The approach
Why this app exists, who it's for, what differs.
Why this app exists
Real progress in Japanese starts when you can read. Once reading clicks, vocabulary and nuance compound naturally from contact with real text. Nihongo Dominos is a recognition-drilling layer for the three written-language building blocks — vocabulary, kanji, grammar — designed for short, frequent sessions alongside whatever else you're using.
Who this app is for
You can read hiragana and katakana, you've worked through a beginner foundation (Genki I, Minna no Nihongo, or equivalent), and you want to push from "I can read simple sentences" to "I can read everyday Japanese — newspapers, novels, blogs, manga."
The app runs on iPhone (portrait, one-handed) and iPad (landscape, with the open back-face sitting beside the grid in an ambient side panel). Your Challenge Box and study progress sync between devices through your own iCloud account.
The three building blocks
Vocabulary, kanji, and grammar share the same shape — frequency tiers — so time in one pillar reinforces the others.
Verbs, adjectives, adverbs, nouns, conjunctions, number paradigms — every entry validated against JMdict. ~98% coverage of everyday text at the top of the pool.
The intermediate core — the full JLPT N5–N2 pool unioned with the entire Kyōiku elementary set — ordered by a blended frequency score across newspaper, Wikipedia, and literature (~95% coverage of everyday reading). The complete JLPT N1 set and the non-Jōyō (表外) kanji are now included too, in the later tiers — there when you need them, but the focus stays intermediate.
One specific surface form per tile, variants consolidated. Curated then audited against Bunpro, DOJG, Genki, Tobira, Quartet. ~94–96% coverage in general texts.
Two smaller reference tiles sit below the four pillars on the home page: 部首 · Bushu (the 214 Kangxi radicals) and 仮名 · Kana (hiragana + katakana). Same domino mechanics, used as cross-reference rather than retrieval drill — see How to play for what they do.
How this app differs
Frequency-driven, not syllabus-driven
Most apps follow JLPT or textbook progression. We rank by how often things appear in real Japanese. Tier 1 is what you'll meet first; Tier 8 is what you'll meet last. If you'd rather study by JLPT level, flip the JLPT lens in the bottom control bar — same pool, re-bucketed by level. The cards lead with the intermediate range, N5 → N2; N1 and 表外 are there too, beyond the focus.
Recognition-first, not production-first
The skill we train is reading. We don't ask you to type, write, or fill in blanks.
Three pillars, one mechanic
Vocabulary, kanji, and grammar under the same tier system. The 6,800+ example sentences are individually composed, not templated: each noun example uses only nouns from the same or earlier tiers and rotates through Bunpou patterns, so studying nouns doubles as grammar exposure.
Targeted spaced repetition only where it pays off
The regular tier games are pure exposure — fast drilling, no scheduling overhead. Items you struggle with (peeked three times) flow into a Challenge Box that does use spaced repetition, then graduates them out once retained for ~2 months.
Open data, transparent methodology
Every list traces back to an authoritative open source — JMdict, KANJIDIC2, KanjiVG — or a documented curation process you can read end-to-end on the Where the data comes from page.
Credits
Created by Juergen Hammer. Multiple Japanese-language teachers have spot-checked and validated the content.