algorithms: three ways to space
There are several ways to organise reviews over time. They all apply the same principle; they differ in how finely they adapt to what you already know.
| Method | How it works | What it does well | Its limit |
|---|---|---|---|
| Fixed schedule | Review a chapter on day 0, 1, 3, 7, 15, 30, then monthly. Very common among French medical students, who call it the "J method". | Simple, no tools needed, easy to put in a calendar. | Same dates for everything: you over-review what you know and under-review what resists. |
| Leitner boxes | Cards sit in several boxes. A card you get right moves to the next box, reviewed less often; a card you miss goes back to the first. | Accounts for mistakes, card by card, with paper. | Each box's interval is fixed in advance and says nothing about your memory. |
| Adaptive algorithm | A calculation estimates, for each card, when you're likely to forget it, and reschedules it after every answer. | Every card gets its own rhythm, adjusted continuously. | You need a tool, and honest answers for it to get things right. |
What adaptive spacing changes
In Mindlet
Mindlet learns your forgetting rhythm, card by card.
For each card, Mindlet estimates two things: how well you know it today, and how fast that memory fades. It brings the card back when your chance of remembering it drops towards 90%. That's the default: high enough to retain, low enough that you don't keep reviewing what you already know.
At first, the calculation starts from average values. Then, every week, Mindlet re-tunes its parameters using all your answers: your memory isn't anyone else's, and the schedule ends up fitting you.
For the curious, the calculation is called FSRS-6, the latest version of the algorithm described by Ye, Su and Cao in 2022. In July 2026 it replaced SM-2 in Mindlet, an algorithm designed in 1987 with fixed rules, the same for everyone.
- Every answer reschedules the card, not just the chapter
- Parameters re-tuned weekly on your own reviews
- Nothing to configure: it runs in the background
Oubli, Flou, Acquis: what the three buttons change
The app interface is in French. On a flashcard, you say whether you knew it, with three buttons: « Oubli » (forgot), « Flou » (fuzzy) and « Acquis » (got it).
Active recall and the testing effect: the other half of the method
Spacing isn't enough if every review means rereading. Rereading feels like mastery because the text looks familiar; but recognising an answer isn't the same as retrieving it. Roediger and Karpicke showed this in 2006: students who reread felt better prepared, and remembered less.
Testing yourself forces memory to go and fetch. A flashcard, a multiple-choice question, a fill-in-the-blank, an open question: each format is a different way of making you retrieve the answer. That's why Mindlet varies the formats for the same idea, and why its hints during a review point you in the right direction without giving the answer away.
A mistake isn't a failure. Searching, getting it wrong, then seeing the right answer leaves a stronger trace than reading the correction straight away. The cards you miss are the ones that teach you most.
How long per day, for how many weeks?
There's no magic duration, but there is a rule: a little, often. A short session every day beats one long session a week, because it lands at the right moment for more of your cards.
In Mindlet, you set the session length to short, medium or long. Each day, the app shows you the cards that are due. On days you can't make it, nothing is lost: the cards wait, and the calculation accounts for the time that has passed.
As for the calendar, start early. Cepeda and colleagues found that the further away the exam, the wider the ideal gap between reviews. For an exam in six weeks, starting six weeks ahead leaves room to space things out; starting three days ahead doesn't.
-
Short sessions
A short format you keep up every day beats a long one you skip.
-
One reminder a day, at most
At the time you choose, or the one Mindlet learns from your habits. Nothing during quiet hours.
-
Start early
The further away the deadline, the more your reviews can spread out, and the lighter each one feels.
An example: one chapter, from day 0 to the exam
On Monday you come out of a cell biology lecture. That evening you drop the handout into Mindlet, which turns it into cards: definitions, the stages of mitosis to put in order, a few multiple-choice questions. You do a first session and miss a good share of them. That's normal: it's the first time you've had to pull these answers from memory.
The next day, the missed cards come back; the ones you knew wait a few days. By the end of the week the session is shorter: only what resists is left. In week two, the easy cards barely show up; the hard ones come back more often, then spread out in turn.
Three weeks later, the night before the exam, you don't need to reread everything. A practice exam tells you what holds and what's still fragile, and your last session focuses on those points. The work was done in small doses, day after day, right when it was useful.
Even without Mindlet
Spaced repetition with a pencil and a calendar.
Nobody owns the method: three habits are enough to apply it with paper cards or any tool you like.
-
Turn the course into questions
One card, one precise question, one short answer. "Define mitosis" works better than "Chapter 3".
-
Test yourself before rereading
Hide the answer, really try, then check. Mark what resisted.
-
Schedule the returns
Review on day 1, 3, 7, 15, 30. Bring missed cards forward, push known ones back.
Common mistakes
- Rereading instead of testing. It's comfortable, and it's the least effective.
- Cramming the night before. Massed study works the next day, and leaves almost nothing a month later.
- Grading yourself too kindly. A generous « Acquis » pushes the card away at the wrong time.
- Cards that are too long. A card that asks you to recite a paragraph will always be partly missed. Split it.
- Memorising without understanding. Spacing helps you retain what you've understood. It doesn't replace the explanation.
- Giving up after a busy week. Just pick up where you are: a few days behind is easy to catch up.
What the research says, and its limits
The evidence is strong for retaining facts, vocabulary, definitions, dates and mechanisms. It's thinner for complex skills such as problem solving or essay writing, where spacing helps but practice matters most.
The 16.5% figure comes from one specific study (Spanish vocabulary, one semester, middle school students). It shows what personalised review can bring; it doesn't promise the same gain for every subject or every learner.
Finally, spaced repetition is for remembering. Understanding is still your job, and your teachers' and your courses'. Mindlet prepares the cards and picks the moment; you make the effort to remember, and that effort is what does the learning.
FAQ
Frequently asked questions
What is spaced repetition?
It's a learning method where you revisit an idea at growing intervals, just before you would forget it, trying each time to retrieve the answer from memory. It builds on the forgetting curve described by Ebbinghaus in 1885 and on many studies, including the meta-analysis by Cepeda et al. (2006).
What are the best spaced repetition intervals?
A common fixed schedule reviews a chapter on the day you learn it, then on day 1, 3, 7, 15 and 30, then monthly. Treat these dates as a starting point: bring forward what resists, push back what you know. An adaptive tool does this for you, card by card.
What is the difference between FSRS and SM-2?
SM-2, designed in 1987, applies fixed rules that are the same for everyone. FSRS estimates, for each card, how strong the memory is and how fast it fades, and tunes its parameters to your own answers. Mindlet has used FSRS-6 since July 2026 and re-tunes those parameters every week.
Does spaced repetition work for maths?
Yes for what needs remembering: definitions, theorems, formulas, methods to recognise. For problem solving, spacing helps too, but it doesn't replace working through exercises. The best approach combines both: cards for the tools, exercises to use them.
Which spaced repetition app should I use?
It depends on what you start from. If you write your own cards, many tools will do. If you start from courses, PDFs or videos, Mindlet generates the cards for you, then schedules reviews with FSRS-6. It also imports Anki decks. The app interface is in French; content can be in English. A free plan lets you try.
How long should I review each day?
There's no universal ideal. The useful rule is consistency: a short session every day beats a long one each week. In Mindlet you choose a short, medium or long session, and the app only shows you the cards that are due.
Keep reading
Let Mindlet pick the right moment.
Drop in a course: Mindlet turns it into cards, then shows them to you right when you were about to forget.
Free to start, no credit card · Web, iPhone, iPad and Android
