On this page
Flashcards work best when reviews are spread out over time. The question is who decides when each card comes back: you with a box of paper cards, or an app with an algorithm. Both can work, and knowing how each schedules reviews helps you choose.
Why does spacing matter at all?
A 2006 review by Cepeda and colleagues in Psychological Bulletin pooled 839 assessments of spaced practice from 317 experiments. One key finding: the gap between study sessions that produced the best retention grew as the time until the final test grew. In practice, material for a test next week suits shorter gaps than material for an exam months away.
Flashcard users often get this wrong. In Nate Kornell's 2009 flashcard experiments, studying one large stack (which spaces each card out) beat studying four smaller stacks separately (which bunches repetitions together). Spacing was more effective than massing for 90% of participants, yet after the first session 72% believed massing had worked better. Spacing feels worse while it works better, which is one reason to hand scheduling to a system.
How does a paper box schedule reviews?
The classic paper method is the Leitner system, proposed by the German science journalist Sebastian Leitner in 1972. Cards move forward one box each time you get them right and go back to the first box when you get them wrong. Later boxes are reviewed less often.
Here is one simple schedule (our example, not Leitner's original, in which the size of each box's partition set the pace):
Box 1: every day
Box 2: every 2 days
Box 3: every 4 days
Box 4: weekly
Box 5: every two weeks
The scheduling is coarse: every card in a box gets the same gap, whether you found it easy or nearly forgot it.
How do app algorithms schedule reviews?
Apps set a separate gap for each card, based on how you answered it. An early rule, on which Anki's legacy scheduler is based, is SuperMemo's SM-2, described by Piotr Wozniak:
The first interval is 1 day, the second is 6 days.
After that, each interval is the previous one multiplied by an "E-Factor" (easiness), which starts at 2.5 and is not allowed to fall below 1.3.
After each review you grade your recall from 0 to 5. Easy recalls raise the E-Factor; hard ones lower it.
Any grade below 3 sends the card back to the start.
Worked example. Grade a card 4 ("correct response after a hesitation") every time and, by the SM-2 formula, its E-Factor stays at 2.5. Its reviews come after 1 day, 6 days, then 15 days, then about 38 days. Grade it 3 instead and the E-Factor drops by 0.14 each time, so the gaps grow more slowly. Fail it and you start again at 1 day.
Newer schedulers model memory more directly. Anki's manual describes FSRS as "an alternative to Anki's legacy SuperMemo 2 (SM-2) algorithm" and lets you choose a "desired retention", default 90%. It warns that above 90% "the workload increases very quickly, and above 97% the workload can be overwhelming." Research systems go further: Duolingo's half-life regression estimates the half-life of a word in a learner's long-term memory, and a 2019 PNAS study using Duolingo data found learners whose reviews followed an optimised schedule memorised more effectively than those following simpler rule-based schedules.
When do apps help, and when is paper better?
Apps help when:
you have hundreds of cards, such as inorganic chemistry facts or NEET biology terms;
you would otherwise skip reviews, since the app shows exactly what is due;
your days are irregular and you need per-card timing.
Paper is better when:
making the card is part of the learning: drawing a mechanism, a circuit or a ray diagram by hand;
your phone distracts you (see why a phone on the desk hurts focus);
your deck is small, under about a hundred cards, where a box is easy to run.
Many students use both.
A setup that works for JEE and NEET
These are our suggestions, not research findings.
Put the conditions on the card. Not just the formula, but when it applies, such as "constant acceleration only".
One idea per card. Split long reactions into separate cards for reagent, conditions and product.
Cap new cards. Start with 15 to 20 new cards a day so daily reviews stay manageable.
Do not use flashcards for problem solving. Cards hold facts and formulas. Numericals need problem sets, with spacing across chapters.
Review daily, briefly. Ten to twenty minutes most days keeps the queue under control better than one long session on Sunday.
When a card keeps failing, the cause is often a concept you never understood, not a fact you forgot. The Fermi Concept Library covers every concept in physics, chemistry, biology, mathematics and computing, each with its own lesson, which helps before you go back to the card. For writing better cards, see how to make flashcards that work.
