Spaced Repetition for German Vocabulary: The 7-Review Schedule Behind Wortschatzmeister

You learn twenty German words on Monday. By Friday, twelve of them are gone. That is not a memory problem. It is a scheduling problem, and it has a known fix.

---

What Is Spaced Repetition?

Spaced repetition means reviewing a word at growing intervals: soon after you learn it, then a few days later, then weeks, then months. Each review lands just before you would have forgotten the word, which is exactly the moment a review does the most good.

The idea is old. In 1885 Hermann Ebbinghaus measured his own memory for lists of nonsense syllables and drew the first forgetting curve: you lose most of what you learn within the first day or two, and then the loss slows down. He also noticed the part that matters for learners. Every time you successfully relearn something, the curve gets flatter. The memory decays more slowly after each review.

graph LR
    A["Learn a word"] --> B["Forget fast\nhours to a day"]
    B --> C["Review 1"]
    C --> D["Forget slower\ndays"]
    D --> E["Review 2"]
    E --> F["Forget much slower\nweeks"]
    F --> G["Review 3 ..."]
    style A fill:#DBEAFE,stroke:#3B82F6,color:#1E3A8A
    style B fill:#FEE2E2,stroke:#EF4444,color:#991B1B
    style C fill:#D1FAE5,stroke:#10B981,color:#065F46
    style D fill:#FEF3C7,stroke:#F59E0B,color:#92400E
    style E fill:#D1FAE5,stroke:#10B981,color:#065F46
    style F fill:#FEF3C7,stroke:#F59E0B,color:#92400E
    style G fill:#D1FAE5,stroke:#10B981,color:#065F46

So the schedule follows the curve: a short gap first, longer gaps as the memory hardens.

---

What Does the Research Say?

Spacing is one of the best-replicated findings in the psychology of learning.

Cepeda and colleagues (2006) reviewed hundreds of experiments on spaced versus massed practice and found the same result almost everywhere: spreading the same amount of study over several sessions produces better long-term recall than doing it all in one block.

Cepeda and colleagues (2008) then asked how long the gaps should be. The answer depends on how long you want to remember. To pass a test next week, a gap of about a day is ideal. To still know a word months from now, the gaps need to stretch into weeks. Longer goal, longer gaps.

Dunlosky and colleagues (2013) rated ten popular study techniques by how well the evidence supports them. Only two earned the top rating: distributed practice, which is spacing, and practice testing, which is active recall. Rereading and highlighting, the techniques most learners actually use, came out near the bottom.

Language teachers got there early. Paul Pimsleur proposed a memory schedule with expanding intervals for language learners in 1967, and Sebastian Leitner turned it into the box of index cards in 1972 that every flashcard app still imitates.

---

Why the Gaps Get Longer

Every review you pass is evidence. If you still knew the word after three days, the memory is stronger than it was, so the next gap can be longer. Reviewing a word you would have remembered anyway wastes your time. Reviewing a word after you have completely forgotten it is a fresh start. The sweet spot is just before the forgetting, and that spot moves further out each time.

Robert Bjork calls this a desirable difficulty: a good review should feel a little hard. If every review feels easy, the gaps are too short.

---

How Spaced Repetition Works in Wortschatzmeister

Every word you learn in Wortschatzmeister runs on the same schedule, whether it comes from a Language Level set, an idiom set, a podcast or a custom set you built yourself. There is one scheduler in the app, and every word goes through it.

One schedule, seven passes

Pass	Day	Next review in
1. First contact	Day 0	1 day
2. Review	Day 1	3 days
3. Review	Day 4	7 days
4. Review	Day 11	14 days
5. Review	Day 25	30 days
6. Review	Day 55	60 days
7. Review	Day 115	Retired. The word is learned

graph LR
    A["Day 0\nFirst contact"] --> B["Day 1"] --> C["Day 4"] --> D["Day 11"] --> E["Day 25"] --> F["Day 55"] --> G["Day 115\nRetired"]
    style A fill:#DBEAFE,stroke:#3B82F6,color:#1E3A8A
    style B fill:#FEF3C7,stroke:#F59E0B,color:#92400E
    style C fill:#FEF3C7,stroke:#F59E0B,color:#92400E
    style D fill:#FEF3C7,stroke:#F59E0B,color:#92400E
    style E fill:#FEF3C7,stroke:#F59E0B,color:#92400E
    style F fill:#FEF3C7,stroke:#F59E0B,color:#92400E
    style G fill:#D1FAE5,stroke:#10B981,color:#065F46

A word you meet today and never fail is finished in 115 days, after seven encounters in total. That is the whole schedule. The gaps never depend on how well you scored, only on how many passes the word has behind it.

