Where are you on your Japanese journey?
Fill in your details and we'll model your personalised learning trajectory using research-backed algorithms.
Your Japanese Learning Trajectory
Your Forgetting Curve
How quickly memories fade — and how spaced repetition fights back
How this works
Hermann Ebbinghaus (1850–1909) was a German psychologist who became the first person to scientifically study memory and forgetting. He memorised lists of nonsense syllables and tested himself at intervals — discovering that memory decays in a predictable curve.
The forgetting curve is modelled as:
R = e−t/S
where R is retention (0–1), t is time in days, and S is the "stability" of the memory — how long it takes to decay to ~37% retention.
Spaced repetition works by reviewing a word just before you'd forget it. Each successful review increases S — so the same word takes progressively longer to forget: 1 day → 3 days → 7 days → 14 days → 30 days → 90 days.
What this means for you: Without review, a new word learned today will be ~50% forgotten in 3–4 days. With Anki or a similar app, the same word can be retained at 90%+ with just a few seconds of review per year once it's "mature".
Vocabulary Acquisition Projection
Where you'll be in 5 years, based on your current study habits
How this works
The projection uses a simple but realistic model of word acquisition:
daily_new_words = (study_minutes × consistency_factor) ÷ 20
The "20 minutes per word" estimate comes from research on deliberate vocabulary study: it takes roughly 10–20 minutes of spaced practice spread over several days to reliably encode a new word family into long-term memory.
Consistency factor is a multiplier based on your days/week. Study 7 days/week = 1.0 (full efficiency). Studying 5 days/week ≈ 0.78 — not because you study less total time, but because irregular practice leads to more forgetting between sessions.
The confidence band represents ±20% around the central estimate. The optimistic scenario assumes you encounter lots of vocabulary in context (reading/listening), making each word easier to retain. The pessimistic scenario accounts for life disruptions, difficult vocabulary domains, and forgetting of rarely-used words.
The horizontal level lines are JLPT vocabulary thresholds — widely cited estimates that the real exam doesn't publish an official word list, but these are the community consensus targets.
Time to Each Level
How long to reach each milestone from where you are now
How this works
Each bar shows estimated months to reach a vocabulary milestone from your current position, based on your study inputs.
The JLPT (Japanese Language Proficiency Test) levels represent:
- N5 — Can read basic hiragana/katakana, understand very simple Japanese. Good for surviving tourism.
- N4 — Everyday greetings, simple conversations, basic kanji (~300). Watch anime with subtitles comfortably.
- N3 — Intermediate. Can handle many daily situations. ~650 kanji. Starting to read simple manga.
- N2 — Upper intermediate. Sufficient for most workplaces. ~1,000 kanji. Read news with dictionary help.
- N1 — Near-native reading. ~2,000 kanji. University level. Rare for non-native speakers.
Important caveat: These are vocabulary-only estimates. Kanji, grammar, listening, and speaking are separate skills. Real JLPT preparation also requires grammar study and reading practice. The model assumes vocabulary is your primary bottleneck, which is often (but not always) true.
The optimistic estimate assumes you hit the upper bound of the confidence band. The realistic estimate uses the central model projection.
Comprehension Estimate
How much of real Japanese you can understand right now
How this works
Based on research by Paul Nation (2001) and Batia Laufer (1989), text comprehension follows a threshold model:
- Knowing 95% of words in a text = "comfortable reading" (can guess unknowns from context)
- Knowing 98% of words = "independent reading" (rare unknown words don't disrupt flow)
- For Japanese, 95% coverage of conversation ≈ ~2,000 word families
- 98% coverage ≈ ~5,000 word families; novels/news ≈ 8,000+
The formula used here is a logarithmic curve:
comprehension% = min(98, log(vocab + 1) / log(8001) × 98)
The logarithmic shape reflects a real phenomenon: the last few percent are disproportionately hard. Going from 0 to 2,000 words buys you enormous comprehension gains. Going from 7,000 to 8,000 barely moves the needle — but those last few percent are what separates "near-fluent" from "fully fluent".
Spaced Repetition Optimal Schedule
When to review each word for maximum retention with minimum effort
How this works
SM-2 (SuperMemo 2) is the algorithm invented by Piotr Wozniak in 1987 and used as the basis for Anki's scheduling. It works by adjusting the interval between reviews based on how well you recalled the card.
After each review, the interval is multiplied by an "ease factor" (default 2.5). If you struggle with a card, the ease factor decreases, making intervals shorter. If you find it easy, the factor stays high, spacing reviews out further.
The key insight: each successful review doesn't just reset the clock — it increases how long the memory will last. A word reviewed 6 times can persist for months without review; a word reviewed once will be forgotten in days.
Daily review load: This grows with your total vocabulary. With 500 words in your deck at mature intervals, you'll have ~15–25 reviews per day. At 3,000 words, expect 60–120 reviews/day — which is why many learners separate "new cards per day" (e.g. 15–20) from "review load" in Anki settings.