Spaced Repetition Calculator

Spaced repetition schedules each item so that it is reviewed near the point where recall probability has fallen to a chosen target, typically 85 to 90 percent. The SM-2 style model used here starts a card at a 1-day interval, moves it to roughly 3 to 6 days on the second successful pass, and thereafter multiplies the previous interval by an ease factor. The default ease of 2.5 turns a 6-day interval into 15, then 38, then 94, then 235 days; lowering ease to 1.3, the floor used in Anki, keeps a difficult card cycling every few weeks indefinitely. The steady-state workload is what most learners underestimate. Because each card returns once per average interval, daily reviews settle at roughly the deck size divided by the average mature interval, so a 5,000-card deck with a low ease factor can generate several hundred daily reviews long after the last new card is introduced. Useful ranges: 10 to 20 new cards per day is sustainable indefinitely, 50 or more is a short-term exam sprint, and 20 to 45 seconds is typical handling time per card. The model treats ease as uniform across the deck and ignores interference between similar cards, the effect of rewriting bad cards, and the fact that a small fraction of poorly formulated items absorbs a disproportionate share of total reviews.

Hermann Ebbinghaus established the empirical foundation in 1885 with Ueber das Gedaechtnis, in which he memorised lists of nonsense syllables on himself and plotted retention against elapsed time, producing the forgetting curve and the first evidence that distributed practice outperforms massed practice. The idea lay largely dormant in applied education until Paul Pimsleur published A Memory Schedule in the Modern Language Journal in 1967, proposing graduated interval recall, and Sebastian Leitner described the physical five-box card system in his 1972 book So lernt man lernen. The algorithmic era began with Piotr Wozniak in Poland, who started logging his own review outcomes in 1985 and formalised SM-2 in 1987, introducing the per-item ease factor and the recall quality rating from 0 to 5 that adjusts it. SuperMemo went through successive algorithm generations, while the open-source Anki, released by Damien Elmes in 2006, carried a modified SM-2 to a mass audience and later added the FSRS scheduler based on fitted memory models. On the research side, the 2006 meta-analysis by Cepeda, Pashler, Vul, Wixted, and Rohrer, covering 254 studies, quantified the spacing advantage and gave the practice a firm evidential base.