What counts as a pass

Every encounter is an attempt scored from 0 to 100, and 50 or more is a pass.

• In Full practice, the score is the pronunciation score the AI gives your recording of the word's sentence. That is the shadowing step.
• In Quiet practice, where you use no microphone, the score comes from the word-recall game: 100 for a clean solve, 70 if you needed hints or made a mistake, 0 if you ran out of guesses.
• The written steps, translating the sentence, producing it from your language, writing your own sentence, give you feedback and corrections. They never fail the card.

What happens when you fail

Nothing is taken away. The word stays on the same step, stays due today, and goes to the back of the same session, up to two more times. If it still does not stick, it is waiting for you tomorrow. Once you pass, the ladder continues exactly where it left off.

That is a deliberate difference from the classic algorithm behind most flashcard apps. In SM-2, which Anki is built on, a lapse sends the card back to the start and permanently lowers its "ease", so one bad evening can follow a word around for months. Anki users call it ease hell. We decided a fail should cost you one more attempt, not your history.

Why a fixed table and not an adaptive algorithm

Adaptive algorithms adjust every card's interval from your score. We chose a fixed table on purpose.

• It is predictable. You always know when a word comes back.
• The score has one job. It decides pass or fail and nothing else. A brilliant pass never buys a longer gap and a scraped pass never shortens one, so there is no reason to grade yourself generously and no way to game the queue.
• One scheduler for everything. A due count means the same thing for idioms, podcasts and Language Levels.

The trade-off is honest: the table does not know that der Löffel is harder for you than das Haus. The fail path handles that. A hard word simply takes more attempts on the same step, and every attempt is a real act of retrieval, which is where the learning happens anyway.

Learning has an end

Most review systems never stop. Cards come back at 200 days, 400 days, 800 days, forever. In Wortschatzmeister, the seventh pass retires the word. It is never scheduled again. Everything you earned stays in your record, the word counts as learned, and if it shows up in a text or a conversation you will simply read it. That is the point of the whole exercise.

Where you see it

• Your dashboard shows how many words are due in each family.
• Every word carries its next review date.
• The Review area of each family, Language Levels, Idioms, Podcasts and Custom sets, is the due queue. Open it and work through it.

---

Tips for Getting the Most Out of It

• Review on the day. A missed day breaks nothing, the word simply waits. But a pile of overdue words makes a bad session, so short daily visits beat a weekend marathon. That is spacing too.
• Add new words in small batches. Every new word comes back after 1, 3 and 7 days, so twenty new words a day turns into sixty reviews a day within a week. Ten to fifteen new words a day is a comfortable pace.
• Never peek. A pass you did not earn schedules the word weeks out when it should have come back tomorrow. Use the reveal button and take the fail. It costs you one attempt.
• Quiet practice counts. A review on the bus without a microphone is a real review and a real pass.
• Let words retire. Do not restart a set to "refresh" it. If a retired word turns up and you have forgotten it, that is rare, and reading it once will usually bring it back.

---

Try It Now

The first set of every Language Level, A1, B1, B2 and C1, plus the first two sets of A2, are free for any signed-in account, about 400 words in total. Free means the whole ladder: first contact, all six reviews and retirement.

1. Create a free account.
2. Open Language Levels, pick your level and start Practice on set 1.
3. Come back tomorrow. The words will be waiting.

New to the app? How to use Wortschatzmeister walks through everything else.

References

• Ebbinghaus, H. (1885). Über das Gedächtnis: Untersuchungen zur experimentellen Psychologie. Duncker & Humblot.
• Cepeda, N. J., Pashler, H., Vul, E., Wixted, J. T., & Rohrer, D. (2006). Distributed practice in verbal recall tasks: A review and quantitative synthesis. Psychological Bulletin, 132(3), 354-380.
• Cepeda, N. J., Vul, E., Rohrer, D., Wixted, J. T., & Pashler, H. (2008). Spacing effects in learning: A temporal ridgeline of optimal retention. Psychological Science, 19(11), 1095-1102.
• Dunlosky, J., Rawson, K. A., Marsh, E. J., Nathan, M. J., & Willingham, D. T. (2013). Improving students' learning with effective learning techniques. Psychological Science in the Public Interest, 14(1), 4-58.
• Bjork, R. A. (1994). Memory and metamemory considerations in the training of human beings. In J. Metcalfe & A. Shimamura (Eds.), Metacognition: Knowing about knowing (pp. 185-205). MIT Press.
• Pimsleur, P. (1967). A memory schedule. The Modern Language Journal, 51(2), 73-75.
• Leitner, S. (1972). So lernt man lernen. Herder.

Have questions about your review queue? Write to us at hello@wortschatzmeister.de. Happy reviewing!