Why flashcards work — when they're used properly
Flashcards are effective for a specific reason, and it is not repetition. It is retrieval. The act of trying to produce an answer from memory, before you see it, changes memory in a way that re-reading the answer does not. This is one of the most replicated findings in the study of learning, usually called the testing effect.
The practical consequence is that how you use a card determines whether it does anything at all. Reading the front, immediately flipping it, and nodding at the back is re-reading with extra steps. Pausing to attempt a full answer — out loud, or written down — and only then checking, is retrieval practice. The two feel similar and differ enormously in effect.
Notably, retrieval still helps when you get the answer wrong, provided you check immediately afterwards. The failed attempt appears to prime you to encode the correction. Getting it wrong is not wasted effort; getting it wrong without checking is.
Writing cards that are worth reviewing
Most poor flashcard decks fail at authoring, not review. Some rules that hold up:
One fact per card. A card asking "Describe the causes of the First World War" cannot be graded — you will always half-know it, and you will never resolve it into known or unknown. Break it into specific, answerable pieces. If you cannot say definitively whether you got a card right, the card is too big.
Make the front a genuine question. A bare term as a prompt tests recognition. "What does X mean, and when would you use it?" tests something closer to what an exam asks.
Write cards from material you already understand. Flashcards consolidate knowledge; they do not build it. Making cards from a topic you have not yet grasped produces a deck of memorised strings you cannot apply. Understand first, then card.
Test both directions where it matters. Recognising a foreign word is not the same skill as producing it. For vocabulary, make cards both ways — production is harder and usually what you are assessed on.
Prefer your own words. Cards written in your own phrasing are recalled better than cards copied verbatim from a textbook, and writing them forces the comprehension check that copying skips.
What flashcards are bad at
Cards suit discrete, factual material: definitions, vocabulary, formulas, dates, classifications, symbols. They are poor at extended reasoning, essay structure, multi-step problem solving and anything requiring synthesis across topics.
For those, work past-paper questions instead. Building a deck for an essay subject is one of the most common ways students spend a great deal of time on revision that does not transfer to the exam.
Spacing: the scheduling half
Retrieval is one half; when you retrieve is the other. Given a fixed amount of review time, spreading it across sessions produces substantially better long-term retention than concentrating it. Reviewing a card twice in a row is nearly wasted the second time because nothing has been forgotten yet — and the forgetting is what makes the return trip effective.
A workable pattern for exam preparation is to review new cards after about a day, then three days, then a week, then two weeks, then a month. The exact intervals matter far less than the principle of expanding gaps.
You can run this manually with three piles — struggling, getting there, known — reviewing the first pile every session, the second weekly, and the third occasionally. The mechanism is what matters, not the software.
A review routine that works
- Read the prompt and stop. Do not flip yet.
- Produce the whole answer, aloud or written. Partial credit does not exist here — either you produced it or you did not.
- Check, and be strict. "I basically knew that" is how cards get marked known while remaining unknown.
- Sort honestly by whether you produced it, not by whether it felt familiar.
- Shuffle between sessions. Fixed order lets you learn the sequence rather than the content — a real and easily missed failure mode.
- Mix topics in one session. Interleaving forces you to work out which kind of question you are looking at, which is usually most of the difficulty in an exam.
Pair this with our Pomodoro timer to bound sessions — card review degrades quickly past about 25 minutes. Our guide to revision techniques that work covers retrieval, spacing and interleaving in more depth.
Common questions
Where are my decks stored?
In your own browser's local storage, on your device. Nothing is uploaded — there is no account system and no server holding your decks. That means they persist between visits in the same browser, but they do not sync to other devices, and clearing your browser data deletes them.
Can I get my decks onto another device?
Use the export option to save a deck as a file and import it on the other device. Because there are no accounts, file transfer is the mechanism — a deliberate trade of convenience for not holding your data.
How many cards should a deck have?
Small enough to complete in one sitting: 20 to 40 for a topic works well. Very large decks tend to be reviewed inconsistently, which defeats the spacing that makes them effective. Several focused decks beat one enormous one.
Should I make cards or use someone else's deck?
Make your own where you can. Authoring is itself a learning activity, and it forces you to decide what matters — a shared deck imports someone else's judgement about that, and often someone else's syllabus. Shared decks are reasonable for standardised material like vocabulary or anatomy.
Is quiz mode different from flashcard mode?
Quiz mode presents questions with scoring, which is closer to exam conditions and useful for checking readiness under mild pressure. Flashcard mode is better for building recall in the first place. Use cards to learn, quizzes to test.