Turn a 2,211-kanji bank into a path you can actually finish.
Random kanji quizzes are useful for a few minutes, but they do not tell you whether you have covered the whole bank or whether one weak group keeps slipping through. Kanji Mastery divides the existing source-backed study bank into 113 bounded decks and gives every deck the same five-step learning contract: Study, Meaning Recall, Reading Recall, Reverse Recognition and a mixed Mastery Check.
Why deck mastery is different from another ten-question quiz
The inherited Japanese Kanji game already has strong individual practice modes. Meaning Recall asks you to type an English meaning for a displayed character. Reading Recall asks for a listed on-reading or kun-reading and accepts kana or common romaji. Reverse Recognition starts from an English meaning and asks you to identify the kanji. Sprint 60 adds time pressure. Due + Weak Review brings back characters that were actually missed or scheduled on this browser. The searchable Kanji Bank exposes the data itself.
Those modes solve session-level problems, but a large bank creates a second problem: coverage. If you keep asking for ten random prompts, you can have many good sessions while still not knowing which characters you have never properly tested. Kanji Mastery adds a completion structure without deleting or replacing the classic modes. The 2,211-character bank stays frozen. The Academy simply groups it into finite decks and records which checkpoints each deck has actually passed.
Decks follow the existing frozen study order within each community-derived N5–N1 band. Most decks contain 20 kanji. The final deck in a band can be smaller because the product never duplicates characters just to make the count look tidy. Every one of the 2,211 kanji belongs to exactly one mastery deck.
The five checkpoints
1. Study
Study is an encoding pass, not a mastery claim. The deck displays every character together with its meanings and accepted readings. Use this step to notice shape, meaning and reading patterns before testing yourself. Marking the deck studied only completes the Study checkpoint. It does not silently mark the recall steps correct.
2. Meaning Recall
Meaning Recall displays every kanji in the deck once and requires a typed English answer. The answer checker uses the accepted meanings already stored in the frozen Kanji authority. The checkpoint is based on first-attempt accuracy. A corrected answer after a miss lets the session continue, but it does not erase the first miss from the checkpoint score.
3. Reading Recall
Reading Recall again covers the full deck, this time asking for a listed reading. You can answer in kana or common romaji. Because Japanese kanji can have multiple readings, the goal is not to type every possible reading. One accepted listed on-reading or kun-reading is enough for that prompt. The product displays the accepted reading set after a first miss so the feedback is concrete rather than just “wrong.”
4. Reverse Recognition
Reverse Recognition starts from a primary English meaning and asks you to select the matching kanji. Distractors are drawn so the visible choices do not share the same normalized primary meaning. This direction matters because recognition from a meaning cue is different from seeing a character and recalling what it means.
5. Mastery Check
The mixed check is the point where the three retrieval directions meet. Every kanji in the deck appears at least once. The check then adds up to ten extra prompts, with directions rotated across Meaning, Reading and Reverse. A standard 20-kanji deck therefore produces 30 prompts: exactly ten Meaning, ten Reading and ten Reverse. Smaller final decks use a shorter check but still cover every character.
Passing is based on first attempts
For Meaning Recall, Reading Recall and Reverse Recognition, the pass line is 90% of the actual deck size. A 20-character deck therefore needs 18 first-attempt correct answers. A 19-character deck needs 18. A six-character final deck needs all six. This avoids pretending every deck has the same size while keeping the target strict.
The Mastery Check requires at least 90% overall first-attempt accuracy and at least 80% in each direction. The direction floor matters because a high overall total should not hide a weak reading score. On a 20-character deck, getting nearly every meaning and reverse prompt right does not compensate for consistently missing the readings.
Passing a checkpoint is persistent in local browser state. Repeating it later can improve your best score, but you do not need to clear the same checkpoint every session just to keep the deck marked complete. A deck is mastered only when all five checkpoints have passed.
Weak Kanji stays tied to real mistakes
The Academy shares one important signal with the inherited Kanji product: Weak Kanji. A character can enter or increase its Weak Kanji record only when that character was genuinely prompted and your first attempt missed. Characters that were never shown are never fabricated as misses. This is especially important in a 2,211-item bank because bulk-marking unseen characters as weak would make review data meaningless.
Correct Academy repetitions are deliberately more conservative. They improve the Academy checkpoint, but they do not artificially accelerate the inherited classic spaced-review interval. The classic Kanji modes remain the authority for normal successful SRS progression. This keeps two ideas separate: “I passed this mastery checkpoint” and “this item has advanced through the classic review schedule.”
The Academy recommendation looks for incomplete decks containing saved Weak Kanji. If one exists, that deck is prioritized. Otherwise the recommendation moves to the next incomplete deck in frozen order. The intent is to give the learner a useful next action instead of only a giant progress percentage.
How the 2,211 kanji are divided
The existing bank uses community-derived N5–N1 study assignments. These are useful learning groupings, but they are not presented as an official JLPT kanji specification. The JLPT does not publish a current official per-level kanji list for this product to reproduce. DailyBrainArc therefore preserves the attribution and boundary already shipped with the frozen dataset.
Within those study bands the Academy creates decks of at most 20 characters: N5 has 79 kanji across 4 decks, N4 has 166 across 9 decks, N3 has 367 across 19 decks, N2 has 367 across 19 decks, and N1 has 1,232 across 62 decks. Together that is 113 decks and 565 five-step checkpoints.
Why keep the source order? The bank is already a frozen authority with frequency and study-level metadata. Re-sorting it through an opaque new algorithm would make migration and verification harder and could silently change what a deck means from version to version. Keeping the inherited order makes every deck deterministic and independently testable.
Use the classic modes alongside Mastery
Kanji Mastery is not intended to swallow every existing feature. Use Sprint 60 when you want fast recognition under time pressure. Use Due + Weak Review when you want the browser to surface scheduled and actually missed characters. Use Daily Kanji for the date-stable specialist challenge. Use Kanji Bank when you need to search the source-backed reference information. Use the Academy when your goal is systematic coverage and a visible completion path.
A practical routine is simple. Work through one Study deck, attempt its Meaning and Reading checkpoints, then use Review when misses accumulate. Return for Reverse and the Mastery Check after the characters no longer feel brand new. There is no requirement to clear all five steps in one sitting, and there is no account or cloud database. Progress is stored locally in the browser.
What this tool does not claim
This is not a handwriting grader, stroke-order tutor or radical decomposition course. Those features require their own verified data and interaction contracts. The Academy does not invent components or copy proprietary mnemonic systems. It works only with the meanings, readings and metadata already present in the source-backed bank.
It also does not claim that passing a deck means you can use every kanji perfectly in real Japanese. Character knowledge is only one part of reading. Vocabulary and context matter, which is why the separate Japanese Vocabulary specialist remains useful. The mastery score here has a narrower meaning: you met the published first-attempt thresholds across the three tested retrieval directions for that frozen deck.
That narrower contract is intentional. A useful learning tool should make it clear what was measured, what was not measured and what the next practice step is.
Start with a bounded deck
If you are early in kanji study, start with N5 Deck 1 and treat Study as permission to slow down. If you already know many basic characters, you can jump to another study band or deck. Your local Weak Kanji history can then influence the recommendation after you begin using the Academy.
The important change in V0.152 is that the 2,211-character inventory now has an end-to-end map. You can see deck progress, checkpoint progress, the number of kanji inside mastered decks, and the next incomplete area that deserves attention. That makes the existing bank more useful without pretending that more raw data alone equals better learning.