I am learning Japanese Kanji via Remembering the Kanji, but this question could apply to any vocab or grammar flashcard sets.

I know that programs such an Anki have an algorithm that decides the best spaced repetition on a card-by-card basis, but what is the best feasibly spacing pattern for physical flash cards? I specify feasible, because tracking the history of each of thousands of flash cards individually would take too much time and space.

I am ~2,100 Kanji in, and have been using an SRS based on the Fibonacci sequence. Newly created flash cards get put in a "0-day" pile and will be reviewed the same day. Any I get correct become a new "1-day" stack dated to practice tomorrow. Tomorrow, I will review them and those I get correct become a "2-day" stack that gets reviewed after 2 days- and so on, for 0>1>2>3>5>8>13>21>34>55>89 etc. So every day I review up to 11 different stacks of cards, with those I get correct being promoted and those I fail being added to that day's "0-day" pile.

Is there a better SRS sequence? I find I loose a lot of cards to the 34>55 gap and very few to the 1>2 gap, but I hesitate to switch to some other arbitrary sequence unless there is evidence it is more effective. Is there an entirely different organization method that will work without making me keep track of hundreds of separate flashcard bundles?

  • Awesome. I would do like that and I also thought about Fibonacci sequence for srs. I don't think there is a strong backed by science theory. But since you experienced loosing a lot of cards to the 34>55 gap and very few to the 1>2 gap, why wouldn't you omit >2> step and add some, say, >45> step? Commented Oct 6, 2021 at 16:59


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.